This page contains discussions that have been archived from Village pump (technical). Please do not edit the contents of this page. If you wish to revive any of these discussions, either start a new thread or use the talk page associated with that topic.
I want to use a reference in a source that has page numbers for virtually all the text - but not for the "Authors note" at the beginning, which has some useful information. If I were referring to any other page in the book, the ref would look like[1]: 99 (for page 99). I am sure that there must be a way of stating "author's note" instead of a page number, but cannot find it. Any ideas? ThoughtIdRetired (talk) 21:12, 18 February 2021 (UTC)
Thanks, that would seem to work, but has a problem if the same source is needed elsewhere in the article (which is highly likely). That means I would have to use the full reference a second time (giving it a slightly different refname) and use the {{rp|99}} code for the page numbers. The problem is that it would be a mess for the reader, as they cannot track all the parts of the article that rely on the one book.
@Izno:I feel that shortened footnotes make it more tedious to work out where the text of an article comes from. You might find the reference more easily, but it is difficult to get a view of how much of the article relies on that work, and it is slower to get back to the point you had read to in the article. I have done quite a bit of trying to work out where some articles are sourced and therefore this is the voice of experience. It's a particular issue if several authors have written more than one work on the subject - I have to make written notes in complex cases. So, the citation method I have chosen may not be pretty, but it scores better on functionality. Obviously, just a point of view.... ThoughtIdRetired (talk) 08:10, 19 February 2021 (UTC)
@ThoughtIdRetired: Typing <ref name="Greenhill 1988">{{cite book |last1=Greenhill |first1=Basil |title=The Merchant Schooners |date=1951 |publisher=Conway Maritime Press |location=London |isbn=0 85177 475 X |edition=1988}}</ref>{{rp|author's note}} gives[1]: author's note . I find shortened refs abominable, I find articles strewn with errors because people get them wrong. DuncanHill (talk) 02:25, 20 February 2021 (UTC)
@Izno, ThoughtIdRetired, and DuncanHill: Just to point out the opposite point of view and what I consider to be the advantages of {{sfn}}s. Using {{sfn|Greenhill|1988|p=47}} adds reusable refs to the reflist without having to resort to the <ref name="Greenhill 1988 p47"> malarkey :) which one often comes across. All refs to the same page are automatically grouped together, and you only have to change the page number(s) in the {sfn} to achieve this.[1] Plus, you only have to hover your mouse over the {sfn} to see the basic ref, and hovering again or clicking on it takes you to the {cite book} in the Bibliography, and the back button returns you to where you were.[1] With the even easier <ref>Greenhill 1988 p. 47</ref>, just as commonly found, you have to physically scroll down to the Bibliography and scroll all the way back up again to find where you were.
I agree that screwed-up {sfn}s are annoying: often that's because the editor didn't include |ref=harv in {{cite book}} etc. The few examples I have come across have mostly been of this type. However, you now don't need it any more: CS1/2 copes with it automatically, and a bot has been removing |ref=harv from {cites} for a while.[2] Since I use {sfn}s all the time I find fixing them only takes a few moments, and I haven't come across a bad one for some time. I also agree that {sfn}s and {cite book} can take longer to learn than <ref name="" />, and more slightly more time-consuming to use in an article. Last time I looked the relevant information was scattered over several Help pages with not much attempt to construct a single comprehensive instructive page about using just {sfn}s and {cite book} etc. Things may have improved. Would anyone know how many articles actually use {sfn}s? I know it's a minority.
In order to use sfns properly, I believe the {cite book} params have to be correct as well: the actual edition cited should be in (parentheses), and the original edition in [square brackets]. I would tend to use |year= and |orig-year= which displays slightly differently: {{cite book |last1=Greenhill |first1=Basil |title=The Merchant Schooners |year=1988 |orig-year=1951 |publisher=Conway Maritime Press |location=London |isbn=0 85177 475 X}}.[3] You can use |ref={{harvid|Greenhill|1988a}} in {cite book} and {{sfn|Greenhill|1988a|p=}} to distinguish different books published in the same year.[4] You can use |loc= as in {{sfn|Greenhill|1988a|loc=Author's note}}[5] Obviously reffing is very personal and can be contentious. I started a section on my talk page with a view to gathering people's ideas and possibly writing an Essay. If anyone would like to contribute, let loose with their gripes etc. with reasons, please feel free. MinorProphet (talk) 05:39, 20 February 2021 (UTC)
Greenhill, Basil (1988). Pleasure Schooners. London: Conway Maritime Press. ISBN 0 85188 478 X
As a note, you do not need to use |ref={{harvid|Greenhill|1988a}} in {cite book}, just set year=1988a: {{cite book |last1=Greenhill |first1=Basil |title=Pleasure Schooners |year=1988a |publisher=Conway Maritime Press |location=London}}, which outputs the necessarily disambiguated citeref: CITEREFGreenhill1988a. --Izno (talk) 07:40, 20 February 2021 (UTC)
@MinorProphet: When building a {{cite book}} (or similar) with a view to using {{sfn}}, the only parameters that matter for the {{cite book}} are: |last1= (or |last= if there is only one author), |last2=, |last3=, |last4= (as many as are applicable) and |date=; other parameters such as |edition= and |orig-date= do not contribute to the sfn anchor. As regards suffix letters for years, see refs 25, 50, 51, 55 at Reading Southern railway station. --Redrose64 🌹 (talk) 11:50, 20 February 2021 (UTC)
@Redrose64: Hi, I think I was aware that {{sfn}} only needs |last=etc. and |date=, but my wording was clumsily expressed. What I was trying to say was that since you can use {{cite}} params in various ways not necessarily envisaged by their authors, are |edition=1988 and |year=1988 equally valid methods for displaying what amounts to the same concept? In other words, is there (or should there be) one and only one way to use eg {{cite book}}? There will probably be both purists and hackers. Also, it's very much a personal choice since everyone seems to have developed their own favourite style of cite, regardless of the many very specific Chicago/Harvard/Oxford/CS1|2 styles. Thanks also for the helpful pointers to Reading SR station - coincidentally I was at school there in the 1970s, and remember well the old GWR station with the original entrance building. >MinorProphet (talk) 13:06, 20 February 2021 (UTC)
At present, if you are not specifying the day and month you can use |year= as an alternative to |date=; but |year= is not guaranteed to work forever - there are those who would seek to eradicate all parameter alternatives for the cite templates. If you don't stick to what's in the documentation, it may not work and you will probably get a bot trampling all over your edit. --Redrose64 🌹 (talk) 13:31, 20 February 2021 (UTC)
I seem to remember a very long time ago that I encountered some difficulty with using |date= with only a year. It would have been when I first started trying to understand {{cite book}} coupled with {{sfn}} etc.; perhaps merely an initial misunderstanding. Anyway I've used |year= ever since with books, and |date= with {{cite journal}} and {{cite news}} but I really don't care what the param is, even |go-forth-and-multiply=...(lol?) If |date= takes any combination of DDMMYYYY etc., I'm quite happy to use it. @ThoughtIdRetired and DuncanHill: I have been slow to understand your preference for a {{cite book}} in the reflist, with perhaps no bibliography. As was pointed out it, is indeed much easier to locate every reference to a book etc. by using <ref name="" /> in the reflist along with {{rp|36}}: 36 etc.: all refs are linked to the source on one line. My personal feeling (along with Izno, I think) is that it somehow slows down the whole flow of the article, as if the reference itself were more important than the information being imparted. It obviously depends on a number of factors: many social science journals use Chicago-style (Greenhill 1988, 38) harv error: multiple targets (2×): CITEREFGreenhill1988 (help) as a matter of course which also (imo) breaks up the flow. Mind you, many scientific articles are simply not concerned with forging a literary style, but merely presenting information in the plainest and simplest way possible. Again, it's a matter of preference. I am not being aggressive or dismissive, merely trying to understand why people prefer certain reffing styles. So, why might someone want to locate every cite of a particular source in the whole of an article? As a general reader, when coming across an article that interests me, I might well attempt to hunt down the source, but I wouldn't necessarily want to pinpoint every ref in a specific book, even as an editor or creator of an article. Thoughts? MinorProphet (talk) 15:31, 20 February 2021 (UTC)
Most of the sfn errors I come across (and I have fixed hundreds) are undefined works, or multiple undifferentiated works, or not matching the name in the sfn brackets with the names in the defined citation. Sometimes they are because of editors copying lumps of text from other articles and not bothering to check the refs afterwards, or editors "cleaning up" bibliographies and further reading sections without realising that some of the works are being called by refs. I've also seen them broken because someone has, cleverly and helpfully, defined refs in a template which happen to have the same name as other refs used on the page calling the template. As a reader I want to be able to check a citation quickly and easily, and for me short citations make that harder. They don't break up the flow of the text because if I just want to read the text I can choose not to look at the refs. When it comes to "literary style" in article space I'm with Beckett, who tried to write without it, and Q, who told us to kill our darlings. "Just the facts, ma'am", as Joe Friday didn't quite say. DuncanHill (talk) 15:58, 20 February 2021 (UTC)
I am caught by surprise by the volume of comments here - I will have to study them in more detail later as a bit rushed right now. An immediate answer to So, why might someone want to locate every cite of a particular source in the whole of an article? As a general reader,.... - in 2 parts. (1) General reader: when I first started using wikipedia - before doing any editing - if I wanted to find out more on a subject it seemed sensible to pick the major sources for an article and read them. Rightly or wrongly, I feel my preferred method of referencing allows you to quickly identify where the bulk of the article is sourced. (I note that "further reading" sections are an unreliable source of helpful reading material - if the books are not used as references in the article, why are they important?) (2) As an editor, more by accident than design, I seem to spend a lot of my rather limited editing time on articles that I feel are something of a disaster (in terms of content). The first job is to work out where the article is sourced. (That often answers the question: why is this so bad?) So often there is an impressive list of references, but when you get into it, you realise that most of the meat of an article is sourced to something one would struggle to define as an RS. The weirdest that I have found is a ref that appeared to be based on Fast Sailing Ships by David MacGregor - it's a pivotal work and cited in Clipper - where you would expect it to be used often. It then turned out to be an on-line link to just the index of this book, was used once, and supporting text that suggested the editor had never actually read the source. (Fixing this is a work in progress.) So - working out the overall reference base for an article seems key to sorting out any content problems.ThoughtIdRetired (talk) 17:20, 20 February 2021 (UTC)
Thanks all for your carefully and generously-expressed views. I feel that the original thread has shifted far enough to be restarted in the following sub-section. MinorProphet (talk) 21:40, 20 February 2021 (UTC)
It has been increasingly obvious to me for some time that there are a number of deep-seated problems with "WP" and its overall attitude to sourcing and referencing. The comments here exemplify what I have began to call WP's psychosis in regard to reffing. Maybe my impending essay could be titled "Why is this so bad?"™
1)@DuncanHill: I sympathise completely with your distress about failed {{sfn}} refs. I haven't experienced a tenth of your discomfort. I regularly click on 'Random article' to get a feel of what's around, but it's mostly Ukrainian 3rd-division footballers, Iranian villages pop. 270 in 2004, Japanese railway stations and species of minute sea snails in Patagonia with approx. 1 ref per article, or 2 if you're lucky. I get the feeling that you are a fan of {{sfn}}, Duncan, but the attempts of ill-informed editors have left you in a state of disrepair. Why do they get it so wrong? I refer to my previous post, which points towards a complete lack of comprehensive and authoritative help in this specific area. The choices of reffing presented to a noob editor are mind-bendingly multifarious. The possibilities presented in the various reffing "Help" pages are utterly contradictory and are displayed as a smorgasbord of equal-valued options, none of which can possibly be "better" than any other, since "consent" is apparently the prime factor.
Why is it so bad? Because although sfns and cite books are one of the best ways to approach referencing (imo), they are also the hardest possible way on WP, and quite frankly the "Help" is derisory. It's no accident that there are so many sfn fails: you might as well ask "Why do so many people die on Mt. Everest?" PS Why do you refer to William Faulkner, 'darling killer' as Q?
2)@ThoughtIdRetired: Thanks for your early reactions: I look forward to some further thoughts. I feel that you perhaps approach almost the same problem from almost 180°: the ability to pinpoint the significant sources (RS) of any article (if any) seems indeed crucial to working out whether it's even worth fixing (WP:Blow it up and start over), or could could just do with a little smartening up around the edges. If you would like some assistance with 'Clipper' I'd be happy to have a look at least. Slocum, Alain Gerbault, Dana, my seafaring literary heroes. I seem to remember having bought a copy of The Log of the Cutty Sark or something similar, a few years ago. I'll be getting the opportunity to dig out some boxes soon.
We all know when we come across a well-written, welcoming, intelligent and cogently-argued article, with enlightening, well-chosen sources and an over-arching sense of style and purpose, understandable by experts and newcomers alike: but these are, alas, most rare. Even the best-contrived, illuminating text in an encyclopedia article must fail unless it is backed up by solid references. Surely the quality of the refs underpin and define the entire edifice of our amazing enterprise: yet the reffing here is an utterly contradictory, gloomy and psychotically destructive enterprise as long as any editor can do what they damn well choose. There are far too many reffing options, and none can be be allowed a priority under the current system. The best reffing takes time and effort to learn, and the Help is pitiful: you can do this, or that, or the other, you can achieve it in 47 other ways which we will mix up and intersperse with other pointless and dead-end options, or you can devise your own insane system: but if you create an article with reffing system X, that's how it will be for the rest of time unless some sort of "consensus" is reached, which usually consists of two equally intellectually unattractive argumentative saddos engaging in an edit war before subjecting themselves to Arbitration presided over by sleeper fifth columnists and QAnon New Page Patrollers who by their billion edits have wormed their way into Adminship. Hmm, perhaps David Foster Wallace might have begun his career with a similar sentence. Sleep beckons. MinorProphet (talk) 21:40, 20 February 2021 (UTC)
Seem to be 3 headings for issues here. Getting citation information into an article accurately and efficiently. My preference for this is the {cite book} template. I have used {sfn} but find the technical requirements tedious and/or difficult to learn (i.e. I have missed the easy way to use it). The multiple problems with short references in Wikipedia suggest I am not alone in this. One problem with {cite book} is that you then have to use an extra template to put a page number in the first time. The second problem with {cite book} is that on the second use of a reference, you have to remember the name you gave the ref. Solutions of this include having a second copy of the whole article open and using "edit find" to get the right one. This is relevant if there is a prolific author on the article subject who has written several useful refs. References I use a lot are stored on my user page (so that gives consistency between articles on similar subjects.)
Then there are problems getting citation information out of an article. Short citations require some adept handling of the mouse to show the full reference details (you already have the page number at this point) - but it works. However, you cannot go to the reference section and make a quick assessment of the number of times a source is cited. (Unless I am missing something.) {cite book} has the problem of "ugly" inclusion of the page number in the text of the article. It is pretty pointless to show that then, because the reader does not know which ref it applies to without hovering the cursor over the reference's number. Playing (amateur) systems analyst for a moment, it is surely possible to improve the display of {cite book} references by only displaying the page number on user request (either by cursor hover, as part of seeing the whole reference, or an on/off preference switch for the page).
The last problem with references is source selection. This is the first and major fundamental role of an editor - decide which sources are important and authoritative, providing the basis for the article. (The second fundamental role is putting all that material into an intelligible and readable form.) To put this trivially, I am sure I could find a source that said the moon was made of cheese (in the children's book section), but not appropriate for Moon. More seriously, I don't know if all subject areas have this problem, but I have been editing in areas where there are a huge number of books that are totally inadequate. For Highland Clearances there are many books written for the "tourist bookshop" that simply pander to the misconceptions of the people who buy them. This is even addressed by academic historians (Tom Devine covers it quite extensively: The Scottish Clearances: A History of the Dispossessed, 1600–1900. London: Allen Lane. ISBN 978-0241304105 pgs 9-11). In maritime history, particularly on subjects like Clippers, there are many "coffee table" books (similar criticism to the Scottish tourist bookshop fare), but also authors who appear authoritative, but have big failings. Eric Kentley (16 years as a curator at the National Maritime Museum) wrote Cutty Sark, the Last of the Tea Clippers, published by Conway Maritime (a very good quality publisher for this subject) (2014). His lack of grasp of the subject is shown by some of the reviews of this book on Google (not the reviews that say "present for...." etc.) (Not sure I want to risk anything legal for Wikipedia by listing all the stuff Kentley got wrong.) Interestingly the worst of these were edited out in the second edition. (I note Cutty Sark does not use this work.) Basil Lubbock is a fantastically prolific writer - and often the only source on some sectors of the subject - but even his Wikipedia page warns that he get things muddled up a lot. In short, you need an understanding of the subject to choose the sources. Another danger is the book that is readily available online. If google books let you see all the text of a book that is not totally ancient, that surely tells you that there is no value left in that work for the publisher. The book of a similar age that you cannot see on google books, but would have to buy or find in the library, is almost certainly more authoritative. The partial views that you can get on google books means that an editor is seeing things out of context - and can misunderstand. Some authors are a real risk for this - Eric Richards (Scottish historian) often lays out the argument he wants to demolish in some detail, then says what he thinks. The snapshot views offered by google books would completely misrepresent his output. Then we have the source that is authoritative on some aspects, but not others. For instance [1], where a philosopher (appropriate in discussing the morality/legality of the bombing of Dresden) was cited for the death toll of the bombing of Hamburg. He simply had the wrong number - and any editor on top of the subject should have spotted this, or if they didn't know, checked it. My thinking here possibly goes further than WP:CONTEXTMATTERS warning about "Information provided in passing by an otherwise reliable source". Overall, Wikipedia editors need higher standards on source selection. A bit more time reading and a bit less time editing. I should add - I have, at times, made some real howlers on Wikipedia - but am fortunate that other editors have politely corrected them. ThoughtIdRetired (talk) 14:55, 21 February 2021 (UTC)
I've found a few howlers caused by poor source selection. A couple of favourites - a misattribution of a famous quote to the wrong speech, because one historian blindly copied another's error, another a totally spurious quotation ultimately sourced to a satirical work. Both of these would (or should) be obvious to anyone familiar with the subject. I'm sure I must have made a few too. Hopefully someone picks them up! DuncanHill (talk) 15:50, 21 February 2021 (UTC)
Perhaps we need a how-to-write-an-article guide that starts with references, as in "First, gather your sources and make sure they are acceptable, reliable and properly referenced." Then it becomes easier to build an article with a well-stocked reference section, on which every subsequently written section and paragraph will depend. I am only a newby in Wikipedia, but this is based on my experience writing other essays in an academic context, where the first thing I did with a new book was to write its reference down (in the locally relevant style) before starting to make notes from it. It saved me from thrashing around later trying to build a ref for a book that I had handed back long before. This might even <pious wish>discourage some of the more fantastical Draft pages from getting as far as being declined for lack of references</pious wish>. Writing in this way leads me a to use a hand-built reference list (all nicely in alphabetical order), with lots of sfn's picking out the relevant page numbers.--Verbarson (talk) 14:05, 22 February 2021 (UTC)
@ThoughtIdRetired:, thanks for your comprehensive comments. 1) Getting cite info into an article. I am a firm believer in {cite book} because it always displays the information in a completely consistent way. It's orientated towards scientific journals with volume, issue and page number as simple numbers, rather than the literary "Oxford" style with 'vol. III' and date at the end. This can be worked around with |others= and |series=. Although you are correct in saying that only <ref name="" /> shows up all the references to a single cite compared to {sfn}, it's not particularly difficult to difficult to scroll through the reflist and make a rough guess as to its relative importance. Do you or anyone really need an exact count?
You correctly identify the many problems of using <ref name="" /> with {cite book}, which is why I started using {sfn}s. I started an instructional article with examples, called The joy of sfn, but I haven't done much work on it recently. Maybe I should have another look. As always, it's very much a personal thing, and you are right that getting to grips with {sfn} takes quite a lot of work. I reckon that most of the effort is fact understanding {cite book}, and when you know it, {sfn} is a mere doddle, although DuncanHill knows how much it can go wrong. I've made a quick list of what can go wrong further down. In fact the whole mechanism of how any combination of [ref]...[/ref] and [sfn}s actually fit together is nowhere explained on an official Help page (most people probably wouldn't need it); although it the basics were kindly laid out for me here, especially the second answer 2.
As I see it, there are various ways of making a ref.
The very easiest for the editor: <ref>Greenhill, p. 36</ref>, and the reader has to go looking in the bibliography section for the relevant work, which can be formatted either manually or by {cite book}.
<ref name=>"Greenhill 1988"</ref> plus {rp|36} (I think your preferred method) which as you say involves "ugly" inclusions in the body of the text. Note that it only works if you include the entire cited source in the (first) instance of |ref name=, whether you use {cite book} or do it by hand.
The get-around is very time-consuming for the editor, which involves creating a separate ref for every page used: <ref name=>"Greenhill 1988 p36"</ref>. There is no intrusion, but the cite then needs to be in the Bibliography section, again either manual or with {cite book}. Note that in all these three cases {cite book} takes no dynamic part in the referencing system, it merely displays information, and the MediaWiki software (I think) just puts them together when it compiles the {reflist}.
You can make an inline {harv} eg {{harv|Greenhill|1988|p=36}} (Greenhill 1988, p. 36) harv error: multiple targets (2×): CITEREFGreenhill1988 (help) but I think this is now deprecated.
In my opinion, {sfn} completely overcomes all these difficulties. {cite book} automatically creates a {harv} which an {sfn} can link to via a CITEREF. {sfn} is actually just a {harv} wrapped in a <ref></ref>, When I create any {cite book} I put in a hidden {sfn} eg <!--{{sfn|Greenhill|1988|p=}}--> All I need to do is remember the author and date of the book (not hard), or it's in the bibliography section hidden in the cite. The entry in the reflist automatically links to the cite in the Bibliography, which appears if you hover over it, or jumps to it if you click on it. What could be easier? (cue distant hollow laughter...)
2) Getting cite info out of the article. I agree with all your points. As to the hovering the mouse, there may well be adjustments in Control Panel → Mouse which might improve things. I agree that the output of {cite book} could be tweaked quite easily, but I think only a very few people have the knowledge and ability to do it, and it would need a concerted effort to bring about any change. 3) Source selection. This is obviously a separate issue to the reffing systems discussed above. It would be somewhat unfeasible to compile a list of "approved" reliable sources, and since anyone can edit WP, you are going to get a wide range of sources chosen for various reasons which you discuss. "Anyone can edit" is both one of the great strengths and weaknesses of WP. Very often editors are not specialists or experts: many of us are merely informed readers with an interest in one or many subjects. But as long as coffee-table books (or worse) are used as sources, Wikipedia will remain a "coffee-table encyclopaedia". I agree with everything you say about Google books. Google search seems to deliberately omit archive.org from its results, and almost everything out of copyright is there in full with pdfs and plain OCR text. Even some contemporary books are there in full for short-term loan online like a proper library.
@DuncanHill: Just to recap on your reasons why {sfn}s get broken:
The {sfn} will appear to be good, but there's nothing it can link to because {cite book} is completely missing
Copying lumps of text from other articles and not copying the associated {cite book}, and not bothering to check the refs afterwards.
Removing works from Bibliographies without checking that there is a corresponding sfn in the text.
Duplicate CITEREF: (I think?) Editor has defined refs in a template (eg sfn, harv) which happen to have the same name as other refs used on the page calling the template.
Other common reasons:
Mismatch between params of {sfn} and {cite book} where both exist:
In {cite book}, any or all of |last1= (etc.), |date= or |year= is either missing or incorrectly defined. Even if the {sfn} is completely correct, the link will fail.
As I pointed out earlier, you must definitely use the correct date of publication using |date=, and |orig-date= and NOT eg |edition=1988, since CITEREF cannot extract information from |edition=.
In the {sfn}, any of the first two params may be incorrect, eg {sfn|Greenwood|1998|p=36} instead of {sfn|Greenhill|1988|p=36}
In the {sfn} if there is more than one authors (etc.) one or more may be missing or incorrectly spelled, even if the {cite book} is completely correct.
An extraneous pipe sign {sfn|Greenwood||1998|p=36} will result in an empty ampersand (Greenwood & 1998, p. 36) harv error: no target: CITEREFGreenwood1998 (help)
If you define two {sfn}s with different pagination schemes eg {sfn|Greenwood|1998|p=36} and {sfn|Greenwood|1998|pp=36}, you will get an error when previewing.
@Jonesey95: I installed Trappist the monk/HarvErrors.js and fixed Koornwinder polynomials as a test - seems to have worked a treat. Btw, there seemed to be no 'Install' button as suggested in the installation instructions. It also pointed out in James Douglas (English Army officer) the problem of using {{sfn|Greenhill|1988|p=36}} along with |edition=1988 in {cite book} as pointed out in the list above, giving Harv error: link from CITEREFKenyon1993 doesn't point to any citation. and Harv warning: There is no link pointing to this citation. The anchor is named CITEREFKenyon1986. (Haven't fixed it yet.) This is exactly what ThoughtIdRetired is using in his OP in the section above, and I'm afraid it is simply wrong and leads into error. It may be why you're not a fan of {sfn} since they would never work if you use {cite book} in this way.
Here's what you posted in your OP: <ref name="Greenhill 1988">{{cite book |last1=Greenhill |first1=Basil |title=The Merchant Schooners |date=1951 |edition=1988 |publisher=Conway Maritime Press |location=London |isbn=0 85177 475 X}}</ref>{{rp|99}}, which gives you this:[1]: 99 which happens to work, although cite book is wrongly formatted.
{sfn} can only ever work with this if you use {{sfn|Greenhill|1951|p=36}}, resulting in (Greenhill 1951, p. 36) harv error: multiple targets (4×): CITEREFGreenhill1951 (help), which is obviously not what you need.
Compare with {{sfn|Greenhill|1988|p=36}} along with this [2] (Greenhill 1988, p. 36) harv error: multiple targets (2×): CITEREFGreenhill1988 (help)
Thanks, User:MinorProphet. I may be working from an insufficient knowledge base, but here are some additional thoughts. (i) The refname I have given: "Greenhill 1988" is intended for the {citebook} template. I have found that it can be confused by some editors to be a Harvard reference, but I could have instead given the refname "strawberry jam" - but then I would have had to remember which one was which. (ii) In some cases it is absolutely crucial to identify which edition a reference is from. Over and above the risk that page numbers do not align, some editions have substantially different content - sometimes even contradicting what was said in an earlier version. I see that I am using year, orig-year and edition differently from you. It would help enormously if {citebook} had a direct link to an easy to read brief set of instructions on what should be in each field - hover over a field to see what information should go there? I have just revisited [[2]] and even knowing the usage preferred by MinorProphet, it is a little challenging to find that interpretation in the guidance as written. Perhaps I am getting a little impatient with instructions that require a bit of work, but can't some of the CE enthusiasts on Wikipedia make this less impenetrable? ThoughtIdRetired (talk) 20:46, 22 February 2021 (UTC)
...and WP:CITE seems to be a little light on "so how do I actually do that?" (though better than the help text given above). Writing instructions is a special skill - there must be someone in the Wikipedia community who can cover this fundamental part of editing. ThoughtIdRetired (talk) 21:21, 22 February 2021 (UTC)
@ThoughtIdRetired:Thanks for your reply. I also had a look at the Date section you referred to. Many things have changed or been updated since I last looked at it, and I would agree that what I said earlier isn't reflected in the Help page. In my opinion, the parameters for {cite book} should be filled out by copying the info on the physical title page and the next page (eg edition, isbn etc.) This means, I feel, that |date= should be the date of publication (whatever edition it is) and not the date of the original edition. You have also highlighted the grim state of actual practical help available in the documentation pages, including Template:Cite book. I absolutely agree that there should be one and only one style of reffing. WP:CITESTYLE says that "A number of citation styles exist" but doesn't tell you that none of these are available through {cite book}, a subset of CS1|2 which is a hodge-podge of Chicago and APA invented by WP:en coding editors and tweaked as they saw fit over the years. You are quite right about writing instructions, it's a particularly specific skill, but the documentation is usually written by the guys who write the templates, and unfortunately the two capabilities are almost never present in the one head. Just to check, |date= can take an {sfn} style disambig for two books from the same year, although the documentation at Date says that |year= fulfils this function. {{sfn|Smith|1988a|p=304}}[3]
...and on source selection - it's largely a matter of self-discipline by editors - "do I understand this enough to chip in?" That does not require you to be a total expert, and the co-operative nature of Wikipedia is the safety device for an editor getting it all wrong. But conversely, editors who go for high edit counts worry me - how on earth do they ever find time to research all those edits? More rules, I guess, would be a problem. Better compliance with the rules we have would help.ThoughtIdRetired (talk) 21:04, 22 February 2021 (UTC)
@Verbarson: Such things do exist, such as Your First Article and Help:Referencing for beginners, and Wikipedia:Reliable sources (RS), plus all the associated 'See also's at the end of the articles. On the other hand, because WP allows almost any referencing style, it is left up to the individual to make a choice with no advice given (or even allowed) as to what is "The Best Way"™. It really does take a fair amount of reading/researching experience to be able to recognise an RS, and as shown above there are very many ways in which any reffing method can go wrong. I've been on WP for around 12 years, but I continue to make basic slips such as missing off the / in a closing [ref]. I remember the confusion when it happened the first time, and how long it took to find the solution. Now I can recognize it instantly, but sometimes I can make multiple mistakes, and it takes five or more 'Previews' to get a single extended {ref} or {efn} right. Many intelligent people have never had the opportunity to write anything at university level (essentially, I feel, the general standard of writing, sources and reffing which WP is aiming for), and I also don't think that the sort of all-in-one guide you suggest exists. Maybe it would be far too long.
Reffing types - {ref}...{/ref} and the rest, discussed above, vs. templates ({sfn}, {harv} etc.
Planning - General layout - Bullet-point outline of sections
Content - style - WP:MOS which I call Ministry of Style, guidelines for GAs and FAs as ideals.
I agree that far too many people have no appreciation of these things, perhaps they just have to learn the hard way which takes time and effort, with no short cuts I'm afraid. MinorProphet (talk) 18:48, 22 February 2021 (UTC)
@DuncanHill: - I wonder if you could say why you don't like {sfn}s at all, even when properly used? I am genuinely interested, and would like to hear your thoughts. I realise that they are quite lot of extra trouble - much of the effort is having to get {cite book} right - and easy to get screw up as 26,000 articles show: instead a manual bibliography and a ref - /ref of some sort achieves much the same thing, if perhaps slightly less elegantly imo but with much less margin for error. MinorProphet (talk) 19:09, 22 February 2021 (UTC)
@MinorProphet: It's not just sfn, it's the whole Harvard style. It seems totally unsuited to Wikipedia. Look at David Lloyd George. Look at refs 1 & 3. Ref 1 I can point my mouse at the little 1 and see the book and author in preview. I click on the 1 and go straight to the details. Click "back" and I'm back where I was in the text. Ref 3 I point my mouse at and it just says "Harnden 2011". I click on the 3 and go to where it says "Harnden 2011", then I click on that and end up at the details. I then have to click "back" twice to return to my point in the text. Yes, I do know I can click on the preview, but that's far less intuitive than clicking on the number. As a reader who looks refs up it's a pain. As an editor who checks refs it's a pain. We should make it as easy as possible for people - readers and editors to look up refs. DuncanHill (talk) 20:08, 22 February 2021 (UTC)
Thanks User:DuncanHill - I am puzzled why a computer-based encyclopedia needs to use a referencing system developed for the paper-based world. One single referencing system, with all the problems sorted out, will surely achieve the highest level of reader-understanding and editor-usability. My vote would be {citebook} with better help text (hover over for help on what should go in each field - click for fullest detail?) and the page numbers of refs only showing on hovering over the ref number, to help the flow of readability of the article (with option to make all ref page numbers visible in the article?). Of course, there would be an absolutely mountainous barrier to consensus forming on such a proposal, but I don't think that should kill a reasonable aspiration.ThoughtIdRetired (talk) 20:55, 22 February 2021 (UTC)
@DuncanHill: Thanks very much for your views. The term "Harvard style" has been badly abused on WP, since it's generally mixed up with {sfn} and {harv} via CS1|2, and in fact Parenthetical referencing or Harvard-style inline Author-date referencing eg {Greenbrook 1988, p.46) which {harv} implements is now officially discouraged since October 2020. If I understand you correctly, are you disagreeing with the whole system of using {sfn} and {harv}, coupled to {cite book} in a separate bibliography? It seems to me that some editors/readers are quite happy with a References section populated with full cites, created by {cite book} using {ref name=""}, although {cite book} now creates its own CITEREF based on author and date anyway, ready to be linked to by [sfn}.
It's interesting that you picked David Lloyd George, which is an appalling mish-mash of reffing styles with many tens of editors in the last year. Having installed the Trappist the monk/HarvErrors.js as recommended by Jonesey95 as above, I can now see all the {cite book}s in the Bibliography with no ref pointing to them (ie roughly 50%).
Ref [1] uses the {ref name=""} style, but the whole {cite book} which appears in the Bibliography section is ignored and duplicated in the ref itself, ie briefly {ref name=""}(cite book |last= etc.}}{/ref}{rp|13} This, in my opinion, is completely wrong in this particular article, since the {cite book} entry in the reflist cancels out the one in the bibliography. You are right that hovering over the [1] shows the complete source, but that's because the entire {cite book} is included in the ref and every subsequent {ref name ="" /}. This is ThoughtIdRetired's favoured approach, I think, and it does indeed allow you to see every link to any source in the reflist. It is also true that you cannot link directly with this this reffing style to a {cite book} in the Biblio. I think may be possible in some other way using {ref}s, but I have never bothered to find out since {sfn} does it for you. As far as I know (and I may well be wrong) the only other widely-used way is a plain {ref name=""}Greenhill, p.37{/ref}, which means that editor and reader alike have to scroll down to the Biblio to find out. In other words, {cite book} should be either exclusively in the notes OR the bibliography, but not both.
Ref [3] uses {sfn} in the completely correct way. As you say, if you just click on the [3] that takes you to the reflist, and and clicking on that takes you to the Biblio, and you have to click back twice. But are you aware that when you point (hover) your mouse over the blue[3] so that it shows "Harnden 2011, p. 11", and then move the mouse slightly so it points to the blue link Harnden 2011, that it then shows the {cite book} entry in the Biblio? You can go there directly by clicking on it. The whole chain is displayed by two very slight movements of the mouse (admittedly two rather than one). If the cite book contains a url, you can even click on that and be taken there direct with an even smaller movement of the mouse.
Not every article needs a Bibliography section, especially short-ish ones, but a major article like this one most certainly does, and it has been seriously abused. In fact, one of the more concerned editors (see Talk:David Lloyd George#More problems with refs - Grigg 2002 vol3.) has been making just these changes, combining {ref name} with {cite book}, for this very reason that {sfn} "takes more clicks" which I feel is simply untrue. Has no-one ever tried hovering their mouse? You don't have to actually click on anything with {sfn}! Looking at the whole hodge-podge has made me feel slightly nauseous. I imagine that an article in this appalling state would take an entire week to fix, since 50% of the refs would have to be changed into either {ref}s or {sfn}s, including getting "consensus" (lol).
At some point this thread is going to get archived (five days of inactivity), so might I suggest if that happens we could carry on this most interesting and fruitful discussion we could do so on my talk page under User talk:MinorProphet#Further thoughts on referencing. Thanks anyway for all your most generous comments so far, it has been really stimulating. >MinorProphet (talk) 03:15, 23 February 2021 (UTC)
Quick reply. "But are you aware that when you point (hover) your mouse over the blue[3] so that it shows "Harnden 2011, p. 11", and then move the mouse slightly so it points to the blue link Harnden 2011, that it then shows the {cite book} entry in the Biblio? " - no, it doesn't - do you have some kind of script which makes it do that? Even if it did, it's still more moves than necessary. The bibliography on Lloyd George is a mess I agree, it has been used to list just about every book or article that a particular editor has found. I've tried trimming it occasionally, and it needs re-ordering. I am the "more concerned" editor in the talk page thread you linked to. I found a mass of duff refs (wrong book) and turned them into correct ones. It is infinitely more important to have accurate refs than ones which please your aesthetic sensibilities. I do not see any reason not to list all books cited after their appearance in the refs section. A "works cited" followed by a "further reading" section is what DLlG needs. The errors which matter are the no pages, no works, wrong authors, wrong dates, which are, in my experience, far more common using sfn or harv. Having a citebook not called by a reference isn't, to my mind, an error at all, just an artefact of the gadget. DuncanHill (talk) 03:47, 23 February 2021 (UTC)
@DuncanHill: Hi, thanks for your swift reply. I'm really sorry not to have recognised your name. I did look through the article history to check if anyone here was involved, but I haven't had had much sleep recently (although not under the influence) and I simply didn't make the connection. I hope you aren't too offended. As far as I know I have no script installed which allows this hovering effect. I'm using an ancient version of Firefox (v. 47) on XP SP3, which can sometimes give strange effects. As far as I remember it just happened one day and it has continued consistently for at least five years. By your description of the {sfn}, you have no other choice except to click directly on the [3] or the Harnden 2011 when you point at /hover over it: either of which takes you to the reflist, from where another click takes you to the {cite book} in the Bibliography section. Can you not point to/hover over the Harnden 2011 link even in the reflist to see the complete cite? What browser/OS are you using? I can certainly see that it would make using {sfn}s a completely and frustratingly tedious bore, which is the exact opposite of my experience and why I'm a fan. I wonder whether @ThoughtIdRetired: has the same experience as you. I think will start a new thread to enquire.
On the subject of DLlG, skipping through the talk page I became aware of the mass of screwed-up refs of what ever type. You are quite right that "Having a citebook not called by a reference" isn't a real error, I was only made aware of it having installed the HarvError script I mentioned. I think this has only come about relatively recently, since changes to CS1|2 meant that |ref=harv was no longer needed in {{cite book}} to allow {sfn} to link to it (a major source of sfn fails), but was able to generate its own CITEREF automatically. Thus the 26,000 articles that use {cite book}, {citation} etc. with no associated ref. In the old days plenty of articles used {cite book} without |ref=harv, simply as a consistent way of formatting the cite, but it has turned into something else, perhaps an error-generating beast whose 'errors' are not necessarily errors at all. I might have a look at some of my own articles...
As I have said, and your experience shows, {sfn} etc. is essentially more complex than almost any {ref}...{/ref} or {ref name=""}, coupled or not with {cite book}, perhaps even by an order of magnitude: but I still maintain that the unless the params of {cite book} are used consistently and 'correctly' (whatever that means), there will still be room for errors even if {sfn} is not being used. My personal approach is: since {cite book} generates its own CITEREF, almost exactly the same as an intelligently-chosen {ref name=""} why not use it? Again it's personal.
It seems there are two approaches to the appearance of the reflist. One is to have everything in it in long format, regardless. The other is for it to contain only short footnotes, generated either by a very simple {ref}Greenbridge, p. 204{/ref} [or (Greenbridge 1998, p. 204) if needed for disambig], or {sfn}s. This keeps the reflist very compact, with only {cite web}s making a difference. Plain {ref}s require everyone to scroll down to the Bibliography, {sfn}s with their hoverability (for me at least) don't. Again, I find {rp|204} intrusive as it breaks up the flow of the text, although this is obviously not a difficulty for everyone. Sorry, I've been rather long-winded. I appreciate your taking the time to reply. Cheers, >MinorProphet (talk) 05:17, 23 February 2021 (UTC)
Well, well, well (three holes in the ground). Check out #Displayed output of sfn below. Izno kindly pinpointed the source of the behaviour. It's to do with Preferences → Gadgets → Browsing and a combination of Navigation popups and Reference Tooltips. For me, when Nav. popups is disabled and Ref. Tooltips is checked, all I have been describing bursts into life. MinorProphet (talk) 08:41, 23 February 2021 (UTC)
If you want to run it yourself, just use the Fork button to create your own copy. And then you can run whenever you want. -- WOSlinker (talk) 15:08, 22 February 2021 (UTC)
Best way to intra wiki link multi language edit-a-thon?
I hosted in the past, an edit-a-thon at Wikipedia:WikiProject Organized Labour/Online edit-a-thon Tech February 2021, however there are no corresponding Labor WikiProjects in German or Italian Wikipedia, so I hosted the pages in my personal username space, e.g. see de:Benutzer:Shushugah/Tech-Arbeiter*innen_Edit-a-thon, and manually inter wiki linked them, since they're ineligible for a WikiData item due to being in username space. Another option would be to...boldly create a WikiProject in German wikipedia, but Italian and other language wikipedias don't have such option.
I regularly link across Wikipedia languages and Wikipedia projects, but requiring new users to prefix every article with the Language AND Project is a high technical burden. Having a Meta place holder to encourage new Wiki platforms does make sense though. Shushugah (talk) 15:45, 23 February 2021 (UTC)
Image resolution on mobile
What is the highest resolution image that will display on mobile? I have noticed that if I go to Commons and click on a high resolution version in some cases it will not load. Is this the operating system or Mediawiki or something else limiting download size? Cheers, · · · Peter Southwood(talk): 18:38, 21 February 2021 (UTC)
RoySmith, My experience has only been with .png files, Android OS, Samsung tablet and a few phones. I haven't done much experimentation because why bother if it is a known limitation. · · · Peter Southwood(talk): 04:33, 22 February 2021 (UTC)
I have had a few friends test. iPhone 11 and 12 seem to manage, but Android phones tested so far fail on the original of File:Dive_sites_of_the_Whittle_Rock_Reef_high_resolution.png (14,040 × 9,930 pixels, file size: 8.71 MB, MIME type: image/png) The next highest resolution available by default is not legible for the small print. Is it likely to be OS, browser, or timeout for some reason? · · · Peter Southwood(talk): 08:50, 22 February 2021 (UTC)
Pbsouthwood, definitely memory. That’s a gigantic image. It either runs out of memory, or isn’t even able to address images of such a size in some level of graphics libraries it uses. —TheDJ (talk • contribs) 08:32, 23 February 2021 (UTC)
Fair enough TheDJ, the problem is then that the highest resolution available from the default set of pngs is too low to read the text. Is there a way to work around this? · · · Peter Southwood(talk): 12:19, 23 February 2021 (UTC)
Pbsouthwood, with default pngs, do you mean the links in the file page with preset sizes ? In that case, open any of them and modify the url to retrieve a width that is somewhere in between (hmm, just tried that, seems that the thumbnail server is having trouble even scaling something to those in between sizes...It just throws errors) In general I would say.. this is not a suitable file format to store that amount of information... SVG or something might be better. —TheDJ (talk • contribs) 13:35, 23 February 2021 (UTC)
Correction: The 6,771 × 4,789 pixels version crashes in zoom, but the original svg seems to zoom in just fine. However tapping the image from the article does not load the svg, it loads a bitmap rendering, presumably a png. · · · Peter Southwood(talk): 17:07, 23 February 2021 (UTC)
Pbsouthwood, That's a VINO (Vector In Name Only). Sure, it's packaged in an SVG format, but it's mostly just a bunch of encoded PNG bitmaps. That doesn't solve the problem that you're dragging a huge amount of data around.
I know what I'm suggesting is a lot of work, but I think the only way to make this work is to actually trace the contour intervals so you can generate a real vector version of this. -- RoySmith(talk)17:48, 23 February 2021 (UTC)
RoySmith, That is the original Inkscape svg from which the png was rendered, so the work has already been done. It is probably possible to trim it down a bit, but not sure how much. · · · Peter Southwood(talk): 18:02, 23 February 2021 (UTC)
Pbsouthwood, I'm not an expert either Inkscape or SVG, but when I look at the data in the file, I see 70 lines that start with
xlink:href="data:image/png;base64
I assume each of those is a bitmap image that you imported into Inkscape. Pulling those lines out, they add up to 78% of the bulk of the original file:
$ du -h x Map_of_Whittle_Rock_dive_sites_2021.svg
7.8M x
10M Map_of_Whittle_Rock_dive_sites_2021.svg
so that's clearly the place to start. Maybe you could down-convert the individual images to something much lower resolution and re-import? PNG uses lossless compression. My hunch is that JPG's algorithms would do a much better job of compression than PNG's on these kinds of images. I would start by converting the original png files to jpg, and experiment with how low you can crank jpg's quality parameter and still get a reasonable rendition.
That sounds like a bot-like task. Could a bot be written to remove the deleted categories from the articles after the categories are deleted? That isn't easier for the bot coder, but it sounds as though it would make it easier for the closer of the CFD. Robert McClenon (talk) 23:54, 22 February 2021 (UTC)
I know little about the details of categories. If my comment above is completely off the mark, then I know even little less than I thought I do. Robert McClenon (talk) 23:54, 22 February 2021 (UTC)
@Fences and windows: Quite so; if a category is deleted legitimately, i.e. via a WP:CFD, the normal procedures of that forum will pass the actual empty-and-delete task to an adminbot. On occasion, a category is populated by a template - for these, an amendment to the template code should be sufficient. There is normally no need to empty a category manually. --Redrose64 🌹 (talk) 10:43, 23 February 2021 (UTC)
Hi all, I've just learned in another lengthy discussion above, (latest posts in #Further thoughts on refs), that my experience of {{sfn}} is somewhat different to at least one other editor. When I hover over a ref generated by an {{sfn}}, it shows me the blue-linked short footnote found in the {{reflist}}, eg Harnden 2011, p. 206. Remaining in the main body of text, If I move my mouse a fraction and hover over Harnden 2011, it displays a second box (not sure of the exact name) which contains the usual output of {{cite book}} which generated the CITEREF in the first place. Clicking on that takes me straight to the Bibliography (or wherever the {{cite book}} is located}. Is this the default behaviour, or is it just me? I'm using ancient FF v47 and XP SP3, which shows certain quirks (eg recently all lists have one column only, whatever the width eg {{reflist|20em}} NB That's not the problem here.) The other editor says he has to click on the ref to get to the reflist, and then again to get to the Biblio, and has to click back twice to get back to where he started. So, if my experience is the proper sequence, why might other users not be able to see the second box through a hover? As far as I know I have no script installed which might do this, but I am very un-technical in this respect. Cheers, >MinorProphet (talk) 05:52, 23 February 2021 (UTC)
When I hover over a shortref, it shows me no additional pop-up/second box. I have to click on that reference to see the whole citation. The whole citation then of course doesn't have a link back to either its use in the text or its shortref. The lack of immediate information in the pop-up, and the lack of the ability to right click a source link from that pop-up without darting around other areas of the page, makes me prefer longform citations. CMD (talk) 06:57, 23 February 2021 (UTC)
@Chipmunkdavis: Hi, I'm not surprised you prefer longform, I'm sorry that your experience seems so poor compared to mine. What browser and OS are you using, to help narrow down where exactly this behaviour occurs? (Could be just me, only 1% of users worldwide use XP, max. possible version of FF is 52.9, and is now at v.85...) MinorProphet (talk) 07:28, 23 February 2021 (UTC)
(I'm to blame for your reduced viewing experience of columns, but you're now one of the <2% using a browser that can't handle them with standard CSS, and most who can't are in fact stuck on Chrome mobile from the same time period... The use case for which will almost always be a single column. I judged that it would be reasonable to stop sending 3x the CSS just for that 2%.)
Out of curiosity, what color is the second box? Secondly, when you go to Gadgets in your preferences, do you have Popups enabled? Third, what options are checked on the Beta page in your preferences? --Izno (talk) 07:35, 23 February 2021 (UTC)
Aha. I didn't even know a second pop-up was possible before this post. Following this tip, I disabled pop-ups in the Wikipedia gadgets, and now I get a second pop-up. How counterintuitive. CMD (talk) 07:44, 23 February 2021 (UTC)
The two kinds of pop ups not playong each other nicely is a known issue, largely caused by the fact that the nav popups gadget was first written some 15 years ago and received little in the way of updates. The other is more modern... --Izno (talk) 09:06, 23 February 2021 (UTC)
All the popups are b/g plain white, f/g black text, links in blue.
If you mean Navigation popups: article previews and editing functions pop up when hovering over links, then no, it's not checked. On the other hand, I do get article previews.
No Beta features enabled at all.
Aha! How about Browsing → Reference Tooltips, which is checked. Just disabled it, and all the incredibly useful info disappears and a vast sadness rolls over me. Restored Tooltips, and all is light and joy. Why on earth is this not default behaviour? All the stunning usefulness of sfn and cite book is otherwise completely negated. >MinorProphet (talk) 08:06, 23 February 2021 (UTC)
My memory says it is the default behavior for accounts registered after a certain date, and I think there was some discussion about making it the default viewing experience also for unregistered editors. But I am unsure. That all said, I've been watching the discussion above and would like to point out the books referencing project, which, whenever WMDE gets off its butt to finish it, will allow one to get both the lesser and fuller citation in one popup (or at least that is in the current backlog), at least for articles which convert to that style. --Izno (talk) 09:06, 23 February 2021 (UTC)
Reference tooltips works for me when logged out and on other browsers, so it might be the default in newer browsers. CMD (talk) 09:03, 23 February 2021 (UTC)
The browser should have nothing to do with it. Whether you get ReferenceTooltips or one of the others is down to your Wikipedia user settings. ReferenceTooltips first became available as a gadget on 4 April 2012, and was initially opt-in for all logged-in users. At 00:26, 17 July 2012 (UTC) it became a default gadget, which means that it was enabled for logged-out users, and was also enabled (on an opt-out basis) for logged-in users registered after that time. Logged in users who registered before 17 July 2012 did not have their setting altered, so unless they had already enabled the gadget, it remained switched off for those users. --Redrose64 🌹 (talk) 23:25, 23 February 2021 (UTC)
If I turn off popups then the ref tooltips thing seems to work, but I don't get page previews. Even if I did I understand that page previews lack the edit/history/talk etc links that popups have and which make them so invaluable. DuncanHill (talk) 10:11, 23 February 2021 (UTC)
There are at least three different kinds of popups, which are not compatible - if you turn one of them on, you also need to turn off all of the others. At Preferences → Gadgets we find:
Navigation popups: article previews and editing functions pop up when hovering over links
(D)Reference Tooltips: hover over inline citations to see reference information without moving away from the article text (does not work if "Navigation popups" is enabled above)
and at Preferences → Beta features there is ⧼popups-refpreview-beta-feature-message⧽ which shows
Show a preview of a reference by hovering over its footnote marker. Please note: If you’re using the Navigation popups gadget or the Reference Tooltips gadget, you won’t see Reference Previews.
The default is that Reference Tooltips is enabled, the other two disabled. Perhaps the compatibility warnings can be strengthened. --Redrose64 🌹 (talk) 10:56, 23 February 2021 (UTC)
Would it be technically possible to unbundle the reference tooltips from the rest of Navigation_popups, if that is the only redundant aspect to ReferenceTooltips? Seems odd there are two gadgets doing the same job with one being worse at it. CMD (talk) 11:07, 23 February 2021 (UTC)
Ber-limey, what an utter and seemingly irreconcilable mess! Welcome to It_Wasn't_Me_Miss.com, aka the daily refuge of maintenance coders the world over. Although I laud everyone's technical and/or coding ability; also people's apparent understanding of the technical depth of what is going on; and our exceptionally refined ability to discuss both technical and emotional issues in the same breath, it appears we may have arrived at a pretty pass. There seem to be a number of previous issues which have caught up with us, and a number of current issues which seem unresolvable unless certain decisions are taken. And I suggest that the responsibility rests with us, here, now. "What, then (sang Plato's ghost), what then?"[3] >MinorProphet (talk) 12:37, 23 February 2021 (UTC)
Hi all, does anyone know why visual editor seems to do this from time to time? I didn't click anywhere near the top of the article and just was expanding the ref. I don't have any other examples off hand, but I have noticed this happening every now and then. Rather odd. --TheSandDoctorTalk16:36, 17 February 2021 (UTC)
@RoySmith: Definitely a weird one. Unfortunately not off hand or that I could find in around a 30 minute search. I know that I have seen it before (typically IPs notice and correct lol), but it seems to happen every few hundred to few thousand edits. Special:Diff/1002506672 is the closest I could find, where it inputted random text at the start of the article. --TheSandDoctorTalk18:24, 17 February 2021 (UTC)
TheSandDoctor, Thanks. Were these both instances of you having copied the completed cite template from somewhere and pasted it into the article? The next time this happens, could you please set your time machine to t minus 5 minutes and video record your earlier editing session? -- RoySmith(talk)19:02, 17 February 2021 (UTC)
@TheSandDoctor: What's your web browser/operating system, and do you find that you "lose focus" after some things? (Like you know where the cursor is supposed to be, and then you click on something, and the cursor seems to be nowhere)? Whatamidoing (WMF) (talk) 00:32, 23 February 2021 (UTC)
@Whatamidoing (WMF): I was using Google Chrome on Windows 10. In the Twitter article case, all I did was update the ref and then click apply & then save the edit. As for the Swift article, it was too long ago for me to remember. --TheSandDoctorTalk03:42, 23 February 2021 (UTC)
@Whatamidoing (WMF): Could you please clarify the question? Do you mean whether it took two clicks to have the "edit" button show up? If so, that sometimes happens but do not know if it happened in this case, unfortunately. If not, please elaborate and I will answer as best I can. --TheSandDoctorTalk18:14, 23 February 2021 (UTC)
It happened to me last night in Firefox, and an editor fortunately undid the whole thing instead of correcting it, which meant that I saw it. (It's not a page that I routinely check the history for). The key detail in both my and @Chipmunkdavis's diff is that it's replacing the first template at the top of the wikitext with the citation template.
As I understand it, it is possible to configure one's preferences at meta to create a global user page that will be transcluded from meta to all wikis, MediaWiki:Help:Extension:GlobalUserPage. When a user does that, a user page is no longer editable on en-wiki and its history log is not viewable here either. Most of the time this is fine, but I can see some potential problems, e.g. if somebody starts putting some inapprorpriate material (spam, using the userpage as a webhost, personal attacks etc) on their global user page, and we are not able to address the issue here. Another situation concerns users who are banned or blocked (e.g. for sockpuppetry) on en-wiki. Often in such cases we tag their userpages accordingly but it would seem that for a user with a globally transcluded user page we don't have this capability. Or do we? I saw a user who got indef blocked at ANI yesterday and they seemed to have implemented a global user page option via meta today. Is there anything that can be done in such situations? Nsk92 (talk) 13:36, 21 February 2021 (UTC)
The easiest way is to create a user page locally. All local user pages override global ones. Note that meta has also policies against inappropriate material like spam, webhosting, attacking others, so for some pages requesting deletion there is an option too. Majavah (talk!) 13:43, 21 February 2021 (UTC)
Nsk92: the local user page can still be created if there is a transcluded meta page. Try it on your example. If someone is using their meta global page for abuse, a global lock is probably in order also. –xenotalk13:46, 21 February 2021 (UTC)
The blocked user in question, User:Tisquesusa, did have a local user page before. Then it suddenly disappeared and got replaced by a global user page transcluded from meta. The globally transcluded page does not contain any abusive or improper material at the moment (there was a G11 user subpage that I CSD tagged today and it got deleted). But the situation still somewhat concerns me. In fact I don't understand how I can try to create/edit a local user page for this user now (assuming I wanted to do that). I can't create a red link. I can't access the history log for the user page. The global user page just sits there. Nsk92 (talk) 13:59, 21 February 2021 (UTC)
The page that was deleted (as both G11 and U5) was a subpage User:Tisquesusa/Más Muisca, not the parent user page. I don't remember what the user page itself contained, but the subpage was an advertisement page for a guided tour/adventure operation run by the user. That's U5 in my book. Nsk92 (talk) 14:30, 21 February 2021 (UTC)
Nsk92: what do you mean "I can’t create a red link"? Unless there was a change I didn’t hear a about, any user can create another user’s user page. –xenotalk14:34, 21 February 2021 (UTC)
Only if it doesn't exist already. When I type 'User:Tisquesusa' in the search window and press 'Go', I am taken to the global user page, User:Tisquesusa. If I press 'Search' instead, I get a line 'There is a page named "User:Tisquesusa" on Wikipedia.' Nsk92 (talk) 14:43, 21 February 2021 (UTC)
When visited locally, it can be edited locally, despite the global presence. Generally if there is abusive content it would simply be turned into a redirect to the user talk page. –xenotalk14:46, 21 February 2021 (UTC)
But how, exactly? How can it be turned into a redirect or edited locally, if necessary? I can't access the local history log for this user page now. And the edit button for it is not available either. Nsk92 (talk) 14:49, 21 February 2021 (UTC)
Your link works, and it allows me to create a local user page (it does look like it has been deleted or displaced somehow, presumably by the global user page being activated). But I have no idea how you created this link. Did you just manually type an entire http address? Interesting and strange ... Nsk92 (talk) 15:08, 21 February 2021 (UTC)
I’m using the desktop monobook responsive view; I have some custom scripts but I don’t think they are what’s adding the edit button for me. –xenotalk15:09, 21 February 2021 (UTC)
Nsk92, the link you are probably looking for is "Add local description" which takes you to the edit page. It is confusingly named, but I think it is because it uses the same system as images on commons use (i.e. random file File:Rueda de prensa sobre la sentencia del tribunal europeo acerca de los desahucios en España (8558751810).jpg when viewed on enwiki has the "Add local description" link instead of "Edit"). The "Add local description" link makes more sense for images, as you are adding a local description of the image. However, it doesn't make as much sense for global userpages. Perhaps this should be renamed for global userpages to something like "Create local userpage"? Dreamy Jazztalk to me | my contributions15:29, 21 February 2021 (UTC)
(edit conflict) Depending on preferences there may be one or more of the tabs "Add local description", "Create", "Create source", "Edit", "Edit source", or MonoBook variants. Don't you have any of them? PrimeHunter (talk) 15:31, 21 February 2021 (UTC)
You are right, there is an "Add local description" button. I didn't realize that it acts as a local edit button (which would presumably override a global user page?) Nsk92 (talk) 15:55, 21 February 2021 (UTC)
← Yes it would Nsk92. Dreamy Jazz, good suggestion for the interface change. The responsive view was giving me an intuitive pencil but in landscape, it is indeed Add local description. –xenotalk16:07, 21 February 2021 (UTC)
Xeno, if I'm not wrong, would that be achieved by placing the text needed in MediaWiki:Create-local? I also presume that mediawiki pages can contain parser functions and magic words which will work per page (so that the page knows where it is being used and then can modify the wording based on the namespace). It could, if I assume correctly for those both, use a parser function to check if the namespace number is 2 (i.e. userspace) and then output "Create local userpage" instead. Dreamy Jazztalk to me | my contributions18:38, 21 February 2021 (UTC)
Edit filter to prevent http: //%5B(and cousins) making good links bad
So while playing on toolforge and running a search on %.%[4](which terrified me that I would blow up a server), I stumbled onto the fact that potentially hundreds of links/refs are broken because http://%5b (as well as %5b%20 , %5E and a few others) are appended to the beginning of the url.
Working my way through cleaning up what I can but wondering if an edit filter would be viable to that warns an editor adding a double http:// in an edit summary (I say warn because possible their are valid reasons - Interner archive comes to mind)
Slywriter (talk) 02:31, 25 February 2021 (UTC)
Here is an example one added. No apparent cause. 10 years ago. Unclear why it's happening. %5B = open square-bracket. Old MediaWiki bug? -- GreenC02:53, 25 February 2021 (UTC)
Wow, sorry, totally forgot to add examples. I've already cleaned up a bunch but let me go back through them and see if there is any similarity in when the edits were added. Slywriter (talk) 03:05, 25 February 2021 (UTC)
Yeah hard to say. BTW notice some of the URLs have trailing %5D that also should be removed. It may be pretty difficult to automate a fix given how many forms it takes. -- GreenC04:18, 25 February 2021 (UTC)
Like this URL http://%5Bhttp://www.drukgyelhss.edu.bt/%20Drukgyel%20school%20%5D%20 in Drugyel higher secondary school should be [http://www.drukgyelhss.edu.bt/ Drukgyel school]
5B & 5D there. Cut and Paste of a wiki link gone wrong?
Tested what happens when you manually type/cut and paste [http://example.com] into the edit screen(&visual editor manual cite wizard) as [http://[example.com]] but that didn't trigger a change to unicode, the brackets remained.
Doubt I'll have much luck figuring out the how they do it but the brackets showing up give me the idea to search against other special wiki characters (<>{}~) unicode equivalents. Might be some other unicode characters that will cause obvious problems at beginning or end of a url.
If the double html remains constant, may be able to at least automate the generation of a list that can be manually checked. Slywriter (talk) 05:31, 25 February 2021 (UTC)
With some sleep and a new look at the data, assuming I am correct in my assumption that my search string is returning an alphanumeric sorted list of ALL websites links found in article space, then the problem is less than a 100 links. Only reason I question my assumption is the surprisingly low number of IP only web addresses that are used for links. Slywriter (talk) 14:49, 25 February 2021 (UTC)
It can happen if you try to use the toolbar link feature on code which already has external link syntax. An article may say [http://www.drukgyelhss.edu.bt/ Drukgyel school]. Click the chain icon in the toolbar and paste the code in the top field. You get a warning and cannot save. If you select "To an external web page" then you can save but it becomes [http://%5Bhttp://www.drukgyelhss.edu.bt/%20Drukgyel%20school%5D [http://www.drukgyelhss.edu.bt/ Drukgyel school<nowiki>]</nowiki>]. I don't know whether nowiki was always added. PrimeHunter (talk) 21:45, 25 February 2021 (UTC)
Thanks for that explanation. The sequence also explains why it's mostly, though not exclusively, low quality pages that its occurring. Situations where it seems the editor was determined to get the link in.
I've found a few other bad link styles like web addresses starting with a period http://.example.com . ::For now, I'll work on cleaning up through toolforge but once I have a good list of the errors is here the best place to see if someone with better backend knowledge can setup a query that I can run on my own or updates a page to show likely bad URLs? Would be better for long term maintenance and I suspect lower overhead than me running constant toolforge searches. Slywriter (talk) 22:08, 25 February 2021 (UTC)
Latest version of Firefox appears to break Wikimedia single sign-on
This clearly either needs WMF liason with Mozilla to whitelist the single sign on mechansim, or to use the Storage Access API to request it be permitted on a site-by-site basis. Where is the best place to report this? -- The Anome (talk) 09:18, 25 February 2021 (UTC)
But does it need to be fixed if it isn't broken? Using firefox as always, version 86, has continued to give great service, and hasn't broken yet. just an observation. -Roxy the grumpy dog.wooF11:03, 25 February 2021 (UTC)
@The Anome: to be clear, simply having FF v86 doesn't appear to be breaking logon, does it? Is this an opt-in feature that users have to select? (I think the option is labeled All third party cookies (may cause websites to break)). — xaosfluxTalk12:06, 25 February 2021 (UTC)
Logon is fine, cross-site logon between sites ending in "wikipedia.org" is also fine. What doesn't work is SSO between top-level domains, such as Commons or Wikidata, if you log in from a wikipedia.org site. -- The Anome (talk) 12:13, 25 February 2021 (UTC)
I believe so. But the system sould support strict mode, as this is what any privacy-conscious user should be using already. -- The Anome (talk) 13:46, 25 February 2021 (UTC)
The mozilla blog says Total Cookie Protection makes a limited exception for cross-site cookies when they are needed for non-tracking purposes, such as those used by popular third-party login providers. – SD0001 (talk) 07:47, 26 February 2021 (UTC)
Ugly workaround
I threw together a hacky little user script at meta:User:Suffusion of Yellow/central.js. Works for me with FF86/Linux with all third party cookies blocked. Install in your global.js, then click "Central login" under "tools" on any site you are currently logged in to, and follow the directions. You will be logged in to all the other wikis, too. Suffusion of Yellow (talk) 21:26, 25 February 2021 (UTC)
This 2017 en.WP RFC determined that once bots and scripts have replaced magic links (ISBN et al.), magic links should be disabled. As far as I know, bots and scripts have done this work (except for new untemplated and nowiki-wrapped ISBNs that are added continually by manual editing and buggy VE copy-paste editing), and the ability to disable magic linking locally has been provided.
This MW page appears to explain how to disable magic linking on a per-wiki basis. It is past time to do it here, if we have this local control and if there are no show-stopping feature requests that depend on magic links. There are a bunch of magic-link-related bugs and feature requests linked at T145589, but I am unable to determine whether any of them will be affected if we disable magic linking locally.
Is there anyone here with the sysadmin-level knowledge to figure out whether we can finally disable magic linking here at en.WP? – Jonesey95 (talk) 03:18, 25 February 2021 (UTC)
I don't disagree that these should be disabled at some point, but there are more than 11,000 ISBN magic links in use and a few thousand of the rest combined. If the magic links are disabled how will they be replaced? Perhaps we could start with an edit filter that does not allow any more magic links to be added, and then hopefully the fixer bot can get rid of those remaining? RudolfRed (talk) 00:57, 26 February 2021 (UTC)
I'm pretty sure the bots don't mess with User-space or Talk-space pages, for good reason, and there is no need to "fix" old AFD or other discussion pages. There should be no problem disabling magic links in those spaces.
That leaves only about 1,000 article-space pages with ISBN magic links, down from an initial population of about 500,000 in 2017, and with many tens or hundreds of thousands of pages fixed up by bots over the years. That's 99.8+% done, which is good enough. Those stragglers presumably exist only because the bots are taking care not to mess with edge cases like this one or this one. I'm fine with keeping the maintenance categories or WPCleaner reports around if we think they are useful, but the RFC said we should disable these magic links. It's time. – Jonesey95 (talk) 04:57, 26 February 2021 (UTC)
I say go for it. The effects of T162291 are the same (I assume they will stop doing that once the change goes through), T179769 is stalled because it has not been requested to disable magic links yet and T145589 is just about autolinking new magic links in a new way. It is possible to make an specific tool to make VE link those links the new way, once T179769 is fixed, see mw:VisualEditor/Gadgets. At the most, just warn VE users that adding magic links the new way is not going to work until T179769 gets fixed. Tidy was repaced with linter errors outside of mainspace still remaining, so that should be fine (seemingly not on enwiki, but elsewhere, see phab:T192821).--Snaevar (talk) 14:29, 26 February 2021 (UTC)
I am using the Wikipedia app to browse through Wikipedia. It really annoys me that it saves your search history. I know how to clear my search history in the app, but I would like to how to prevent the app from storing my search history so I don’t have to keep clearing it. Same goes with the article viewing history. Thanks, Interstellarity (talk) 00:06, 28 February 2021 (UTC)
my question is when I enable the HIDE->bots parameter - in order to reduce the size of my watchlist - the problem is the hidden articles may contain as well non-bot new edits before the latest/recent bot edit, that have been made earlier (but was not present in my earlier listing, since those edits did not have been made then, but between my two listings). With other words, in case the last=(recent) edit has been made by a bot, then that article will not displayed on the actual listing of the watchlist, and it will remain until a non-bot edit will be performed, however, if again a bot edit would follow (and meanhile I would not make a fresh listing by any purpose), again it would be/remain hidden, and I would not know anything about any intermediary non-bot edits, like that. This is the problem. I would like just to hide those articles in my actual listing, which had just and only both edits from a given time, but not the non-bot edits have been done after the given time, but before the recent/latest bot edit. I hope I formulated my problem in a way to be correctly understood. Thank You.(KIENGIR (talk) 22:34, 24 February 2021 (UTC))
Thank, I looked through the thread, but as I see the outcome tended the opposite direction I wish to achieve...or I would be wrong?(KIENGIR (talk) 00:55, 25 February 2021 (UTC))
thank you, I am very afraid to make tryout changes, since in the past a village pump discussion was needed to fix another issue with the watchlist, and I had luck with the tweak that was proposed, however, the whole system did not work as should be, so my current result is a test&tweak alltogether, far from perfect (not listing back to the set days, in a way number of diffs limited, etc.). Explanding watchlist to show all changes - even with after hiding bot edits - is out of qustion, since I wish to reduce, it would dramatically enlarge it (I assumed the directive you put as a sequential step and not separate). The group changes I don't understand what would exactly means, can someone explain it?(KIENGIR (talk) 22:47, 25 February 2021 (UTC))
No need to be afraid; if you don't like the results of this, uncheck those options. You want to see all edits except bot edits. Until T11790 is fixed, this is the only way I know of to accomplish that. 'Group changes by page' groups the changes into collapsed lists so that grouped entries on your watchlist watchlist look something like this (except a lot of it will be wikilinked):
► 22:47 Wikipedia:Village pump (technical) 24 changes history +8,012 [TheDJ; Suffusion of Yellow; Roxy the dog; PrimeHunter; Pelagic; ONUnicorn; KIENGIR; Dreamy Jazz; 0mtwb9gd5wx; Slywriter (3×); Xaosflux (4×); The Anome (8×)]
When you click on the ►, it changes to ▼ and shows the changes for the page sort of like this:
▼ 22:47 Wikipedia:Village pump (technical) 24 changes history +8,012 [TheDJ; Suffusion of Yellow; Roxy the dog; PrimeHunter; Pelagic; ONUnicorn; KIENGIR; Dreamy Jazz; 0mtwb9gd5wx; Slywriter (3×); Xaosflux (4×); The Anome (8×)]
22:47 (cur | prev)..(+828)..KIENGIR (talk contribs) (→Watchlist parameter rollback)
22:08 (cur | prev)..(+1)..Slywriter (talk contribs) (→Edit filter to prevent http: //%5B(and cousins) making good links bad: Indent (Tags: Mobile edit Mobile web edit)
22:08 (cur | prev)..(+793)..Slywriter (talk contribs) (→Edit filter to prevent http: //%5B(and cousins) making good links bad: Thanks PrimeHunter and question on long term Maintenance (Tags: Mobile edit Mobile web edit)
etc
Click ▼ to collapse the list. Line wrapping in the real watchlist is prettier. Try it. If you don't like, undo it.
Thank you very much for the demonstration...well maybe once I will try, but as I see I have to really wait for the fix...(KIENGIR (talk) 19:32, 27 February 2021 (UTC))
As far as I know, "Group changes by page in recent changes and watchlist" is irrelevant in this equation. Older edits to pages whose latest edits are hidden show up as long as "Expand watchlist to show all changes, not just the most recent" is checked. Nardog (talk) 15:20, 28 February 2021 (UTC)
Optional Class Parameter
Someone recently pointed out to me that the class parameter for a Wikiproject banner is better left blank for Drafts and Redirects because those articles are likely to change in the future and the class is auto-populated simply by virtue of the article being in Draft or Redirect space. I was curious what the standard conventions for WikiProject banners in general, but I'm specifically interested in whether it would be better to have a shortened WikiProject banner for Categories and Files because they are unlikely to change, but would be auto-populated. Would a short banner without a class parameter be best or should the class be included? Or if it doesn't matter, what would you suggest? I would assume that it's best to keep a standard convention across a large number of articles so if I'm making sure it's one specific way for an entire WikiProject what would be ideal?
I've never used the Village Pump so let me know if this is the wrong place. I figured it was a technical question because I want if there is a benefit to adding optional parameters. TipsyElephant (talk) 02:18, 28 February 2021 (UTC)
I think most WikiProjects also auto-assess class on categories and files. So I think the answer to Would a short banner without a class parameter be best? is yes? @Redrose64: --Izno (talk) 18:45, 28 February 2021 (UTC)
Not just categories and files - rather than listing the namespaces explicitly, it's easier to say "all talk namespaces except for Talk: itself". More specifically, if the |class= parameter is omitted or left blank, all WikiProject banners that have class ratings (with the exception of {{WikiProject Military history}}) will automatically set the class if either (i) the subject page is a redirect; or (ii) the subject page is not in main (article) space - if both apply, it's classified as a redirect. In the case of {{WikiProject Military history}}, which doesn't autodetect redirects, it will automatically set the class if the subject page is not in main (article) space. So, generally speaking, the |class= parameter should only be filled in if the subject page is an article or disambiguation page. Remember that pages in Draft: space are not articles (although they may become articles when developed sufficiently) so WikiProject banners in Draft talk: space should not have the |class= filled in. --Redrose64 🌹 (talk) 00:18, 1 March 2021 (UTC)
@Possibly: it isn't "allowed there" it is just ignored - when you put in whitespace you don't trigger the link generator - the only reason you are then seeing the link is that a separate link generator is formatting your bare link. — xaosfluxTalk10:47, 1 March 2021 (UTC)
My ignorance of technical matters is complete. My IP address is currently subject to a range block.
(1) The instructions given to IPs are confusing and misleading, and I think they ought to be changed. Is this the right place to raise this issue?
(2) Up until today, this has not been a problem for me, since I can simply log in to my account. Today, the system kept telling me that I was not allowed to edit Wikipedia because my account had been blocked – it was logging me out in the middle of a simple edit. Sweet6970 (talk) 12:40, 21 February 2021 (UTC)
Depends on whether it’s a global block on Meta or a local one on English Wikipedia. Can you share part of the message you receive, Sweet6970? (No need to divulge the IP address if you’re not comfortable with that.) — Pelagic ( messages ) – (08:15 Fri 26, AEDT) 21:15, 25 February 2021 (UTC)
The wording I receive as an IP is
"""""""""""""""""""""""""""
View source for Melvyn Bragg
← Melvyn Bragg
Jump to navigationJump to search
You do not have permission to edit this page, for the following reason:
Your account has been blocked from editing Wikipedia.
This does not affect your ability to read Wikipedia pages.
Editing from 2A02:C7D:0:0:0:0:0:0/33 has been blocked (disabled) by Drmies for the following reason(s):
Persistent addition of unsourced content: Persistent addition of incorrect content: change to site-wide
This block has been set to expire: 13:15, 29 June 2021.
Even when blocked, you will usually still be able to edit your user talk page and email other editors and administrators.
For information on how to proceed, first see the FAQ for blocked users and the guideline on block appeals. The guide to appealing blocks may also be helpful.
Other useful links: Blocking policy · Help:I have been blocked
"""""""""""""""""""""""""
This wording is incorrect for me, since my account is not blocked, it is my IP address which is blocked. More importantly, it is incorrect for the IPs, because there is no point in appealing, they have to create an account if they want to edit.
I am no longer having trouble being logged out, but I am still concerned about the wording, which does not make it clear to IPs that if they want to edit, they have to create an account. You can see that 8 people have created a Talk page in order to appeal, which is not what the intention is. I have checked some of the IP addresses, and found various parts of London, Middlesbrough, and Glasgow, and I am wondering if everyone in Britain who uses my ISP is blocked from editing as an IP, and not being told that they have to create an account if they want to edit.
Generally IP blocks extend to accounts created using that IP or the IPs cannot create new accounts. Otherwise they would be too easy to avoid, and telling the users how to circumvent a ban wouldn't be a good idea either even if it would be possible. Things can cascade from there, see Wikipedia:Autoblock. --mfb (talk) 12:18, 26 February 2021 (UTC)
@Mfb: In this case, the intention is that any IP affected should create an account. See [7] The range of this block is absolutely enormous, and as I said, it looks like the range block covers anyone in Britain who uses the same ISP as I do. But the problem is that the instructions received by the IPs do not tell them that they should create an account. Sweet6970 (talk) 12:32, 26 February 2021 (UTC)
@Sweet6970: Typically, any significant range block should use the {{rangeblock}} or {{anonblock}} (or {{schoolblock}} or similar) templates, which I believe are relatively clear. Most admins are quite good about doing this. I don't enjoy taking over such large range blocks, which is what would happen if I were to change the message; what we can do is ping the blocking admin User:Drmies, or you can head over to his talk page, and ask him simply and nicely to please add a template. -- zzuuzz(talk)12:45, 26 February 2021 (UTC)
@Zzuuzz: Thank you for your comment. It looks like the {{anonblock}} would be the most suitable, but it still does not actually say ‘You are allowed to create an account, and this is what you should do’. (Under this particular block, there is no need to make a special request to create an account.) I did raise the question of this range block with Drmies in January. [8] The reply I received [9] felt like a brush-off, and I felt that it implied that I was on the side of vandals. I expect that if I raise this matter again I will be told that this is none of my business, so there is no point. Sweet6970 (talk) 13:41, 26 February 2021 (UTC)
I believe your previous note to Drmies didn't clearly identify the problem, or the solution (which you probably weren't to know about). But no worries, hopefully my ping from here will clarify the issue and result in a positive outcome. If not, ping me again, and I'll begrudgingly take over Drmies' block with a suitable message. Generally, the anonblock template is tried and tested, and I think it should be good enough, but suggestions to improve the wording are always welcome. -- zzuuzz(talk)13:50, 26 February 2021 (UTC)
I have to admit I don't quite get the problem. If it's that the block notice isn't inclusive enough, sure, someone higher up on the technical food chain can fix that, I suppose. But still, Sweet6970's note on my talk page was odd: "It looked like it would have been possible for me to create an account (I didn’t go through with it)"--well, they already had an account. The template's second sentence is "you are still able to edit if you sign in with an account", and that links Wikipedia:Why create an account?, where as-yet unregistered users can create an account. Many if not all as-yet unexperienced editors are able to "work this out", to use Sweet6970's phrasing, so why not here? And "If the idea is that anyone who is affected by the block should create an account, then the instructions need to be improved to give a clear route to this option"--well, the instruction is right there, and it's clickable. So no, I'm not saying this is none of their business, and I appreciate advocacy, and if someone thinks the template should be improved, go for it. Drmies (talk) 15:59, 26 February 2021 (UTC)
@Drmies: (1) Regarding my note being ‘odd’: Yes, of course I have an account. Now. But I am looking at this from the point of view of someone who does not have an account, and remembering the bafflement, frustration, and annoyance which I felt when I was subject to a range block before I created my account.
(2) I don’t understand what you mean when you say The template's second sentence is "you are still able to edit if you sign in with an account". That’s not what I get when I go to edit logged out. I have already copied the wording in my previous post above:
“”””””””””””””””””””””””””””””””””””””””””””””””””””””””
Jump to navigationJump to search
You do not have permission to edit this page, for the following reason:
Your account has been blocked from editing Wikipedia.
This does not affect your ability to read Wikipedia pages.
Editing from 2A02:C7D:0:0:0:0:0:0/33 has been blocked (disabled) by Drmies for the following reason(s):
Persistent addition of unsourced content: Persistent addition of incorrect content: change to site-wide
This block has been set to expire: 13:15, 29 June 2021.
Even when blocked, you will usually still be able to edit your user talk page and email other editors and administrators.
For information on how to proceed, first see the FAQ for blocked users and the guideline on block appeals. The guide to appealing blocks may also be helpful.
Other useful links: Blocking policy · Help:I have been blocked
“””””””””””””””””””
There is nothing here about creating an account. The instructions lead you to appeal the block (as several IPs have done) which is pointless.
(3) Your block prevents people from all over Britain from editing as IPs, and does not notify them that they have the option to create an account. Sweet6970 (talk) 16:44, 26 February 2021 (UTC)
Would it not be possible to cancel the existing block and immediately replace it with a block using the anonblock template? Alternatively, the template which is currently in place includes your own wording for the reason for the block. Presumably, you could add something here like ‘If you wish to make constructive edits to Wikipedia, please create an account in order to do so’? This would be clearer for IPs subject to the block. Is it possible to cancel the existing block, and immediately replace it with another block with the same template, but with the additional wording? Sweet6970 (talk) 18:33, 26 February 2021 (UTC)
Hi, just trying to reconcile eveything above. Is the technical problem that we are trying to resolve mostly: Logged in users that are using an IP address range that is hard blocked are getting the wrong text from MediaWiki:Blockedtext's headline. ? — xaosfluxTalk19:20, 26 February 2021 (UTC)
Sorry, no. The problem is that IPs who are blocked under a very large range block are not getting the message that they should create an account if they want to edit. (I know very little about these things, but I think this is what is called a 'soft block', because my IP address is covered by the block, but I usually have no problem editing if I log in.) Sweet6970 (talk) 20:01, 26 February 2021 (UTC)
Indeed. The issue is that the default (interface) block message is not helpful, compared to having the {{anonblock}} template in the block (log) message. This is why we have the block log templates. The range is actually soft blocked with AC enabled. All we need is for someone to copy-paste {{anonblock}} into the current block reason. -- zzuuzz(talk)20:05, 26 February 2021 (UTC)
Indeed it should, and that I can't readily explain. On another IPv4 range, I've only been able to confirm that MediaWiki:Blockedtext-composite is working correctly, but I'm not in a position to check this situation. However, I think my point remains, this block - any block of this size - needs a block log template. -- zzuuzz(talk)20:49, 26 February 2021 (UTC)
@Zzuuzz: You first mentioned the {{rangeblock}} and {{anonblock}} templates in your post of 12:45, 26 February 2021 (UTC). If I block a range of IPs (something I've never done, but recently I have had cause to), where should these templates be placed? I normally put templates like {{subst:uw-vblock}} on the user talk page - but an IP range may have hundreds (if not thousands) of talk pages. Am I expected to put the template on all of them? --Redrose64 🌹 (talk) 21:31, 26 February 2021 (UTC)
Any template placed in the block reason is transcluded to the user. So, if you pick one of the templates in the dropdown list (the bottom half of this list), the full template will be displayed when the blocked user tries to edit. Every IP in the range will see the transcluded template when they try to edit. You can also just place the template in the 'custom reason' field, or mix it up as I recently did here. That user will get the full username block template, plus the custom message I added. -- zzuuzz(talk)21:46, 26 February 2021 (UTC)
@Zzuuzz: I’m sorry to trouble you. The message I get when I go to edit logged-out is still the same. You have said that all we need is for someone to copy-paste anonblock into the current block reason. Presumably only an admin could do this. Could you fix this, please? Sweet6970 (talk) 16:45, 1 March 2021 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
Wikis using the Growth team tools can now show the name of a newcomer's mentor anywhere through a magic word. This can be used for welcome messages or userboxes.
A new version of the VideoCutTool is now available. It enables cropping, trimming, audio disabling, and rotating video content. It is being created as part of the developer outreach programs.
Problems
There was a problem with the job queue. This meant some functions did not save changes and mass messages were delayed. This did not affect wiki edits. [10]
Some editors may not be logged in to their accounts automatically in the latest versions of Firefox and Safari. [11]
Changes later this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from 2 March. It will be on non-Wikipedia wikis and some Wikipedias from 3 March. It will be on all wikis from 4 March (calendar).
'Publish changes' should ask for confirmation but doesn't
When finished editing a Wikipedia article and clicking 'Publish changes', your changes go live, immediately and irrevocably. That is arguably reckless and wrong.
When editing, I will click 'Show preview' incrementally, many times, to make sure each change I make is correct. But it is very easy to click the adjacent 'Publish changes' button by mistake. Especially because 'Publish changes' is hi-lited in blue but 'Show preview' is white, visually tempting you in a moment of inattention to click the former when you didn't mean it.
Allowing all comers to commit, with a single click, a permanent write action on a Web site used by untold millions egregiously fails to err on the side of caution. A good user interface, let alone publishing to a world-wide audience, demands that on clicking 'Publish changes', you put up a box asking 'Are you sure you want your changes to go live? Yes / No,' with 'No' the default, hi-lited in blue.
It should obtain for all editing on Wikipedia—articles, Talk pages, this Village Pump page, all contexts.
I, for one, desperately want a safeguard against my own inattention.
This is unlikely ever to occur for the 2010 wikitext editor. There are simply too many people who want to be able to edit now, and most of them are the power editors who care.
Allowing all comers to commit, with a single click, a permanent write action on a Web site used by untold millions egregiously fails to err on the side of caution. is overblown. Bad actions are readily reversed. Good actions - well, why are you complaining? :)
I have a list of ~80,000 pages that I am fetching daily. Is there a way to find out which pages in my list have been edited in the last day, so that I can fetch only those? Sam at Megaputer (talk) 19:53, 1 March 2021 (UTC)
@Firefly:: I'm fetching the pages through the API so that they can be evaluated for promotional content as part of an anti-promo tool I've built. Right now the major limitation I am facing is the computing resources, so the more efficient I can make this thing the more pages we can defend. Can you link me to the documentation on those database queries? Sam at Megaputer (talk) 21:02, 1 March 2021 (UTC)
@Sam at Megaputer: so as a quick example this query pulls in all pages that have been touched in the past 24h (touched means edited, had protection level changed, had included templates changed, etc.). I’ve limited it to return only 1000 results for now, but obviously you could tweak it for your own use. Some documentation here about querying the replicas on Toolforge if it helps. ƒirefly ( t · c ) 21:25, 1 March 2021 (UTC)
Assuming it is the same set of 80k, SD0001's idea is far better than mine. I had no idea you could use the watchlist via the API. ƒirefly ( t · c ) 15:32, 2 March 2021 (UTC)
You can also fetch the watchlist using RSS format with a secret token, may be even easier since it is basic authentication. — xaosfluxTalk15:40, 2 March 2021 (UTC)
I'm assuming you mean purely when displayed somewhere, as opposed to being part of the image itself? Where is this needed & why? Thanks! ƒirefly ( t · c ) 11:40, 2 March 2021 (UTC)
I don't know whether image syntax can do it but you can display an image without a border and surround it with a border made in other ways. PrimeHunter (talk) 12:06, 2 March 2021 (UTC)
I'm trying to get Fantendo's Template:MenuSprite to have an optional border round it, which it now does; I was missing the word 'solid'. Using div forces the sprites on to several lines, so I've used span instead. However this does mean the boxes are running on to each other, but I can play around from here.--Launchballer12:21, 2 March 2021 (UTC)
Wikipedia is not affiliated with Fandom. Please say when your post is about another wiki. Your code [12] has mismatched <span>...</div>. PrimeHunter (talk) 12:48, 2 March 2021 (UTC)
See WP:EIS#Border: images not having either |thumb or |frame may be given a |border option, it's styling is due to the <img /> tag being given the attribute class="thumbborder" for which our style sheets have this rule:
Just checking in to see if anyone else has information for me about mw:Talk pages project/Replying. I'm talking to the devs about offering this as a default-off Beta Feature here. There are lots of comments at Wikipedia talk:Talk pages project#Experiences from people who've been using it, but we might be biased. ;-) So if you're aware of any problems that make you nervous about having more people try it out, or if you've been secretly cleaning up messes every time I use it and you just never wanted to mention it, please ping me. (Try it out here: https://en.wikipedia.org/wiki/Wikipedia:Village_pump_(technical)?dtenable=1 and look for the [reply] buttons. If you switch to the visual mode, it'll make it really easy to @-mention editors. As long as nobody breaks the page with, say, an unescaped half of a parser code, an unclosed div tag, or a broken wikitext table, then it should work even on a page as big as this one.) Whatamidoing (WMF) (talk) 00:45, 23 February 2021 (UTC)
I love the MW tool a lot more than the old reply-link system. It looks clean and I like the visual editor. Woohoo. Aasim (talk) 16:51, 2 March 2021 (UTC)
When you've got it working, please install it everywhere which is not an actual article. For example, on this very page. Using wiki software for discussion is Wikipedia's original sin. Rollo (talk) 22:25, 2 March 2021 (UTC)
Two-dimensional schematic diagram
I would like to do a two-dimensional schematic diagram to represent a streetcar network of 3 north/south lines and one interconnecting east/west line. I would like to use Template:Routemap but I would want the east/west line to be a horizontal line instead of twisting it into a vertical line. Thus, the schematic would resemble the city street grid. However, I suspect I would have problems indicating and labelling east/west stops. Is there any existing examples of doing this?
My understanding is that user.user_editcount (and thus the editcount field of API:Users) is back-filled by an external process and lags real time. Anybody know how much lag is typical? I'm working on improving performance of SPI Tools by cacheing in redis, which leads directly to how to do cache invalidation intelligently. My current thought is to use the user's editcount. If it lags by a few seconds, that's perfectly fine. If it lags by a day, that's going to suck. -- RoySmith(talk)16:33, 2 March 2021 (UTC)
I just tested it. I ran this oneliner, and then I started editing:
From my limited sample (3 edits), it updated within a second after making the edit. So it should be fine for you to use. --rchard2scout (talk) 09:29, 3 March 2021 (UTC)
Hello, the helpful tool that finds all articles of a certain quality and/or importance in a Wikiproject produces duplicate listings sometimes. For example, the list here has multiple entries for Greece runestones. Luckily it's a small list, but half of the items are duplicates of the same article, which might go poorly on a longer list. Oddly the table on Wikipedia:WikiProject Writing systems correctly tabulates the number of FAs as five. Any insights appreciated, CMD (talk) 10:23, 4 March 2021 (UTC)
Hi, When I warn a user or revert a vandal on an article I've never edited before it adds the user or article to my watchlist but for 31 days only - Is there way this can be changed so that it can be added permanently like it used too?, Thanks, –Davey2010Talk16:15, 4 March 2021 (UTC)
Hi Xaosflux, Thank you so much!, Looking at the Twinkle preferences there were a lot of entries that said "watch for a month" so I would assume my issues was indeed related to that, I've changed everything back to indefinitely so hopefully all will be fine again, Thank you for your help it's greatly appreciated, Many thanks, –Davey2010Talk16:36, 4 March 2021 (UTC)
Two edits at once?
Just curious, especially for admins running deletion scripts: Is it possible to make two actions at precisely the same moment? Heymid (contribs) 22:03, 4 March 2021 (UTC)
Surely it depends what you mean by "the same moment". There is always going to be an order to edits—one will have a greater revision ID than the other—but it's possible for many edits to have the same timestamp, which only has precision up to the second. — The Earwig (talk) 00:27, 5 March 2021 (UTC)
Hi @Steve M! I'm making the final preparations to post announcements about the script to various places, including English Wikipedia. It was initially created for Russian Wikipedia, so substantial efforts had to be made to internationalize it. I've drafted a configuration file for English Wikipedia in my user space, but I'm only superficially familiar with English Wikipedia's customs, templates, etc., so I'd be happy if someone helps to make fine-tuning for EnWP (see the instructions) and possibly other wikis. (The script works quite well even without a configuration file, but it would definitely improve user experience in several respects.) Then wikis can decide if it's worthy to have it included in the gadget list. Indeed, having the script run as a gadget would make it load a bit faster and easier to install. Jack who built the house (talk) 14:43, 2 March 2021 (UTC)
Wow, this tool looks really amazing! It even notified me live when someone replied to be me in another section. – SD0001 (talk) 15:51, 2 March 2021 (UTC)
Pages where the script should and shouldn't run. So, if you see a page where the script wrongly run or haven't run, if it's not a one-off, it could make sense to add it to a blacklist/whitelist.
Comments that the script shouldn't detect as such. For example, the script doesn't need to detect comments in technical templates such as {{Moved discussion to}}, as it doesn't make sense to reply to them anyway. The script uses class names and template names to filter out such templates from both web view and wikitext.
More grained configuration includes various details on
which discussions should be considered closed/archived, based on class names and page names,
which templates should be used for different purposes,
what wikitext patterns should be taken into account when handling wikitext,
which elements shouldn't be highlighted when highlighting a comment, based on class names,
and so on. I've already added some of this info for English Wikipedia, but there could be more.The source of the English Wikipedia configuration is here. Since I'm just learning English, I could have also messed up with interface translation, so your contributions are welcome here (or you can contact me directly if you don't have a GitHub account). Translations to other languages are at Translatewiki. Bug reports are accepted at Phabricator or Commons. Thank you very much! Jack who built the house (talk) 17:23, 2 March 2021 (UTC)
Jack who built the house, one issue is that rolling over sections does not always show the edit, link, and thank button. It starts showing after you reply to a subsection once. It is possible to investigate this bug? Thanks, Steve M (talk) 00:49, 3 March 2021 (UTC)
@Steve M: Do you mean the comment menu not showing up when you hover over a comment? And the comment isn't being highlighted with blue? Does this apply to all comments or only some specific ones? Does "not always show" mean "sometimes it does, sometimes it doesn't (with the same comment)"? Or with different comments?A screenshot would be nice. The behavior you mentioned could also mean a JavaScript error. Please, open the browser's developer tools (usually F12 will do it), go to the "Console" tab and copypaste the text of the errors (on red or yellow background).It starts showing after you reply to a subsection once. That possibly means that this happens after a page refresh. Does clicking the reload button on the navigation panel (on the left side of the page) give the same effect? Jack who built the house (talk) 01:07, 3 March 2021 (UTC)
@Steve M: Hm, so, at the first page load, it doesn't work. But after a refresh by means of the script (reply / refresh using the navigation panel) it works.Which browser and Wikipedia skin do you use?This looks to me as an effect some script or gadget could give. They could mess with the page's content and "kill" some events attached to elements. This could explain why after a refresh it works. I've tried to add all your user scripts, but I didn't see the bug. I wonder if it could be a gadget that you have installed. Could you try disabling your gadgets and then, if the problem is gone, enable them one by one to see which gadget causes the problem? Jack who built the house (talk) 01:22, 3 March 2021 (UTC)
@Steve M: Hm, could it be "(U) Change UTC-based times and dates, such as those used in signatures, to be relative to local time (documentation)" gadget? I have tested it, and it indeed gives the effect you mentioned. To use Convenient Discussions together with that script, you need to turn it off as a gadget and import it like this in your common.js:
mw.hook('convenientDiscussions.commentsReady').add(function(){importScript('User:Gary/comments in local time.js');});
Jack who built the house, thanks for figuring out the issue. Best to show this on the troubleshooting page, so that other users will find it out. Disabling it did fix the issue. Is it an issue with Convenient Discussions or that script?~ Steve M (talk) 01:39, 3 March 2021 (UTC)
That script alters timestamps, converting "(UTC)" to "(UTC±N)" and transforming dates, thus making the timestamps invisible to Conveneint Discussions, which relies on a predefined date format.Best to show this on the troubleshooting page Well, it is currently mentioned in the "Installation" section, I just didn't know it's a gadget in EnWP. Maybe a "Troubleshooting" section is needed as well, I'll look into it. Jack who built the house (talk) 02:08, 3 March 2021 (UTC)
Templates differ between sites, and link syntax is present everywhere.
The user might want to change the displayed name, and it's easy with the link: press ⇧ Shift+↵ Enter, and you can enter the displayed name, for example @Steve. You can do that with the {{u}} template: {{u|Steve M|Steve}} → Steve, while with the {{ping}} template you have to write like this: {{re|Steve M|label=Steve|p=}} → @Steve, which is too complex. Also, usernames with = are allowed, and you would have to put 1= before the parameter content in these cases.
The reply to section always shows. I am talking about the link, edit, and reply think that shows when you roll over a comment and enables you to reply to any comment in one click. The reload button on the navigation panel fixes it until I go to another page. Reloading the page via browser resets it, and I have to reply to a subsection again in order for it to work. Steve M (talk) 01:10, 3 March 2021 (UTC)
Stop Chrome from scrolling the editing box when I press enter
I asked this already at the reference desk but was unable to arrive at a diagnosis or solution. When I am editing Wikipedia, Chrome has some weird annoying feature that is horrendously annoying. Usually, but not always, when I press the enter key, the editing textbox will scroll so that the new line is at the very top of the text box, rather than the text box staying where it is and the new line pushing the text down under it (as is the expected behaviour). This only happens on Wikipedia (although I have not found any sites that use plain multiline text boxes to test... everything is script-based these days), and it happens on every computer I try it on. It happens logged out as well as in incognito mode, so this is the behaviour that IPs would be experiencing unless they use the visual editor.
I've used Chrome for years, but this behaviour has only been happening to my knowledge for the past several months. How do I stop text boxes from scrolling when I press enter? I'm up to date with v88.0.4324.150. - Floydianτ¢17:46, 21 February 2021 (UTC)
This happens to me when I paste in one or more lines of text, but it is not consistent. Subsequent pastes do not always make it happen. Using Chrome on Lubuntu.--Verbarson (talk) 15:20, 22 February 2021 (UTC)
Gonna try one last bump... otherwise it looks like I have to pick between the inconvenience of scrolling text in Chrome, or of pages refreshing when I press the back button on Firefox and me losing my changes. - Floydianτ¢06:21, 4 March 2021 (UTC)
The default behavior for the Firefox backspace button is changing soon, on that note. I'm not sure when, since it was originally expected to be in Firefox 86 but my browser still goes back when I hit backspace. Even today though, there is an about:config option to turn that off.[13].
The core issue may need reporting on Phabricator. At least then it will not be lost. I can say that I have had issues with syntax highlighting with Firefox for mobile, but that's not your problem apparently. --Izno (talk) 20:26, 4 March 2021 (UTC)
I meant when I have a big unsaved edit and accidentally leave the page through an image to commons, press the back button, and get a fresh edit page without any of the changes I typed in. In Chrome, the text is exactly how I left it. - Floydianτ¢00:19, 5 March 2021 (UTC)
I think that the way MediaWiki communicates with browsers has changed. My Firefox version hasn't changed lately, but its behaviour has. Example: when editing, I often use the "Show preview" feature; in that, I might click one of the links I just added to ensure that it goes to the right page. When returning from the linked page to the preview screen, Firefox used to present me with a normal edit page, with the "This is only a preview; your changes have not yet been saved! → Go to editing area" box at the top. In the last few days, this has changed so that on returning from the linked page to the preview screen, Firefox now throws the message
Document Expired
This document is no longer available.
The requested document is not available in Firefox’s cache.
As a security precaution, Firefox does not automatically re-request sensitive documents.
Click Try Again to re-request the document from the website.
I've been seeing something similar with chrome. The message I get is:
Confirm Form Resubmission
This webpage requires data that you entered earlier in order to be properly displayed. You can send this data again, but by doing so you will repeat any action this page previously performed.
Press the reload button to resubmit the data needed to load the page.
ERR_CACHE_MISS
From that screen, F5↵ Enter to get back to where I was.
What should we do when content in non-Roman characters or ideograms is added to an article, but is not then rendered correctly? For example the article on the Tamil language includes Tamil words which are not rendering in my browser. — Cheers, Steelpillow (Talk) 16:02, 2 March 2021 (UTC)
OK, thanks. Presumably that is the answer - learn to install fonts on your system, or weep. Not a very user-centric solution, is this really the best web experience that the 21st century Mobile era can deliver? — Cheers, Steelpillow (Talk) 18:13, 2 March 2021 (UTC)
Steelpillow, pretty much. Many operating systems do come with most fontsets (95% of common scripts). Just android sometimes has missing fonts, to save on storage space and it's possible that Windows was not installed with full language support (also as a space saving measure). And then there is Linux of course, but then you'd be used to things like this. There are some techniques to deliver fonts directly to the browser, but its.. complicated. See also mediawikiwiki:Universal_Language_Selector/WebFonts. By default we don't use that unless we know that the language support for a language edition of a project is very problematic even for it's own user base. —TheDJ (talk • contribs) 11:09, 3 March 2021 (UTC)
Thank you, especially for the link. What a mess you guys have to cope with! Hats off to you for taking it this far. Seems to me the Unicode and w3c folks should get together to define a set of web-compatible default fonts for all languages accepted into unicode: no default font, no Unicode ratification. But hey, they've only coexisted side by side for 30 years, we can't expect miracles. Hardly your problem, though. — Cheers, Steelpillow (Talk) 12:40, 3 March 2021 (UTC)
It kind of works the other way around: without assigned Unicode points, a foundry can't create a font file including the required characters (other than placing them into the private use area, but that has limited usefulness). But in any case, there are people who work on creating reference typefaces for all Unicode characters. The problem is one of bandwidth: web fonts is a widespread solution that works great when a manageable set of glyphs is enough, but leads to font downloading delays when many glyphs are needed. To help improve responsiveness, web developers will create custom font files that just contain a necessary subset to display all of the characters actually used on a web site. Of course, for sites based on user-generated content, that's not possible. isaacl (talk) 21:50, 3 March 2021 (UTC)
There is a difference between a placeholder for a future character set and a ratified release of the standard, but that discussion's just chicken-and-egg nit picking. And if language fonts took up a quarter as much bandwidth as spyware does, they would be pretty dam' fast. But mainly, "Oh, that's bloatware, we better not do that" does not feature strongly in the history of the web; would they really add an impractical amount to a browser's zipped executable? Treat new language sets as updates to the browser, only download 'em once. A little systems engineering can go a long way in making good things happen. Anyway we are going off-topic. I have said more than enough here. — Cheers, Steelpillow (Talk) 10:00, 4 March 2021 (UTC)
Typically people rely on their OS supplier to supply fonts for language support. (They're a big part of the impetus to get new Unicode revisions out and new fonts made, as it enables their revenue.) Optimizing web page downloads is very much a concern for high traffic sites, but nonetheless, downloading and using fonts automatically via web fonts is common now. isaacl (talk) 06:03, 5 March 2021 (UTC)
@Steelpillow: The w3c folks concern themselves with fonts (as a set of glyphs of differing shape) only in a very arms-length way. The HTML working group deprecated and then obsoleted their <font>...</font> tags, and for 20 years (plus) have said "this is purely a presentation matter, so use CSS"; and the CSS people in their turn do their utmost to avoid saying what a font should look like except in the most general terms (normal weight, boldface, upright, italic, size etc.). In particular, they don't state what characters should be included in any font family, other than the U+0020 space char. The CSS Fonts Module Level 3 defines a number of properties but you need to look at the spec for the font-family: property in order to see examples of use - no actual font names are defined apart from five generic families, and even those are not explicity described except in somewhat general terms (it even says They are also not guaranteed to always be different from each other.). It does say what a browser should do if the specified font family is not available.
@Redrose64: Indeed. All of which is precisely why the problem which brought me here occurs, and why we Wikipedians will be stuck with it unless and until the w3c and Unicode folks decide to work together and do things better. — Cheers, Steelpillow (Talk) 17:30, 5 March 2021 (UTC)
I'm attempting to clean up the formatting of this section, but am having difficulty. The 2010 census description is nudged far to the right by a table, and there is a large blank area before the 2000 census description that appears to be due to the historical population table and racial dot map that appear on the right. Can someone assist with cleaning this up? Home Lander (talk) 21:08, 5 March 2021 (UTC)
@Home Lander: You could try putting a {{clear left}} between the table and the text. But even before we get to fixing the layout, there are some horrible accessibility issues with that table. Get rid of the id="toc" attribute - that is the same id as is used by the page's autogenerated Table of Contents, and ids must be unique. Then lose the font-size: 75%; declaration, it's below what is permitted by MOS:FONTSIZE and it's too small for me to read, consequently it's certainly far too small for RexxS (talk·contribs) to read. You should also drop the width: 39%; declaration and the cellspacing="3" attribute, it would be a good idea to replace the latter with class=wikitable. The whole table seems to be marked up with data cells, even the caption, column headers and row headers which must all use the appropriate markup. For example, replace these two lines
|-
!colspan="3"|'''Huntsville Demographics'''[[United States Census, 2010|2010 Census]]
with the single line
|+ Huntsville Demographics [[United States Census, 2010|2010 Census]]
Redrose64, thanks. I think that's a fair bit better. I stumbled upon this section by chance and found it to be quite a mess. Appears more readable now. Home Lander (talk) 21:48, 5 March 2021 (UTC)
Toolforge and MediaWiki.org - "no secure connection" on WiFi
Hi. I am using Safari on an iPhone 6. When I try to load Toolforge or Mediawiki.org on my TalkTalk WiFi connection, I get a message from Safari saying it cannot load the page because "it could not establish a secure connection to the server". However, if I switch to Three mobile data, the pages load just fine. Anyone know why it does not work on the former? SK2242 (talk) 22:01, 6 March 2021 (UTC)
Edit conflicts and the rise of irrecoverability
My recollection is that in the last 15ish years, for the most part, if I were to encounter an edit conflict, I could simply go back and my original text would still be available. But in the past six months or so (estimate), I've found that the opposite has been happening. Now losing the original text in edit conflict seems to happen nearly every time, to the point that I've been forced to copy my text before entering a submission as a matter of course. But I still forget to do it sometimes, which then often leads to said super-annoying outcome. Not using any wiki-extras (whatsoever). I've had the sysop flag enabled starting 2005. El_C16:12, 5 March 2021 (UTC)
PrimeHunter, right, if that was the case (and sometimes it still is), I wouldn't even need to go back; again, going back, itself, used to also work well enough in such instances, but now rarely does. Hence, reaching the final stage of total irrecoverability. El_C03:36, 6 March 2021 (UTC)
Maybe it's unique to my setup somehow...? But, yes, it continues to be an nuisance. Oh well. Just gotta remember to always copy before submitting, I suppose.¯\_(ツ)_/¯El_C22:39, 6 March 2021 (UTC)
Which editor do you use and have you checked the bottom of the edit conflict window? Your edits aren't tagged so I guess you either use the default source editor like me, or wikEd at Special:Preferences#mw-prefsection-gadgets. I don't know whether wikEd can mess with edit conflict windows. PrimeHunter (talk) 23:12, 6 March 2021 (UTC)
Wikidata Query Namespace and relation to the English Wikipedia Gadget namespaces
A while ago, I noticed the Query namespace on Wikidata and the connections it has to the gadget namespaces. First of all, MediaWiki:Namespaceprotected is the message shown to users editing restricted namespaces. Unlike the Spanish Wikipedia File namespace (Archivo in Spanish,) attempting to protect the page does allow you to view the protection levels of the page. However, Gadget namespace on Wikipedia cannot be protected. This does not appear to be a built in feature with the gadget ns, as over at Miraheze this does not appear to be a problem. My observation: Any ns that cannot be edited by anyone locally will give an error when trying to protect the page. This is the same for the Query namespace. However, query ns appears to be even harder to be edited as d:Special:ListGroupRights returns zero results for an attempt to find (query-update) on the page, nor is the right listed anywhere on m:Special:GlobalGroupPermissions. Unlike gadget ns, Query ns appears to be a custom namespace for Wikidata. Is this ns related to query.wikidata.org? And who can edit pages in query ns? Just curious as it seems like a custom namespace for Wikidata that is not currently used and is protected for no specific reason. 54nd60x (talk) 13:51, 7 March 2021 (UTC)
Just thought I'd let people know of this bug with the Toolforge/Quarry replica DBs. It seems a recent change broke the actor table replications, and they no longer contain any rows for IP users. See this Quarry example. ƒirefly ( t · c ) 15:42, 7 March 2021 (UTC)
"where actor_user=NULL" is always null, so of course you got no results. Replacing that with the correct "WHERE actor_user IS NULL" times out - there's likely no index on actor_user - but you can show that the rows are still there by looking for a specific one, like in quarry:query/53047. —Cryptic21:47, 7 March 2021 (UTC)
Oh, hrm. I see what you're saying now - while the rows are there in the actor_revision view and actor_logging and actor_recentchanges and so on, they aren't in the more general actor view. (Your demonstration query's still wrong.) —Cryptic21:52, 7 March 2021 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
Section translation now works on Bengali Wikipedia. It helps mobile editors translate sections of articles. It will come to more wikis later. The first focus is active wikis with a smaller number of articles. You can test it and leave feedback.
When someone links to a Wikipedia article on Twitter this will now show a preview of the article. [15]
Problems
Many graphs have JavaScript errors. Graph editors can check their graphs in their browser's developer console after editing. [16]
Changes later this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from 9 March. It will be on non-Wikipedia wikis and some Wikipedias from 10 March. It will be on all wikis from 11 March (calendar).
The New Discussion tool will soon be a new discussion tools beta feature for on most Wikipedias. The goal is to make it easier to start new discussions. [17]
Future changes
There will be a number of changes to make it easier to work with templates. Some will come to the first wikis in March. Other changes will come to the first wikis in June. This is both for those who use templates and those who create or maintain them. You can read more.
Reference Previews will become a default feature on some wikis on 17 March. They will share a setting with Page Previews. If you prefer the Reference Tooltips or Navigation-Popups gadget you can keep using them. If so Reference Previews won't be shown. [18][19]
New JavaScript-based functions will not work in Internet Explorer 11. This is because Internet Explorer is an old browser that doesn't work with how JavaScript is written today. Everything that works in Internet Explorer 11 today will continue working in Internet Explorer for now. You can read more.
This is not an issue for this talk page, editors decide on both the titles of articles, and the categories they are in. And not all articles in the Romania category have "castle" in the title, e.g. Fortress of Arad or Dinogetia don't. Fram (talk) 17:08, 8 March 2021 (UTC)
@POS78: It appears you created most or all the "castle" articles. If you came here because you don't know how to change article titles then see Wikipedia:Moving a page. A page gets the capitalization you used when you created it, except the initial letter (of the whole title, not of individual words) which is automatically capitalized. PrimeHunter (talk) 20:26, 8 March 2021 (UTC)
I expect External Data is not an extension that will ever be enabled on en.wp due to the security/privacy implications of loading data from an arbitrary other place. That said, #2/3: You cannot. You must request it at phab:. You will need to show sufficient consensus here that it is something our community wants.
What data do you actually want in a specific template? That will be a starting place for showing consensus. An alternative would be to expose that data in Scribunto, which will (also) need change request and subsequent implementation, but which may be easier to persuade the technical side about. --Izno (talk) 05:50, 9 March 2021 (UTC)
I'll try to answer your questions here @Izno, as it sounds like this is the right forum. I am trying to get a current edit count for a user from the en.Wiki API using globaluserinfo and use it as an input for {{Service Awards}} so that it auto updates with no interaction from the user. As to your security concerns (which I think are quite valid), the External Data extension comes standard with a method to alleviate the majority of the concern. There is a local setting $edgAllowExternalDataFrom which sets a whitelist of URLs from which data can be pulled from. There is also the $edgAllowSSL local setting which can require HTTPS. These local settings applies to #get_web_data and #get_soap_data. This, combined with the fact that #get_file_data, #get_db_data, and #get_ldap_data (all of the remaining parser functions which actually retrieve data) must be explicitly allowed in the local settings would allow us significant control over security. Since all I need is the local en.Wiki API, I would be perfectly happy with the local settings $edgAllowExternalDataFrom = 'https://en.wikipedia.org';$edgAllowSSL = false;$edgHTTPOptions['sslVerifyCert'] = true;.
As to the other idea "to expose that data in Scribunto", (as a new editor) I don't know what that means. If that would be easier to gain consensus on and it meets the stated need, I would be fine with that. I just don't even know where to begin. Brvhelios (talk) 06:50, 9 March 2021 (UTC)
Yes, I am sure Yaron has done his homework. What I doubt is WMF's interest in reviewing security for any extension which does not have a sufficiently strong use case behind it.
Regarding edit count, phab:T213110 has already seen it declined for non-security related reasons, which doubles the unlikelihood that such an extension would be enabled. The next solution would be a bot to update some core data location of user edit counts (perhaps similar to how Ahechtbot updates transclusion counts), or if the user is sufficiently high in edit counts to do some string parsing of the list of Wikipedians by edit count (or even to update that process to store some back end more feasible for machine consumption so that a template could use that data more trivially too). That's two separate possible solutions, the former to be pursued at WP:Bot requests and the latter at that bot op's talk page (plus possibly a request at WT:Lua).
If you do not need everyone to be able to see a count, you might 3rdly consider using one of the Javascript solutions that I am pretty sure are available today. --Izno (talk) 14:58, 9 March 2021 (UTC)
Using OSM maps in Wikipedia, how do you create a mouse over for a pog.svg label?
I'm using an OpenStreetMap as a figure in a Wikipedia page and I need to placed pins (Red pog.svg) all over it in order to show distributions of a subject. Fine, I've got that far, but now here's the problem.
I really want the pin to hide the label, and only reveal the label when the mouse hovers over it. It can be done with a static map by using ( |position = none). eg:
Suppressing categories with the babel parser function
the babel parser function is great since it (1) doesn't require the creation of all the little "lang-level" userboxes, and (2) when the "lang-level" userbox doesn't exist, it doesn't try to transclude it (which helps reduce the entries in Special:WantedTemplates). unfortunately, it doesn't seem to have a "nocat" parameter (unlike template:babel). does anyone know if there are any plans to add a "nocat" parameter, or if it is possible with some undocumented method? this is particularly useful when displaying the userbox in the category where you don't want circular subcategories. thank you. Frietjes (talk) 21:17, 8 March 2021 (UTC)
thanks, good to know I am not the only one who wants this feature. hopefully, this will be implemented at some point in the near future, although no visible activity for a year doesn't look promising. Frietjes (talk) 17:39, 10 March 2021 (UTC)
It is recommended that this abuse filter range block /64 for IPv6
@IN: This seems like a good idea in theory, but it would have no effect here. The "block" action isn't even enabled. You'll need the support of communities where blocking filters are used. Look for $wgAbuseFilterActions['block'] = true; in this file. Suffusion of Yellow (talk) 21:12, 10 March 2021 (UTC)
Redlinked in NewPagesFeed
For some reason, even though I've got a global meta userpage that is copied into a local enWP userpage, my username would still show redlinked after the "Created by" in Special:NewPagesFeed. I think the script over there might still not be compatible with global userpages and probably need some tweaking. Assem Khidhr (talk) 10:10, 10 March 2021 (UTC)
Oh, it's been reported for almost 5 years now! I can see from Phab that a proposed fix, PageCuration should be using Title::isKnown() instead of Title::exists(), seems plausible. Do you have any idea why it is taking this long? Assem Khidhr (talk) 20:04, 10 March 2021 (UTC)
Five years is not very long in MediaWiki land, especially for a bug triaged as "lowest" priority. At quick glance Title::exists() is not being used anywhere in the PageCuration codebase, so I don't think it's that simple of a fix. — MusikAnimaltalk04:06, 11 March 2021 (UTC)
Proposal for bot to populate the size of categories over time
I’d like to change the name of the article Augusta, Lady Gregory to Lady Gregory, and I read the process described on Wikipedia:Moving a page. Perhaps somebody here could confirm for me: To change the name of the article I click on “Move” then a “Move Augusta, Lady Gregory” page drops down, and there is a box - the first line at the top of the box says “Move page”. The second line says “New title:” The third line contains two small boxes— the box on the left says “(Article)” (which I will leave alone). The box to the right says the current title “Augusta, Lady Gregory”. I assume in that box I should just delete the “Augusta,” leaving behind the new title “Lady Gregory”. Am I right about that?
Then I finish by typing into the “Reason” box a reason, and there are a couple of things to check, and finally I click on the blue button that says: “Move page”. Thanks. - GümsGrammatiçus (talk) 22:38, 10 March 2021 (UTC)
It is possible for an edit conflict to be missed by the software?
Last night, I made this edit to WP:ANI. I was replying to an editor's query in the middle of the section. I never went anywhere near the bottom of the section, yet another editor's comment was deleted. I don't recollect seeing that reply at the time I was replying, and the timestamp indicates that it is possible there may have been an edit conflict, as I did take a while to compose and post my reply. So, is there a possibility an edit conflict got missed by the software? Mjroots (talk) 06:44, 10 March 2021 (UTC)
@Mjroots: So lets look at the history. M Imtiaz makes an edit at 19:12, Serial Number 54129 makes an edit at 19:14, Levivich and TimothyBlue make several more edits starting at 19:16, then you save your edit at 19:22. Only SN's was wiped out; L's and TB's were merged succesfully. This is what would happen if you tried editing MI's revision after SN's was saved, but before L's was. Except, you were editing a section, and section editing isn't enabled on old revisions, unless you were somehow meddling with the HTML of the edit form. So I don't know what happened, but a bug seems plausible. Unless it can be reproduced, there's not much that can be done. It might involve some very unlucky timing. Suffusion of Yellow (talk) 21:58, 10 March 2021 (UTC)
Could someone help me upload the audio file of this conversation between AG Robert F. Kennedy, President John F. Kennedy and the Governor of Mississippi. I'm not sure how to do this (Commons & copyright is my weak point). Would this call be in the public domain because it was recorded by the federal government? Help would be appreciated. Not sure if this is the right place to ask. ~ HAL33301:58, 11 March 2021 (UTC)
The category was inappropriate for several of them as overcome by the passage of time. There should just be the two nominated and Template:Music, and Template:Music maybe should have a more dedicated discussion on that template's talk page if it needs to continue doing what it's doing (to wit, that's substituting images where text may be sensible these days). --Izno (talk) 03:50, 11 March 2021 (UTC)
Yeah, Template:Music probably doesn't need to do it today, but I don't think the category is necessary even so. --Izno (talk) 03:54, 11 March 2021 (UTC)
Thanks for the responses. I was not sure if it could be 'deprecation' only, for backward-support. And somehow the category was reduced to 2 ;-) The TfD can roll on. -DePiep (talk) 22:19, 11 March 2021 (UTC)
Edit raw watchlist, on time-limited watchlist entries
I like the "Edit raw watchlist" feature, as it allows my watchlist to be edited en-masse in my text editor, and it allows easy data portability should I decide to watch or unwatch nontrivial amount of articles at a time. (I also do not have client-side scripting enabled in my browser)
But as Wikipedia rolled out the time-limited watchlist feature, I found that I could not see which entries are non-permanent on the text-editable raw watchlist; and had to switch over to watchlist editor to see them. Also, neither places seem to allow me to change nor remove the expiration time— save for going into each applicable article, then unwatch and rewatch each of them manually. These are feature disparities that should be fixed.
While I have not checked the actual source code that underlies this feature, I have guesstimated that this is implemented by having extra timestamp field on each watchlist record that denotes its expiration date; and the actual expiration mechanism is kicked in whenever the user view their watchlist, or opened (raw) watchlist editor.
To present this information in the raw watchlist, I suggest extending each line with MediaWiki template parameter-like syntax, by appending "|" to each non-permanent entry, followed by property expire, followed with = and the value of expiration timestamp in (subset of) ISO 8601 format.
For permanent entries, they remain the same:
PAGENAME
For expiring entries, extend each of them with this syntax:
PAGENAME|expire=TIMESTAMP
For example:
Crystal earpiece|expire=2021-04-05T10:56:54Z
Notes on input processing:
If the list user submitted has entries with timestamp that is in the past, treat such entries as if they have been deleted off the list.
If the expiration dates really have to be limited in regard of length or granularity, round them up to the nearest value, or convert to non-expiring if that need be.
If invalid timestamp or delimiting syntax is entered, reject the edit, and return the list that user just entered, together with error messages (and if possible, whereabouts of each error).
Doing this would bring flexibility of raw watchlist editing up to date with the new feature introduced, allow better transparency and portability of user's own settings, and this method allows the raw watchlist to represent possible future properties that might get added to each entry as well.
The issue with the Uruguay national football team article was raised on my talk page by @Kokoeist: - in effect, when viewing this article on a mobile device, there is no option to maximise/minimise sections - everything is automatically maximised. Any idea why? It makes reading/navigation much harder. ~ — Preceding unsigned comment added by GiantSnowman (talk • contribs) 21:29, 11 March 2021 (UTC)
Collapsible content has never worked on mobile, which I assume is the question. If the headers were not collapsed, that may be a separate issue. --Izno (talk) 00:41, 12 March 2021 (UTC)
GiantSnowman, the page has around 1100 images, which is above the maximum of 1000 that MobileFrontend supports, meaning that the sections are not formatted to be collapsible. – BrandonXLF (talk)04:02, 12 March 2021 (UTC)
Toolforge and MediaWiki.org - "no secure connection" on WiFi
Hi. I am using Safari on an iPhone 6. When I try to load Toolforge or Mediawiki.org on my TalkTalk WiFi connection, I get a message from Safari saying it cannot load the page because "it could not establish a secure connection to the server". However, if I switch to Three mobile data, the pages load just fine. Anyone know why it does not work on the former? SK2242 (talk) 15:30, 12 March 2021 (UTC)
If you receive that message on wifi, that likely means there is someone who is attempting to downgrade your secure connection when using that ISP. You will need to contact the ISP in question to understand what they are doing and/or ask them to stop and/or ask if they understand what is going on. --Izno (talk) 19:37, 12 March 2021 (UTC)
Proposal for a "probably unreviewed Recent Changes" feature
Hi everyone. I'm wondering if there is community interest in the following feature.
Background: Sometimes I see hours-old or days-old obvious vandalism on my watchlist, and I imagine nobody was patrolling for vandalism at the time the edit was made. Here is an example of vandalism that any Recent Changes patroller would have caught, but it lasted for six days.[22] Anti-vandalism work, like everything else, is done when volunteers have the free time to do it. Sometimes there are plenty of people on Recent Changes patrol and they are racing to be the first to revert vandalism, and sometimes nobody is on vandalism patrol. The purpose of this proposed feature is to bring changes from those time periods to the attention of vandalism patrollers.
My proposal is this: I suspect that if you were to graph the (number of reverts per minute) against time, you would see some dips in the graph where the rate of reverts is nearly zero. These are the periods when changes were probably unreviewed. I'd like the Recent Changes page, and the log that feeds into tools like Huggle, to give users an option to view changes from these periods so I can check them for vandalism. I think having this option would enable the community to fight vandalism and BLP violations much more thoroughly, especially for under-watched articles. I've run this idea by the WMF Machine Learning team (although it's not really an application of machine learning) and they are watching this thread.[23]Clayoquot (talk | contribs) 22:28, 12 March 2021 (UTC)
@Clayoquot:WP:STiki did a very good job of surfacing such edits. It used User:ClueBot NG's machine learning to flag edits that were scored as being possible vandalism, but the bot for whatever reason did not revert. Alas, the server the STiki backend runs on has been down for a year now. I think reviving it or rewriting a new similar tool is probably the best solution here. ClueBot NG's scoring seems pretty good, but we could also go by mw:ORES which is in the Machine Learning team's ballpark.
As for reverts per minute, you can see some historical data by reviewing the history of User:EnterpriseyBot/defcon. At quick glance it doesn't seem to ever be close to zero, but I agree that lingering vandalism can often be attributed to periods when there were fewer humans patrolling. — MusikAnimaltalk01:48, 13 March 2021 (UTC)
I agree STiki was awesome and it should be a priority to get STiki or something like it back online. But even when STiki was active, I was regularly seeing obvious vandalism in my watchlist. We should bring back STiki but we should also try do even better than we were doing a year ago. Your point about Defcon levels always showing some reverts is a good one. If no volunteers are patrolling then we would still see reverts at ~40% of normal because of ClueBot NG and Xlinkbot, plus page watchers reverting edits that aren't blatant vandalism. Clayoquot (talk | contribs) 05:22, 13 March 2021 (UTC)
Image map caption pop-up title
At Confederate States of America, holding the mouse over the map in the infobox at the top shows a pop-up message which includes a lot of gobbledy-gook (see "text" below). Is this a known feature? Can it be readily fixed?
The infobox includes:
| image_map_caption = {{plainlist|style=padding-left: 0.6em; text-align: left;|
*{{Legend|#008000|The Confederate States in 1862}}
*{{Legend|#55c255|Claims made by the Confederacy}}
*{{Legend|#3d983d|[[West Virginia#Separation from Virginia|Separated]] West Virginia}}
*{{Legend|#00e6e6|Contested Native American territory}}
}}
This will need to be fixed in Module:InfoboxImage, which is apparently expanding the templatestyles tags instead of removing the associated strip markers when it inserts content into the title of the image. It may need to be adjusted in some other way also.
It's not the fault of Module:InfoboxImage here. It's that the title of the image is set when the image is not a thumbnail.
[[File:Confederate States of America (orthographic projection).svg|frameless|alt=Map of northern hemisphere with Confederate States of America highlighted|upright=1.15|{{plainlist|style=padding-left: 0.6em; text-align: left;|
*{{Legend|#008000|The Confederate States in 1862}}
*{{Legend|#55c255|Claims made by the Confederacy}}
*{{Legend|#3d983d|[[West Virginia#Separation from Virginia|Separated]] West Virginia}}
*{{Legend|#00e6e6|Contested Native American territory}}
}}]]
Hi. I made a number of twinkle personal options but unintentionally took Twinkle options off the screen. So are my options above the screen. What can I do to get Twinkle to reappear in my options? DrSalvus (talk) 20:15, 14 March 2021 (UTC)
Is it possible to pull a list of all articles created by a user when the xtools search limit is exceeded?
The xtools pages created tool doesn't work if the searched user has over 500,000 edits. Is there another tool or method that allows you to pull up a list of all articles created by a user with a large number of edits, or is there no practical way to do this? Hog FarmTalk19:31, 13 March 2021 (UTC)
If you go their contribs page, pull down "Search for contributions", choose Namespace = Article and check "Only show edits that are page creations", you can see up to 500 at a time there. Black Kite (talk)19:34, 13 March 2021 (UTC)
@Black Kite and Shushugah: - Thanks. The situation I'm working on is a mass-creation of tens of thousands of stubs, some of which are based on misinterpretations of sources, and me and a few other editors are trying to determine the extent. So this will be a slow process, but it doesn't look like there's an easier way. Is there a way to exclude tags in that method? Being able to filter out the ones with the tag New redirect would make things a bit easier, as the redirects are not problematic. Hog FarmTalk02:41, 14 March 2021 (UTC)
If you give me a username, I can try to query the database replicas directly to get a full list. — JJMC89 (T·C) 06:49, 14 March 2021 (UTC)
And another approach would be to use the API to download a list of all contributions by the user, then extract those with the "N" flag. If other attempts don't work, I could try that. Johnuniq (talk) 08:54, 14 March 2021 (UTC)
@Johnuniq and JJMC89: - It's Carlossuarez46. See this for some background. Essentially, we discovered that they mass created several thousands stubs for California "unincorporated communities" that are largely non-notable, errors, or original research. It has further been discovered they similarly created stuff like that for places in Iran. See this, some of their place creations appear to actually be stuff like cement plants; there was one deleted awhile back at AFD that was actually a gas station, not a community. We've also found a few in the US state of Virginia. The idea is to find out how far this mass-creation extends; I think well over 1,000 of their articles have been deleted. Hog FarmTalk22:57, 14 March 2021 (UTC)
Very cool, thanks (that saves my mucking around). Spoiler alert, the list shows 81,806 articles! Now some kind of tool is needed that takes the list as input then outputs a smaller list after filtering out pages meeting a magic criterion. For example, remove if edited by at least one non-bot editor other than the creator. Or perhaps filter on date of creation? Or number of bytes? Johnuniq (talk) 06:14, 15 March 2021 (UTC)
That roughly matches the number at Wikipedia:List of Wikipedians by article count. Is this list filtered for redirect creations? That said, I wouldn't consider edits by a non-bot separate editor to be a good metric for disqualifying problems. You would need to know how big the change was - a few characters, or a major expansion. BD2412T06:24, 15 March 2021 (UTC)
Current redirects are filtered out. You can't check if it was a redirect or not at creation using the replicas. I could filter out disambiguation pages too. That would bring it down to 72,238. The size in bytes (at creation or current) could also be taken into account. — JJMC89 (T·C) 07:48, 15 March 2021 (UTC)
I purged the page, and it rendered properly using 8 seconds of Lua time. Nevertheless, this is only a temporary fix, and a longer-term solution needs to be found. * Pppery *it has begun...19:19, 15 March 2021 (UTC)
Like move Administrative divisions to its own article (list of ....) perhaps. This is what to expect when trying to squeeze so much wikidata info in through scripting. — xaosfluxTalk19:42, 15 March 2021 (UTC)
That idea was suggested on the article TP four years ago but there was an "article is not too big" objection. I've boldly commented it out as a temp solution. MB22:19, 15 March 2021 (UTC)
Is there a way to get a count of edits per day for a certain article?
I'd be interested to know, for some data analysis I'm doing, how many edits certain Wikipedia pages get each day. I can get the average pretty easily, but the day-to-day fluctuations are hard to find without manually going through a page's history and counting. XTools only seems to go so far as to break it down by edits for each month. How can I get the edit count for each day for a given article? Thanks for your help. Ganesha811 (talk) 21:45, 14 March 2021 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
Wikis that are part of the desktop improvements project can now use a new search function. The desktop improvements and the new search will come to more wikis later. You can also test it early.
Editors who put up banners or change site-wide JavaScript code should use the client error graph to see that their changes has not caused problems. You can read more. [25]
The new version of MediaWiki will be on test wikis and MediaWiki.org from 16 March. It will be on non-Wikipedia wikis and some Wikipedias from 17 March. It will be on all wikis from 18 March (calendar).
Future changes
You can add a newline or carriage return character to a custom signature if you use a template. There is a proposal to not allow them in the future. This is because they can cause formatting problems. [26][27]
You will be able to read but not edit 12 wikis for a short period of time on 23 March at 06:00 (UTC). This could take 30 minutes but will probably be much faster.
You can use Quarry for SQL queries to the Wiki Replicas. Cross-database JOINS will no longer work from 23 March. There will be a new field to specify the database to connect to. If you think this affects you and you need help you can post on Phabricator or on Wikitech. PAWS and other ways to do SQL queries to the Wiki Replicas will be affected later. [28]
hi @Ponyo, thanks for the report and @Xaosflux thanks for putting it on Phabricator
As a possible workaround, can you try to add yourself to "pending changes reviewers" group to see if that helps you to workaround the issue? Martin Urbanec (talk) 02:09, 20 February 2021 (UTC)
(I know it can sound unrelated, but when debugging the previous occurance of the same issue, it sometimes worked with reviewer, but not without for some weird reason, so that's why I'm suggesting it) Martin Urbanec (talk) 02:10, 20 February 2021 (UTC)
In this this thread last month, several users (including myself) mentioned an error where our edits are not being auto-accepted on semi-protected articles. This techincal issue was eventually resolved. Now I'm having the same issue again, as seen here (same article as last time). Can this technical issue be fixed again? Thanks. Maestro2016 (talk) 02:13, 20 February 2021 (UTC)
Hello, I don't know where to put this, but my edits to pending changes pages have recently not been accepted automatically. This is weird because I've been able to make many edits to pending changes pages without any problem earlier this month. My account is almost a year old and I have over 500 edits, I've been able to edit extend-confirmed and semi-protected pages without any issues. Clear Looking Glass (talk) 05:47, 25 February 2021 (UTC)
Looks like the relevant phab ticket for this issue is phab:T233561. This is indeed still a problem, see [30], which should have been automatically accepted because it was submitted by an autoconfirmed user and there were no pending edits before it. Mz7 (talk) 20:07, 10 March 2021 (UTC)
There's this recent problem of mine that when I edit pending changes protected pages it says that my edits need to be reviewed first and not "automatically accepted". The first time that this happened to me was at Dire wolf, where I reverted an edit and didn't automatically accept it. Two edits of mine at Carnosauria [31][32] also weren't automatically accepted, and there were no previous edits pending review. I've been extended confirmed here for a long time already (almost 8,000 edits and 1 and a half years), so why do my edits at pending changes protected pages need to be reviewed? Can someone please explain what's happening and how can this be fixed? JurassicClassic767 (talk | contribs) 06:28, 28 February 2021 (UTC)
No, I know I'm not a pending changes reviewer, what I'm saying is that my edits at pending changes protected pages don't appear as "automatically reviewed". And what I meant was "autoconfirmed" (which is what you need for your edits to become "automatically accepted"), and I've been an autoconfirmed user a long time ago. JurassicClassic767 (talk | contribs) 13:48, 28 February 2021 (UTC)
@JurassicClassic767: Ah I see what you mean now, apologies. Edits from logged-in users to pending-changes protected pages are automatically accepted if and only if there are no prior edits pending review. If there are prior edits pending review, then the edits will go into the queue to be reviewed. ƒirefly ( t · c ) 13:57, 28 February 2021 (UTC)
But that's not what's happening though, in my examples above (Furileusauria and Carnosauria), there were no previous edits pending review, and when I edited, my own edit needed to be reviewed, that's the strange thing. JurassicClassic767 (talk | contribs) 14:12, 28 February 2021 (UTC)
Seems unlikely. FlaggedRevs (i.e. pending changes) is abandoned. Few/no devs want to touch the codebase. Although, it seems like there is a patch for review for that particular regression. ProcrastinatingReader (talk) 22:36, 1 March 2021 (UTC)
Ugh. Yet another part of MediaWiki that's mostly abandoned. Apologies, I didn't know FlaggedRevs had fallen down that particular hole. ƒirefly ( t · c ) 09:35, 2 March 2021 (UTC)
ProcrastinatingReader, that is quite unfortunate. If we cannot ensure the reliability of this software, then perhaps we need to start to have the conversation about whether we should continue to have this software at all. (And I say that as a long-time advocate of pending changes.) Mz7 (talk) 19:44, 12 March 2021 (UTC)
@Mz7: it's funny, I wrote up a small sub-section twice to suggest exactly that but refrained from hitting "submit changes" both times. I really like the idea of PC (and have suggested expansion), but imo the current implementation and UI of FlaggedRevs is inadequate, and it's so buggy and unmaintained that I'm not sure it has much of a future. I'm just cautious that if it's removed we perhaps won't be able to get momentum to have a suitable replacement developed, but it doesn't seem like they're going to fix this one up anyway. But really imo what's needed is start from scratch, from basics with the problem and then design & engineer a new solution. ProcrastinatingReader (talk) 19:50, 12 March 2021 (UTC)
A "review before commit" protection method is valuable for where it is successful (low-edited pages esp. BLPs, apparently TFA sooner rather than later, and gosh-it-would-be-great-if-we-could-do-that-with-templates-and-scripts-except-PC-doesn't-work-there), so starting from scratch doesn't particularly make sense. (What also doesn't make sense is that the WMF built the German version which PCs everything, then started there to make the English version after feedback here, which is part of why the code is a mess.) --Izno (talk) 20:00, 12 March 2021 (UTC)
Funky idea: create a bot that automatically accepts edits made by autoconfirmed users where there are no pending edits before them. Would be a stopgap measure to restore the status quo while we find a developer willing to spend time to deal with the codebase. Mz7 (talk) 20:35, 12 March 2021 (UTC)
There's also the issue of when a page is both semi-protected and pending changes protected, and yet some edits are still not accepted. Ex. [34]... This will need some further bug fixing... Cheers, RandomCanadian (talk / contribs) 00:02, 2 March 2021 (UTC)
I did notice that my edits were not automatically accepted and instead required approval. Consider me clueless. — Paper LuigiT • C01:36, 2 March 2021 (UTC)
Seems like any page that has a "Pending changes" lock (including pages that are both PC and semi-protected) has seemingly stopped automatically accepting the edits of established users. I hope the issues are fixed soon. Clear Looking Glass (talk) 00:11, 16 March 2021 (UTC)
Apparently that this problem is not present at dewiki, per this. So it might be an issue of both configuration as it could be one to do with pendings changes (likely both). RandomCanadian (talk / contribs) 05:05, 8 March 2021 (UTC)
For what it's worth, I've just filed a BRFA for a bot that would effectively 'patch' this bug until it can be fixed properly. ƒirefly ( t · c ) 17:36, 15 March 2021 (UTC)
My first suspect for "invisible" input is usually Wikidata, and the page does, in fact, have a Wikidata entry. Unfortunately, the WD interface is entirely opaque to me and I can't see or say what needs to be changed there. I do see that that map is visible (under "coordinate location") and is shown as imported from Russian Wikipedia, which would explain komical not wery good spellink of English city.
This is not Wikidata's fault. The "imported from" only indicates the source of the coordinates, it does not have any impact on the display. --Izno (talk) 21:02, 14 March 2021 (UTC)
AIUI the labels on maps are taken from OpenStreetMap Data, and a bit of searching there finds that it is listed as the Latin-spelling for the name of the city in Serbian.[35] there is no obvious reason why that would be the source though, as the infobox map at Agrarian Bank Building shows the city name as "Belgrade" rather than the sr-Latn "Beograd". Google also tells me that it is the city's name in Veps (a minority language spoken around the Finland-Russia border region), but that's not in the OSM data so seems an even more unlikely source. Neither language edition of Wikipedia has an entry for Heritage Plaza fwiw. Thryduulf (talk) 22:32, 15 March 2021 (UTC)
For what it's worth, it seems that the "default"(?) text for the city has been changed somewhere that I can't seem to find - see this OSM fragment, which is what is seen in the article (has lang=en), along with the French (same thing, no specific name defined). However, the Russian name displays correctly; most bizarrely if you don't specify a language at all it displays the English spelling correctly! Perhaps someone more familiar with OSM can use this to track down what's going on? ƒirefly ( t · c ) 11:31, 17 March 2021 (UTC)
Should I have been able to create these lowercase Cyrillic redirects?
I was reading an article that linked to the Cyrillic letters ԛ and ԝ. I was surprised to see one-character redlinks, so I went ahead and created them, pointing to the appropriate Cyrillic letter articles. I then took a look at other one-character redlinks (at User:Cbuckley's helpful User:Cbuckley/Unicode), and the first one I went to create, Ƀ, brought me to this RFD from 2013, which concluded that the ƀ title wasn't supposed to be creatable at all (since it's lowercase, and titles aren't case-insensitive on the first letter), resulting in a VPT thread. Since then, something seems to have changed, perhaps related to this action in 2020 by the maintenance script: Uppercasing title for Unicode upgrade. After reading this, it occurred to me that the two I created are also defined by Unicode as lowercase, and perhaps something was amiss here.
So:
On a technical level, was I supposed to be able to create these? If so, why, given that they're lowercase?
Assuming I was supposed to be able to, is there any tech- or policy-based reason I shouldn't have?
(If the answer to either of these is "no," please consider this a preemptive consent to G7 deletion.)
This is a historical oddity that HHVM used to treat those letters as uppercase, but modern PHP/Unicode that we now use correctly has them marked as lowercase. It looks like the migration never finished per phab:T219279...I'll bump the task. I would recommend deleting both because they're inevitably going to be moved to the uppercase letter. Legoktm (talk) 21:20, 17 March 2021 (UTC)
When editing a talk page from a mobile phone, there is not a link to add a new section, or a link to edit an existing section. The edit page only has section 0 (not all sections as one would expect).
So an existing thread cannot be commented, and a new thread will be above older threads, instead of below, as usual. --62.98.99.49 (talk) 03:49, 18 March 2021 (UTC)
Indeed, I see that too, but only when logged out. For example, logged in, viewing https://en.m.wikipedia.org/wiki/Talk:Foo there's a big blue "Add discussion" button at the top, and a "reply" box in existing threads, but both disappear when I view the same page in a private tab. You can still edit logged out by clicking "Read as wiki page" at the bottom. I don't know why the other buttons are hidden from logged-out editors; IIRC it wasn't always this way. Given that the tmboxes (FAQs, BLP warning, etc.) are hidden from mobile editors, I'd almost consider this a feature. :-) Suffusion of Yellow (talk) 04:03, 18 March 2021 (UTC)
Infobox canal
I'm wondering why Deokali Canal is displaying "Lua error in Module:Location_map at line 422: No value was provided for longitude." I recently ranted about Wikidata here and am wondering if this error is due to a bot creating d:Q105759593 (with no coordinates) on 5 March 2021 which triggered {{Infobox canal}} to get the coordinates from Wikidata. Does anyone know how to tell the infobox not to attempt to display a map? Or is there some way of fixing the problem other than working out the required coordinates which might be tricky? Johnuniq (talk) 06:49, 18 March 2021 (UTC)
Removing the pushpin_map parameter removes the location map. The Infobox canal template does not attempt to fetch an coordinate from wikidata, because it does not use the useWikidata parameter, which is required by Module:Location map to fetch that from wikidata. The module is only complaining about the lack of local coordinates.--Snaevar (talk) 08:21, 18 March 2021 (UTC)
OK, thanks. I removed that and the error is gone. I jumped to a conclusion because editing the article shows "Wikidata entities used in this page Deokali Canal: Title, Sitelink, Description: en". Something is calling Q105759593. Johnuniq (talk) 08:56, 18 March 2021 (UTC)
Is there a way for templates to know they've been called via a redirect?
When a proposed deletion has been endorsed, the template says "If you remove the {{proposed deletion/dated}} tag above, please also remove this {{Proposed deletion endorsed}} tag.", however at e.g. Meon Valley Passing loop the endorsement was done using the redirect {{endorse prod}}, which might be confusing for a new user. Proposing to change that to show the actual code used (and working out the best place to propose it) is only worth it if it is technically possible and I don't know where to even start looking to find out (other than here). Thryduulf (talk) 22:13, 15 March 2021 (UTC)
Could you make the redirecting template a wrapper instead? Like, add a parameter of please also remove this {{tlc|{{{templatename|Proposed deletion endorsed}}}}} tag, with the redirect turned into {{Proposed deletion endorsed|comment={{{comment|}}}|templatename=Endorse prod}}. -- Tamzin (they/she) | o toki tawa mi.21:23, 17 March 2021 (UTC)
In Lua there is an way to do this. It is mw.title.makeTitle('Template', frame:getParent():getTitle()).isRedirect. It returns true if it transcludes an redirect, otherwise its false. It is an expensive function, so you would need to limit its use to under 500, assuming no other expensive functions are on the page. This could be added to Module:Redirect, for example, it does similar things.--Snaevar (talk) 09:22, 18 March 2021 (UTC)
Looking at where the stricken text started, there was an unclosed <s>. I have closed it.[37] Our software has changed. Previously it automatically closed more tags at certain points, e.g. headings. I think it was Tidy. See mw:Parsing/Replacing Tidy. PrimeHunter (talk) 14:19, 12 March 2021 (UTC)
I see. So no one fixed it because it did not affect the rendering on the page. I've seen other archives where this happens as well. --Heymid (contribs) 14:31, 12 March 2021 (UTC)
This error was an example of a Lint error, in this case a "Missing end tag" error, of which there are more than eight million across all namespaces. "Missing end tag" is by far the most common actual error (obsolete HTML tags are more common, but they are still fully supported by all modern browsers, so they tend not to cause problems yet). See Wikipedia:Linter for more information.
Many of the errors are still corrected automatically by the MediaWiki software rendering engine, but some of them cause display problems, and there is no guarantee that the MediaWiki software will continue to render them correctly forever.
Gnomes are welcome to join the very small group of editors who are working on fixing these errors; article space is substantially cleaned up, with only about 200,000 errors (plus obsolete tags) remaining, down from about 700,000 two and a half years ago. Template space, with the exception of DYK pages, which are in the wrong namespace, is essentially free of Lint errors. – Jonesey95 (talk) 23:38, 12 March 2021 (UTC)
Right, the IP fixed it but didn't make an edit summary so it looked like random removal of valid code. That happens a lot and many users would have reverted it. PrimeHunter (talk) 10:47, 18 March 2021 (UTC)
Grant proposal to build a Visual Editor for Citoid Web Translators
Hi everyone! With Diegodlh we are presenting a grant proposal to build a visual editor for Citoid web translators. The goal of this visual editor is to make it easier for non-technical users to create and edit Citoid web translators to increase website coverage of this citation metadata retrieval service. The grant proposal is open for feedback and endorsement, and we are also seeking for collaborators and volunteers, so if you are interested, please visit the grant proposal page! --Scann (talk) 12:25, 18 March 2021 (UTC)
Actually, I figured it out. The problem was that I forgot to include the page name when I added {{subst:mfd-inline}}, which caused it to default to the name of the current page and include a link to that page. That link didn't include the leading colon, so it was interpreted as a categorization. * Pppery *it has begun...00:41, 19 March 2021 (UTC)
Infobox scientist doesn't display a portrait from Wikidata
BTW, do I correctly recall that some infoboxes do not display locally defined captions when they display Wikidata-defined images? --CiaPan (talk) 23:20, 18 March 2021 (UTC)
Infobox scientist isn't Wikidata-fed, just like most infoboxes aren't Wikidata-fed. Images are something each wiki can decide for themselves, and someone replacing the main image on Wikidata shouldn't influence what is shown here (e.g. there are many articles with an image in the infobox and another one or two further down: if someone at Wikidata or from another language decides that that second image should be the default, then we would suddenly show that image twice in the same article, and the other image not at all). Fram (talk) 13:35, 19 March 2021 (UTC)
Requesting assistance in clubbing of tables
Greetings
As part of restructuring of the article Black Sea I am looking for help in clubbing following two together to make them compact.
Currently both tables are arranged in the sequence of highest length/area to lowest.
The technical help I am looking for is after clubbing of the tables, user should be able to interactively click on column heading length or area and be able to choose to view as per his preference. I have seen such interactive tables in some articles but I do not know how to create them. I would be thankful if some one helps out.
This is pretty easy, but first let me ask: what do you want to call the combined table? How are you relating coastline length and economic zones? What's the connection (besides belonging to these 6 countries)? — JohnFromPinckney (talk)05:08, 19 March 2021 (UTC)
@JohnFromPinckney: Simply I prefer to call it 'Coastline and exclusive economic zone'. Maritime exclusive zones are basically area of economic rights from the coastline, 200 nautical miles (nmi) from the coast of the state in question. and hence are connected.
Unless community at talk page insists too much I do not find point in keeping them separate since it unnecessarily occupies additional space in the article. After combining I will either present it in Population section with some paragraph writing or in economics section. For economics section a draft page is already opened.
I am not too keen arguing to my point is right, it is just an attempt to make page a little more neat. How other users want it can be discussed at article talk page where a discussion for that purpose is already open. But actual discussion is likely to take too long a time though I invited some users, by checking 30 largest contributing editors only six seem to be active and they too are on the way copy edit fellows so to join discussion they seem largely disinterested.
So I prefer going ahead improving article. If you are interested in Black Sea issues you are most welcome to put forward your views too.
Okay, Bookku, I just wanted to make sure you thought there was a logical connection for the data, so the table makes sense to the reader. Should I just whack the tables together for you, so you can see what I did (maybe questions at my Talk), or were you hoping for something like a tutorial? The former would take me about 5 minutes, the latter... um, longer (more like a half-hour). — JohnFromPinckney (talk)06:45, 19 March 2021 (UTC)
@JohnFromPinckney: As of now I prefer putting the table together and may be you can give general tutorial link of some one else too will do which I can go into at leisure. Thanks it's very nice of you to have your support warm regards. Bookku (talk) 07:01, 19 March 2021 (UTC)
@Bookku: can you clarify which way you want to go? It doesn't matter to me (I'm glad to help either way), but if I don't know which you want (I merge the tables or you merge them), then I won't do anything before tomorrow. — JohnFromPinckney (talk)08:26, 19 March 2021 (UTC)
Coastline and exclusive economic zone
@JohnFromPinckney: I have merged the data below till there I know, you please add that interactive arrow thing. I am not sure I will be able to focus on tutorial as of now so if you can do it for me will be better IMHO. Thanks and warm regards Bookku (talk) 08:39, 19 March 2021 (UTC)
You didn't really need to post your intermediate results here; we already had the link to Black Sea and in fact, I have exceptionally taken the liberty of deleting that part of your previous post, so that it ends with your sig. I hope you don't mind. In fact, I hope nobody minds.
But my efforts are now there under "Exclusive economic zone", a title which might need adjustment. I deleted the other section about coastline length (including its table), reworked the merged table for accessibility (added scope attributes, removed unnecessary wikicode and the entire "Number" column, and added a caption close to what you specified), added the class "sortable" to allow the sorting you wanted, right-aligned the table by default so the number columns look good, made the Totals row stay at the bottom when sorting, and moved the ref citations to the column headings to show what they're supporting. I fleshed out the cite for EEZs (Sea Around Us) and reused the named ref for coastline length at bsnn.org (which is still a bare URL on that page and should be expanded).
Any Q&A can happen on my Talk; it doesn't have to distract people here further. Good luck with the page (I see there's another nasty table farther down...)! — JohnFromPinckney (talk)09:42, 19 March 2021 (UTC)
NO! Absolutely under no circumstances should Bookku or you or I or anyone on the planet make a separate, single-column "table" of numbers from 1 to whatever. That's one of the most inaccessible ideas I have seen on all of Wikipedia, not counting maps and colored graphics.
In the above case the table is only six rows long, so a numbering is unnecessary, even if it always indicated something true. But adding a separate table in the hopes that most visually able people will actually discern the intended alignment on any given device in most circumstances is worse than silly. Don't do that, and don't recommend that. To anyone. Ever. — JohnFromPinckney (talk)11:00, 19 March 2021 (UTC)
"clubbing following two together to make them compact" meant to combine them into a single table. The idea was clearly to have the same number of rows but enough columns to display all the original information. It was independent of the request to make them sortable. PrimeHunter (talk) 15:19, 19 March 2021 (UTC)
"9. To unite, or contribute, for the accomplishment of a common end."
"clubbing following two together to make them compact" clearly means something like join/combine/unite. The context with the tables made it clear enough to me (and JohnFromPinckney). PrimeHunter (talk) 16:15, 19 March 2021 (UTC)
@PrimeHunter: I am really not good at grammar, so thanks for helping out.
@JohnFromPinckney: Thanks for all the help in clubbing @ Black Sea :). COVID is hampering clubbing mood then may be you want to have a virtual look at recent South Asian light-hearted neologism 'pawri' for 'party', it comes from unexpected corner of the world that is Peshawar still more difficult than COVID terrain to party, but You can join virtually. Bookku (talk) 06:21, 20 March 2021 (UTC)
How to apply CSS to a template?
I'm trying to make a graphical element in a template go away by applying visibility:hidden to it. This works if I open Element Inspector in Chrome and manually apply the style to the offending element, but I can't figure out how to do this in wiki markup. I know you can do <span style="...">, but that doesn't seem to work with CSS selectors so you can apply it to just the one element. See [38] for more details. — Preceding unsigned comment added by RoySmith (talk • contribs) 15:25, 20 March 2021 (UTC)
Recently, there have been a RIDICULOUSLY HUGE number of page moves resulting from a user rename by Superpes15. The count of user and user talk pages (including subpages) as of when the "Submit Query" button was last pressed can be found at quarry:query/47167. I will keep pressing that button until the first two counts reduce all the way down to zero or one.
@GeoffreyT2000: Wow there are so many subpages! Since the user was blocked after the rename I will have to revert the rename when it's completed. If you want I can not move the pages when I proceed to revert the rename or I don't know if you can delete all these pages! as far as I know unfortunately these actions always remain in the logs :( --Superpes15 (talk) 17:35, 20 March 2021 (UTC)
As I understand it, the block was for WP:SOCK, and the rename was a WP:RTV which was requested before the sock was found out. If the user subpages that were moved were created by the sockpuppet account (not the sockmaster), there should not be a need to move them back since they are deletable under WP:CSD#G5. --Redrose64 🌹 (talk) 18:08, 20 March 2021 (UTC)
linking to a Wikidata entry within a WP article? or not?
I've been doing a little copyediting in the WP-en article for Null Island. The author has created an internal link to Achowa Point, which leads me to believe the author considers that important. But there's no article for Achowa Point -- although there is a Wikidata entry which is maybe? a little useful. So my question: is there a template to link from a WP article to a Wikidata entry? Redwidgeon (talk) 18:25, 20 March 2021 (UTC)
There's not just a wikidata entry, but from there I see that two other languages' wikipedia sites have articles. Consider {{ill}}? DMacks (talk) 04:46, 21 March 2021 (UTC)
Calling it with {{x33||two}} returns "two", as expected. However, {{x33|one}}should return "one", but instead produces the error "Template loop detected".
A template can call itself once on its own template page. That means a template can show documentation with calls of itself. If another page calls a template then the called template cannot call itself. PrimeHunter (talk) 00:15, 22 March 2021 (UTC)
I have a question regarding Wikiversity. This question I'm not sure where it's best answered at, as this question is not about English Wikiversity, but about non-existent Wikiversities. The link to the image above was an image I uploaded a while ago. Before that, when I was viewing pages on Wikiwand, I observed a pattern. An existing page shows the page. However, when a non-existing page, or at the very least, a page that cannot be properly shown, gives a 404 instead of MediaWiki:Noarticletext. However, MediaWiki:Noarticletext does show the no article text. https://wikiwand.com/en/v%3aaa%3a (v for Wikiversity and aa for Afar = Afar Wikiversity which does not exist) gives a 404, which likely means that the error message was not manually created at aa.wikiversity.org. Given that there are Wikimedia projects in a total of 300+ languages and only 17 Wikiversities, it doesn't seem good to create the error message at ~300 subdomains. However, https://tsz.wikiversity.org or any other *.wikiversity.org subdomain with a currently unused language code on Wikimedia gives nothing at all. So how does the software know what subdomains to show this error message at, and where is the source code for the error message? I have been wanting the source code for some time so I can test around with the message in my own sandbox. 54nd60x (talk) 11:53, 22 March 2021 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
There is a Wikipedia app for KaiOS phones. They don't have a touch screen so readers navigate with the phone keys. There is now a simulator so you can see what it looks like.
You will be able to read but not edit twelve wikis for a short period of time on 23 March at 06:00 (UTC). This can also affect password changes, logging in to new wikis, global renames and changing or confirming emails. This could take 30 minutes but will probably be much faster.
Changes later this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from 23 March. It will be on non-Wikipedia wikis and some Wikipedias from 24 March. It will be on all wikis from 25 March (calendar).
Flagged revisions will no longer have multiple tags like "tone" or "depth". It will also only have one tier. This was changed because very few wikis used these features and they make the tool difficult to maintain. [41][42]
Gadgets and user scripts can access variables about the current page in JavaScript. In 2015 this was moved from wg* to mw.config. wg* will soon no longer work. [43]
Request signature button in new talk page interface
Hi all
Today there is a really nice new fancy interface for adding new subjects on talk pages (I'm using it right now). I'm not sure what its called or how to find the project on Phabricator, but I'd like to request that a signature button is added to the row of buttons, its such a common thing and feels very missing to me. Does anyone know what its called and where I could file a request on Phabricator?
Oh, it seems to add them automatically? This is very nice, would it be possible to have a little pop up the first time explaining this or something? John Cummings (talk) 16:14, 17 March 2021 (UTC)
John, you were in the visual mode. (It tries to guess, on your first edit, which one you prefer, based on how you edit articles.) If you're in the wikitext source mode, then you'll see a live preview that includes a preview of your signature. If you type four tildes (Ghost is correct: it doesn't know about the three- or five-tilde options), it'll show your signature and give you a message that you don't have to add it. Whatamidoing (WMF) (talk) 18:19, 18 March 2021 (UTC)
I think this is just a new-old user adjustment that will happen, not cause to change or add to the user interface. Izno (talk) 18:39, 18 March 2021 (UTC)
Indeed, the entire point is to not ever have to worry about signatures ever again, so no signature buttons or more explanatory text, please! :) I suspect experienced users will figure out the auto-signing feature on the first try, as was the case with John Cummings. The real advantage is of course for newbies. I personally can't wait until this is a default-on feature! — MusikAnimaltalk03:07, 19 March 2021 (UTC)
The bigger issue is that sometimes indentation doesn't quite work as wanted; and it has difficulties when there are some other things in the way; for example {{ctop}} (and the corresponding cbot) sometimes mess with it; see for ex. this one which didn't quite work out... RandomCanadian (talk / contribs) 04:00, 19 March 2021 (UTC)
But feedback/suggestions where to document is always welcome. For example, I'd love to dynamically edit the messages, with more context of what countries/projects I added in talk page/how this was created etc..
Problem
I am in the process of creating templates such as Template:Diplomatic missions in the United States by copying code I generated with Ruby language in my local machine, and pasting it into a structured URL my script editor opens automatically. After looking at the preview, I verify manually if there are any links inside Category:Diplomatic missions in the United States and even fix them for consistency.
Possible solution
This seemed like a task for preloading, and I am currently attempting that, but having difficult preloading both static and dynamic categories. Enclosed are my templates and the url I'd call for say "Venus". I have already included the Categories in the referenced preload template, but they refuse to show up, which you can find here User:Shushugah/Diplomatic Missions In X
Tl;Dr
Tl;dr I need three categories to be added in within each Preload session, the three categories are (with $1 interpolated):
The ruby code I use locally...to open the URL's (not needed) e.g. call this instead:
Code demo
# wiki.rb file can be called with #=> ruby wiki.rb Peru# or ruby wiki.rb Peru t #=> opens talk pagedeftemplate_link(country)"https://en.wikipedia.org/wiki/Template:Diplomatic_missions_in_#{country.gsub('_','%20')}?veaction=editsource"enddeftemplate_talk_link(country)"https://en.wikipedia.org/wiki/Template_talk:Diplomatic_missions_in_#{country.gsub('_',' ')}?action=edit"enddefpreload(country)country=country.gsub('_','%20')"https://en.wikipedia.org/wiki/Template:Diplomatic_missions_in_#{country}?veaction=editsource&redlink=1&preload=User%3AShushugah%2FDiplomatic_Missions_In_X&preloadparams[]=#{country}"endcountry=ARGV.reject{|v|v=='t'}.join('_')print"{{WikiProject International Relations}}\n{{WikiProject #{country.gsub('_',' ')}}}\n"ifARGV.include?"t"system("open",template_talk_link(country))ifARGV.include?"t"system("open",preload(country))unlessARGV.include?"t"
Agreed. What's the best practice for loading categories, without calling them within preload template itself? At best I can imagine making them hyperlinkable like Category:Example and remove the colons in the link right before publishing new template Shushugah (talk) 18:44, 21 March 2021 (UTC)
@PrimeHunter Thank you PrimeHunter! This allows me to avoid the fuss of encoding/decoding the preloadParams!
In general, I am interesting in using my software development skills more, to edit/improve templates on Wikipedia. Do you have any recommended tutorials/links to start with? For most purposes, categories work fine, but for many to many lateral relations like this, across multiple languages. I find it useful to have more structured templates. Shushugah (talk) 20:08, 21 March 2021 (UTC)
@PrimeHunter I'm having difficulty...with nested include/noincludes again.
Right now if Template:Diplomatic missions in Chad will leak its own categories, onto whatever page it's transcluded on, because the categories are missing a wrapper tag. However, I am having difficulty...transcluding that within the preloader. User:Shushugah/Diplomatic_Missions_In_X. I was thinking of substituting them, but I am having hard time understanding how to write {{tl|noinclude}} and accept parameters at same time. Shushugah (talk) 21:05, 22 March 2021 (UTC)
Overview of capabilities and limitations for the mobile editing experience
Simple question really:
Is there a place where every difference to the regular desktop editing environment is listed? Just to take one random example, you cannot Thank using the mobile interface.
I'm talking about articles directed towards the general readership (i.e. wiki pages), not links into the code itself.
I have looked at both those articles (assuming you didn't mean the actual talk page there) but none go into any kind of detail. They seem to be the place where you'd find a detailed summary, but nope. CapnZapp (talk) 13:40, 22 March 2021 (UTC)
Sans some Phabricator/MediaWiki links (this is after all not exclusively a Wikipedia issue) the above two places would be good place to add Known limitations or something to that effect. The increased attention on mobile can only be a positive thing for the global movement Shushugah (talk) 13:44, 22 March 2021 (UTC)
@CapnZapp: I don't think there is a single page but if there is it isn't likely to be specific to the English Wikipedia. Also there are multiple "mobile" interfaces (Mobile Frontend, Mobile Apps). You can try following up at mw:Talk:Reading/Web/Mobile. — xaosfluxTalk13:47, 22 March 2021 (UTC)
While I'm slightly surprised we aren't telling our readers what to expect - other than a general and vague "don't expect everything to work on mobile" - and think that what might have been acceptable ten years ago when mobile editing was a curiosity, no longer is when mobile editing is the standard... (far too many of the discussed pages still come off as talking to the "adventurous few" rather than presenting what billions of people today simply assume is the default)... I'm personally good with an answer of "no, there isn't".— Preceding unsigned comment added by CapnZapp (talk • contribs) 16:47, 22 March 2021 (UTC)
The mobile interface is limited compared to the desktop one. So, I suppose on the list of missing items for readers about Wikipedia are search, media support (I recently added info on this just now on Help:Mobile access), navigation, Special:Nearby and maybe wikimedia commons.--Snaevar (talk) 17:21, 22 March 2021 (UTC)
To complicate further, there are 4 skins of mobile. The default does include WP:Nearby but skips other features. And SOY's table is specifically about messaging users so I wouldn't call it "not even close to complete", just lacks information of other features because it's not intended to be a general comparison table. Slywriter (talk) 01:26, 23 March 2021 (UTC)
Of course, it isn't really complicated. I mean, if we really prioritized the mobile experience we would make this easy. Instead of today's situation, where it's apparently deemed a good UI to feature several skins, each with its own set of limitations, and essentially no non-technical documentation. But I'm a desktop user, so I don't care. CapnZapp (talk) 14:30, 23 March 2021 (UTC)
At WP:Good article nominations, the counter of unreviewed nominations at the top of the page is currently about five higher than the actual number. As discussed here, the root of the issue appears to be that Category:Good article nominees awaiting review lists (currently) 327 pages, but claims to have 332—in other words, something appears to be making the category add five pages to its count. Does anyone have any idea what might be causing this? Thanks, --Usernameunique (talk) 04:31, 22 March 2021 (UTC)
The count at WP:Good article nominations is made with {{PAGESINCATEGORY:Good article nominees awaiting review|pages}} which produces 525. {{PAGESINCATEGORY}} gives the number claimed on the category page and not the actual number if there is a difference. There is no way in wikitext to circumvent the bug and get the actual number. We have customized MediaWiki:Category-article-count to say "approximately" for numbers above 200. PrimeHunter (talk) 09:12, 22 March 2021 (UTC)
The odd thing is that the count seems to consistently be five over the actual number of pages in that category. Every time I've checked over the past several days, the difference is exactly five. In fact, while I was writing this post and rechecking, the count went from 327 to 328 and the actual number of category members went from 322 to 323. I've run into a similar thing in at least one other category as well. If it's going to be a consistent difference, perhaps we could adjust template {{GAN counter}}, to return the actual number by subtracting that five overage produced by {{PAGESINCATEGORY:Good article nominees awaiting review|pages}}. BlueMoonset (talk) 02:02, 23 March 2021 (UTC)
It's unlikely to remain 5. The count is updated by adding or subtracting 1 when a page is added or removed. The update occasionally fails to happen. Sometimes there is a full recount. PrimeHunter (talk) 10:03, 23 March 2021 (UTC)
I'm seeing 318 count and 313 actual pages in the category when I click over to the second display page for the category which has 113 rather than the expected 118. (The first display page has 200.) But it sounds like there's nothing to be done in terms of fixing the count, so I think I'll monitor the situation for a while to see whether this behavior continues. Thanks for the information. BlueMoonset (talk) 00:37, 24 March 2021 (UTC)
Does it matter if http URLs that are unlikely to be the target of computer security attacks are used as references instead of the https equivalent?
You asked if it matters and Izno said yes, it matters. Don't change https to http if https works. Even if the target size is completely safe and harmless, https gives more privacy and security against eavesdroppers on the connection. Wikipedia:External links#Specifying protocols says https is preferred. Saving one byte in the wikitext would be a really dumb reason for changing to http. PrimeHunter (talk) 19:26, 23 March 2021 (UTC)
The size of the link's expansion into HTML is many times larger weight than the character you add at the end of the day in the wikitext, never mind the page size associated with images. Do not worry about it. Text is cheap. Izno (talk) 00:41, 24 March 2021 (UTC)
When I make a citation from a YouTube video using the VE citation tool e.g from a news organisation's channel its not providing very much information, only the name of the video and a link, and doesn't provide additional information which available e.g channel its from, publication date etc. One other idea (although I'm not sure how it could be done would be allowing users to easily add a start and end time for the citation, which would be very helpful in citing information from longer videos eg documentaries. Is there a place on phabricator etc I could make those requests?
You can request a change to the information returned in the Citoid project. You will probably not get a start/end time, but channel and date are very possible. Izno (talk) 00:37, 24 March 2021 (UTC)
The development process for Visual Editor did not pay much attention to the need for references, and as referencing ability has been added in it can be confusing. Using Cite is not the only way. Any template which has Wikipedia:TemplateData can be added with Insert Template. Most of the citation templates do and Template:Cite AV media might give you what you need. StarryGrandma (talk) 00:53, 24 March 2021 (UTC)
Citoid uses Zotero for citation lookup and in Zotero there are translators for some websites. One of the websites with translators is youtube, it is configured to get title, url, duration, date, author and description. This is mostly an upstream issue, Zotero does for the most part get the same result as Citoid, but Citoid for some reason skips the website name. For duration and description there is a double issue, because even if Zotero would pick that up, the TemplateData Citoid mapping (in Template:Citation) is not configured for these fields, i.e. it would not know in what parameter that should go.
Google and Microsoft translators and Wikipedia tools
Hello, I'm from Central Kurdish Wikipedia. Google (Not official yet) and Microsoft have recently added our language to their translators. It's good news for us, but the problem is that some of our users use them without review, and they've created a lot of problems for us. Here, I want to know if there are any tools, gadgets, user scripts, filters or anything else to reveal how much content of an article was taken from those automatic translators? or we appreciate any other idea that makes our work easier. Thanks! ⇒ AramTalk19:08, 18 March 2021 (UTC)
Maybe a warning in the edit window (similar to the one for not copying from random internet sources) could help catch some of them62.98.99.49 (talk) 13:26, 19 March 2021 (UTC)
Aram, I don't think there is such a tool; I've never seen one mentioned in posts about awkward machine translations. If there were one, it could be helpful here as well as at the Central Kurdish Wikipedia. Perhaps some of the Google Summer of Code people would be interested in the idea. BlackcurrantTea (talk) 01:00, 23 March 2021 (UTC)
What does it mean when a table column font has an unwanted enbiggenment not commanded by wikitext?
The rightmost table column has text the size of the article, the other table text is smaller (but still easily readable in landscape on my c. half megapixel screen, and given the great length of this table the small font is probably preferable since this shortish note alone embiggens this row's height by almost 3 times (the average photo height would otherwise set the typical row height and it is only ~1/eth this note)Before touching any linksThe links I touched shrunk
This also happens to links sometimes and if you touch one it will shrink to normal size and the entire page will suddenly rearrange to what it would look like if the link you touched was the right form factor in the first place. This will often result in clicking an unwanted link which has now squirmed to under your finger. Sagittarian Milky Way (talk) 01:10, 23 March 2021 (UTC)
For example Help:Table#Using_the_toolbar has right column text about twice the size of the other two but the producing wikicode is identical and identically sized. On Chrome browser, Android smartphone, MonoBook skin, wikipedia not m.wikipedia, desktop site checkbox on Chrome. It doesn't happen all the time though. An example for links is List of Mars-crossing minor planets, the sea of links are the same form factor as the lead text but when I touch or long press the links they sometimes shrink and rearrange the page to what it would've been if the ones I had shrunk and only those had never been big. They do not stay shrunk when clicking a link and immediately navigating back. One of the times I viewed the asteroid page today some of the links were already small before I touched anything, a column or two of a section but not the other column(s). Sagittarian Milky Way (talk) 15:47, 23 March 2021 (UTC)
I don't see any size difference or page change in Google Chrome 89.0.4389.90 on Windows 10 or iOS 14.4 on an iPhone. I don't have an Android device. PrimeHunter (talk) 19:14, 23 March 2021 (UTC)
Please provide the version numbers of your device and browser software. While I agree it is probably the browser's fault, it may also be related to the age of your device (I vaguely recall similar behavior a long long time ago). Izno (talk) 00:40, 24 March 2021 (UTC)
Chrome 89.0.4389.105. I don't know when my smartphone was designed but I bought it new less than a year ago and the battery still works great after thousands of hours of screen on which is a sign of youth, it's just very low-end (at least by recent name-brand standards). When links are in certain configurations like in prose or section hatnotes they seem to shrink never or maybe rarely while most loadings of pages with lists of links have shrinkage and the links after the touched one always move to fill the extra space, they never just stay where they were. Sagittarian Milky Way (talk) 02:41, 24 March 2021 (UTC)
This is built in functionality of mobile devices to make text more readable when you are using desktop websites. It is called "text size inflation" where it makes the font size of larger blocks of text as well as headers and links bigger based on the idea that it is more likely you will want to read them or touch them (links). Visited links have already been read, so it is unlikely you need to go there again and they are smaller. You shouldn't have this problem if you use the mobile site of Wikipedia. —TheDJ (talk • contribs) 10:37, 24 March 2021 (UTC)
I wish already visited links stayed shrunk for more than 0 seconds, if it was 30 minutes I wouldn't care. But it's good to learn that if I make a long article even a half megapixel desktop monitor would probably cause less scrolling. Sagittarian Milky Way (talk) 13:50, 24 March 2021 (UTC)
Help me coding for collapsible CS1 error & maintenance messages
I want to make CS1 error messages collapsible to make them less annoying. On this case, I want to show a small clickable icon (i.e. exclamation mark) on a citation to show / hide an error & maintenance messages by modifying the code from Help:CS1 errors#Controlling error message display. How should I code it?
It is worthtless. If I am looking for problems to correct them I want to see them all. Otherwise I want to hide them all. So global option in private CSS is enough for me. However, there is build-in mechanism to make collapsible elements in mediawiki described in mw:Manual:Collapsible elements. So if you want this feature optional you have to write a gadget. The manual page consists of some source code, which might be adjusted for your needs. Paweł Ziemian (talk) 08:54, 24 March 2021 (UTC)
After a recent expansion and GA review of Martensdale, California, an infobox has been added to that article. Trick is, coordinates are needed to pull up a map. Exact coordinates for the site are not known, but the GA reviewer was able to pull up some coordinates that are so close it would have no visible effect on the pushpin map output. But it's still not best to pass off slightly inexact coordinates as exact ones. I thought I knew a method to pull this off, but after further looking around, it doesn't work like I thought it did. And looking at the documentation for Template:Coord (which is pulled by Template:Infobox settlement), there doesn't seem to be a way to not display the coordinates. Does anyone know of a workaround for here? Hog FarmTalk05:42, 24 March 2021 (UTC)
It's usually added in VisualEditor by what-does-this-button-do users. A search [46] currently finds 52 in mainspace. It's never appropriate except on Main Page. There are probably editors clearing them periodically. PrimeHunter (talk) 13:56, 23 March 2021 (UTC)
I cleaned up a bunch of these round about the end of December. Many of the affected pages also had other inappropriate magic words, such as __DISAMBIG__ which VE also allows indiscriminate and inappropriate use of. --Redrose64 🌹 (talk) 20:26, 23 March 2021 (UTC)
Based on phab:T118796 and phab:T100691 it seems this issue will not be fixed in VisualEditor. We can hide these options from the Page Settings panel using sitewide CSS pretty easily though, and do so only for the mainspace. @PrimeHunter, Doug Weller, and Redrose64: Thoughts? Frankly I think everything but the "Redirect this page to" should be hidden. — MusikAnimaltalk19:43, 24 March 2021 (UTC)
I am not a fan of using CSS for VE making it too easy to access these. I'd prefer to see an edit filter on the point. Izno (talk) 20:22, 24 March 2021 (UTC)
I thought of that, but an edit filter adds an unnecessary expense (albeit tiny), and since this issue is mostly with new users, they may not be able to make sense of what they did wrong or even know how to get back to the Page Settings panel to fix it. They might instead abandon their edit. A configuration flag in VisualEditor would be most optimal but as I said it seems maintainers are adverse to doing that for whatever reason. — MusikAnimaltalk20:44, 24 March 2021 (UTC)
@MusikAnimal: Wait, we can hide these options now? Years ago there was a discussion about hiding the option to add __INDEX__ (which trips 930 (hist·log)), but no one could find a way, because OOUI. Odd. There's a ve-test-page-settings-noeditsection class for the __NOEDITSECTION__ option, but I can't finding anything unique for __INDEX__. Suffusion of Yellow (talk) 21:13, 24 March 2021 (UTC)
Correct, unlike the other options discussed here, the option to add __INDEX__ does not appear to have a unique CSS class (though a patch to add one I'm sure would be merged). So we can't use CSS for that, but I think disallowing __INDEX__ in the userspace warrants a filter anyway because that can be abused (promo-only accounts, etc.), be it with VisualEditor or not. The other magic words we're discussing aren't ever abuse, just newbies clicking buttons. — MusikAnimaltalk21:18, 24 March 2021 (UTC)
They're both a cost. We can add 3 lines of CSS or we can prevent the edit. I'm just generally going to side on the prevent or warn on edit instead. :/ Izno (talk) 21:20, 24 March 2021 (UTC)
The cost of both I think is pretty trivial. You could argue the CSS is a hacky solution, but it's effective. The point is these options shouldn't be available to editors in the mainspace to begin with, hence why hiding them seems ideal to me. I'm okay with also having a filter too, if we want. While in this case I don't think we're dealing with vandals or bad intended editors, a vandal certainly could abuse these magic words (using VE or otherwise). In the case of Rudy Rochman, the editor apparently intentionally added __NOEDITSECTION__. While it wasn't quite vandalism, it suggests the editor would have manually added magic word via wikitext editing if they knew how. — MusikAnimaltalk22:20, 24 March 2021 (UTC)
Should keep in mind, that if a user omits their signature, SigmaBot should not be prevented from identifying the user, because the url in a signature is from a forbidden list. I don't know the technical implementations though. Shushugah (talk) 18:41, 21 March 2021 (UTC)
@Paul Erik: from a non-technical perspective, in general these are disallowed and an editor may be blocked for link spamming (c.f. Wikipedia:Signatures#External_links), as far as other controls sites on the spam blacklist added to signature should get stopped by the blacklist, but the external site in that deleted edit doesn't appear to be on the SBL. — xaosfluxTalk13:57, 22 March 2021 (UTC)
Sorry to BEANs this thread but unless it's been fixed in the last couple of years, there was disruption from someone who found they could insert blocked URLs into pages by putting the URL in their signature then using three tildes. If this is still correct and Too Much Information, please delete my comment. Johnuniq (talk) 22:18, 22 March 2021 (UTC)
@Johnuniq: I just tried putting a blacklisted link in my signature, and the spam blacklist worked as expected.[47] If you find a way around that, best to file a private phab task, though. :-) Suffusion of Yellow (talk) 20:22, 24 March 2021 (UTC)
Thanks. I wish there was a somewhat more intuitive timeline builder. Ideally, we should have a tool where the editor can sketch out the bar inputs and it will spit out the syntax. BD2412T
How do I print/save a wikipedia article without the brackets/numbers that are interspersed throughout the article?
I am trying to make a long post on reddit that includes me copy/pasting about 40 paragraphs of text from wikpiedia but all the text I grab includes numbers like this[1] and this[2]. Is there a fast way to remove the numbers with brackets around them? Also what are the bracketed numbers formally called? 2600:1006:B054:DE27:9042:63B0:7DA9:3E23 (talk) 02:10, 26 March 2021 (UTC)
They are called footnote markers or reference markers. Clicking them leads to a footnote or reference. Displaying the page with the added CSS .reference{display:none} will hide them. Your browser may have a way to do that. PrimeHunter (talk) 10:02, 26 March 2021 (UTC)
Two user-scripts that I invoke via on meta:User:Pigsonthewing/global.js have stopped working (but were working eysterday). That page has not been changed, and I have restarted my machine to no effect.
I tried the first script, and it's displaying the wikidata description under the article title for me. I checked with Firefox 87.0 on Linux and Chrome on my Android phone. William Avery (talk) 14:00, 26 March 2021 (UTC)
@Pigsonthewing: if you are using "strict mode" privacy settings in your browser you may run in to certain issues as well (that could be failing without a way to give you the error directly). You could try setting meta as "allowed" in such privacy settings. — xaosfluxTalk19:34, 26 March 2021 (UTC)
List of Stuff You Should Know episodes
Five years ago there was an issue at List of Stuff You Should Know episodes where having too many templates was preventing the references from showing. Fred Gandt came to the rescue and split the article into multiple lists. That worked for a time, until Teafed recently updated the article. Now the original problem is back, and no references are being displayed. Can anyone help? --Slugger O'Toole (talk) 14:07, 26 March 2021 (UTC)
It transcludes around 1300 {{Cite web}} from the season pages. That's a large part of the reason for breaking the template limit. The list only shows episode number, title, length, air date. I think it can do without the references (most of them don't work anyway) when they are on the linked season pages. If they are placed in <noinclude>...</noinclude> on the season pages then they will only display there. Then the "Further information" links would also actually show further information. They haven't really done that since episode summaries were removed from the season pages as copyright violations. PrimeHunter (talk) 14:51, 26 March 2021 (UTC)
PrimeHunter, thank you. When you say "for now," does that mean you have a more permanent solution coming? Because as it stands now you just removed some information so that others might be displayed. -- Slugger O'Toole (talk) 17:46, 26 March 2021 (UTC)
The page was broken before. It isn't now. All the information can still be found by clicking the "Main article" links. I already suggested another solution: Don't transclude the references. I haven't tested whether it's enough to get below the limit and it requires more work to implement so I made a quick fix for now, so we don't show a broken page to readers and others get a chance to chime in. If there aren't objections then I may try the suggestion in a couple of days. You and others are of course also free to do something. PrimeHunter (talk) 18:01, 26 March 2021 (UTC)
Another option would be to show the 1300 references as raw url's on the long list and only do the expensive cite web formatting on the season pages. This could be done by inserting two <noinclude>...</noinclude> like this:
<noinclude>{{cite web|url = </noinclude>https://www.iheart.com/podcast/105-stuff-you-should-know-26940277/episode/how-safecracking-works-54991133/<noinclude> | title = How Safecracking Works | publisher = iHeartRadio | accessdate = March 25, 2021 | date = January 2, 2020}}</noinclude>
The fundamental solution is not to transclude season pages into a single article. That practice causes this issue and the issue will not go away any other way. Izno (talk) 21:29, 26 March 2021 (UTC)
Whether it is common and accepted and whether it can be fixed by anything other than hacks and workarounds are mutually exclusive considerations. Izno (talk) 23:00, 26 March 2021 (UTC)
Find and replace
On Fandom, the editing window has a find and replace function. How can I get this on Wikipedia's editing window?--Launchballer18:47, 27 March 2021 (UTC)
I've enabled New wikitext mode. It's not quite right - I'm used to having 'Save' at the bottom and being able to enter my edit summary before I click Save - but this is better. Thank you.--Launchballer20:29, 27 March 2021 (UTC)
What are edits with #suggestededit-add 1.0 in the edit summary? Example: diff. I have noticed a few which changed the first letter of the short description from upper to lower case. Was that just coincidence or is there some gadget encouraging that? Johnuniq (talk) 04:17, 29 March 2021 (UTC)
Amazing, thanks. I followed a few links but can't see anywhere that there is a suggested edit to "fix" the short description. At Commons, there appears to be a tag (c:Special:RecentChanges/tagfilter=apps-suggested-edits) that lists changes by that tool, but I can't see a corresponding method at enwiki. Presumably the tool which searches edit summaries would find them. If anyone has tried the app, I'd be interested to hear what you do and what you see. Johnuniq (talk) 06:14, 29 March 2021 (UTC)
'you have been blocked from editing' message (mobile using desktop, logged out)
I tried to fix a typo and a notice came up that said I had been blocked from editing. There was no explanation why, nor any indication of how I could find out why. How can I find out? This has happened before, and I think it happens when I'm not using WiFi.
Also I feel like I've seen a reference to a discussion of this or something very similar somewhere recently. If that's still going on, can anyone point me to it? Thanks! —valereee (talk) 12:23, 30 March 2021 (UTC)
Is there a way to link (either wikilink or URL) to a template, including parameters, so that clicking on the link produces a page consisting only of the result of the template call?
The closest thing I could come up with is, using a simplified example, [https://en.wikipedia.org/wiki/Special:ExpandTemplates?wpInput={{URLencode:{{Convert{{!}}2{{!}}mi}}}} link] which produces a link that shows the desired result as a "Preview", below lots of undesired stuff.
As a potential solution, is there any way for a page to access a query string appended to its URL, or the section specified by #Section? MANdARAX•XAЯAbИAM20:13, 25 March 2021 (UTC)
Perfect! I tested it with my own CSS and it's exactly what I wanted. Thank you so much!I'll probably submit a request tomorrow to create the CSS page described above, unless an interface administrator happens to see this here and takes care of it before then. MANdARAX•XAЯAbИAM00:13, 26 March 2021 (UTC)
I'm not aware of MediaWiki pages made solely for withCSS or withJS. There are other possible uses like this to only display the page names in search results:
I was so happy to see the solution that I said it was "exactly what I wanted", but I now notice that there's one more tiny thing. Can something be added to the CSS to also hide the question mark and Help link? MANdARAX•XAЯAbИAM07:12, 30 March 2021 (UTC)
You can request , #mw-indicator-mw-helplink is added:
I'm a user of WikEdDiff, and I'm very interested in being able to customize the colors for this tool, i.e. the background colors of the added and deleted blocks. I'd be very grateful if someone could help me find an easy way to do it (that is, not copying the whole script and then modifying it). --Jotamar (talk) 21:13, 29 March 2021 (UTC)
I don't either, exactly. The Talk page is still there (so not renamed, and not deleted), but the link for their contributions is gone (I presume the last, undeleted edit has been revdeld, too, as it was pretty nasty itself). Are you trying to do something with/to/for/about them? An admin may have more insights for you. — JohnFromPinckney (talk)01:26, 30 March 2021 (UTC)
I just don't see the technical explanation. At first I thought it had something to do with a hyphen being added to the username in page history. I went to Meta and saw a number of revdeld log entries shortly after the user’s last edit. --Heymid (contribs) 01:53, 30 March 2021 (UTC)
It's still in Special:ListUsers and in the page history; when an account is hidden it is removed from these and from the contributions and logs but is still recognised as an registered account on the user, talk and contributions pages (unless something has changed recently). Their edits were restored by another account that has been hidden[49] - the difference can be seen in the page history. The user ID (41529844) is in the page history (unlike hidden accounts) but put it into Special:Redirect and it returns "permission error" the same as for missing IDs such as 41529896 and 41529932 (the hidden account is probably one of them), if an account doesn't exist with that ID it's usually "value not found". Peter James (talk) 12:29, 30 March 2021 (UTC)
I've found the hidden account, the name is almost the identical to this and others that were not hidden. It looks like hidden accounts now appear to be not registered, and it's likely that the only way to check that they have been hidden and not renamed, even when the username is already known, would be by trying to create a new account with the same name. Peter James (talk) 17:04, 30 March 2021 (UTC)
This account was handled by the steward team that does things like lock and hide global accounts. As it had activity here, there are local remnants left behind. I can't go in to details, but it will not be getting used anymore and there really isn't anything else that should need to be done right now - you won't be able to investigate it more due to overlapping visibility restrictions. — xaosfluxTalk16:13, 30 March 2021 (UTC)
The username is still in the page history; it usually disappears when the account is hidden but it looks like the process was not completed (either that or some data had to be restored from a backup). There are also three similar accounts that are not hidden (IDs 41529763, 41530029 and 41530100) - the first is locked but not hidden, the second is blocked but not locked, and the third is blocked here but not locked and the global account is either missing or hidden. Peter James (talk) 17:04, 30 March 2021 (UTC)
I don't know about hiding accounts, but renames were still local in 2012, and global accounts could be deleted, or were not created if a username was the same as an account on another wiki. Peter James (talk) 19:46, 30 March 2021 (UTC)
Working from a vector skin on GC v.89 (last version), Windows 10 20H2 (last version), I've noticed that the EngvarB script buttons only show in debug mode, although they don't seem to be working in that case. They just appear. Otherwise, they're completely missing, even after clearing the cache. No relevant js error was found. Script developer contacted here. Assem Khidhr (talk) 12:56, 31 March 2021 (UTC)
We've lots tons of bots and tools that we depend on for the smooth running of enwiki. It's always surprised me that we don't have any standardized logging, monitoring, and reporting system for them. WMF has such systems for their internal use based on the common Graphite and Icinga packages, but they're not available for user-maintained tools, which means we tend to rely on somebody noticing and posting a "Hey, I haven't seen any XXX updates in a while, is it down?" query. So, a few questions:
Am I correct that no such facility exists for user-maintained tools?
Assuming I am, what do we want to do about it? The likely alternatives are:
phab:T53434 is the relevant bug. Toolforge Kubernetes and VPS tools get Grafana graphs, i.e. Pageviews and XTools, while VPS also gets something called Nagf (Not another Graphite frontend). None of these offer alerts, as far as I know. According to wikitech:Portal:Toolforge/Admin/Kubernetes#Monitoring, Prometheus is/was the intended standardized replacement, with alerts, but I'm not sure how to use it or if that effort was abandoned.
If you just want alerts that your tool is down, UptimeRobot is pretty great and it's free. However this will only help if the maintainer is around to fix it when it goes down. If it is believed a simple restart will do it (which I think is often the case with reFill, etc.), you can ask a Toolforge admin to restart it on #wikimedia-cloudconnect. If someone is around they'll surely do it for you. Many maintainers (such as myself) write cron jobs to restart the webservice automatically after so many hours. Hacky, sure, but it works well. — MusikAnimaltalk04:21, 29 March 2021 (UTC)
Hmmm, this seems like a rather fragmented landscape :-) Graphina looks like the kind of thing I had in mind, but it's not clear how I can configure my tool to send data to the graphite database, and it looks like you need puppet access to actually configure dashboards. -- RoySmith(talk)15:21, 31 March 2021 (UTC)
Request: Copying an en.Wikipedia template to Commons
Hi all
Currently the Commons: namespace Commons has Visual Editor disabled, I really want to use VE on some documentation for Structured Data on Commons I'm creating to make it easier to contribute to.
There's an extremely helpful template on English Wikipedia called Template:VEFriendly which allows you to easily edit a page in VE but it isn't available on Commons yet. Please could someone who knows how to make templates make it available on Commons for me? I tried making it in my userspace first but it just won't work and I don't want to guess at things and break something by accident.
WP:REFLINKShas been down for three days now. Usually routine maintenance takes less time than that. If anyone can find out what is up with the tool that would be helpful. Formatting bare urls becomes laborious without it so any help will be appreciated. MarnetteD|Talk17:44, 22 March 2021 (UTC)
Thanks for the suggestion Xaosflux. I use it extensively - in fact I usually have one window with refill2, one with reflinks and one with citer when I'm working on those pesky bare urls. reflinks big plus is that it formats many PDFs which refill2 can't handle so when it is down those have to be done manually which slows everything down. An hour ago refill2 went down as well :-( Hopefully they are both being worked on. If not I'll be finding new wikignome tasks. Cheers. MarnetteD|Talk22:42, 22 March 2021 (UTC)
ReFill is working, just not when asking it to process an article on the English wikipedia. Is there something that may have changed on the English wikipedia in the last week that is causing the code to struggle to process the markup that it is ingesting? If I run the following cURL commands, the first (specifying the version of the Microsoft article on the German language wikipedia works as intended, but the second command to do the same on the English language wikipedia does not.
From what I can discern there isn't anyone actually working on this. To add to the frustration refill is working on other language wikis. Does anyone have any thoughts about publicizing this problem more widely? Perhaps the idea lab or even the next Signpost. If the old tools can't be fixed perhaps one of the programming boffins can create a new one. Otherwise it might be necessary to amend the guidelines to say that linkrot is simply an unavoidable fact of life on the English language WikiP yes that is gallows humor. MarnetteD|Talk00:55, 30 March 2021 (UTC)
@MarnetteD: I mentioned something over in one of the Wiki-related Facebook groups to which I belong. Didn't ask for anything...just floated a complaint balloon. If there's no motion on this in another day or so I may make a more forceful and pointed request for assistance/suggestions/etc. --Ser Amantio di NicolaoChe dicono a Signa?Lo dicono a Signa.01:28, 30 March 2021 (UTC)
Thanks for the info Ser. It would just be nice to know that our concerns are being shared by people who could actually do something to fix the situation. I am soldiering on using citer but it can only fix one ref at a time and that can take a lot of time on an article with only three or four refs let alone the ones with 10 or more. Cheers. MarnetteD|Talk01:52, 30 March 2021 (UTC)
@MarnetteD: At the very least. Personally, I think it's beyond time we get a new tool - reFill is outdated, and I keep having to clean up a lot of its citations. Besides which, this isn't the first time we've done this dance, though I think it's the most frustrating by far that I can remember. --Ser Amantio di NicolaoChe dicono a Signa?Lo dicono a Signa.01:59, 30 March 2021 (UTC)
Has anyone noticed their watchlist changing in the past several hours? Mine used to look like File:Wikipedia watchlist.png, but now it's formatted like * 20:50 Wikipedia:Administrators' noticeboard/Incidents (diff|hist) . . +367 . . Tenryuu talk contribs →User:Marveldc111: Closing discussion (DiscussionCloser v.1.7.3) That doesn't quite do it justice, as there's a massive gap between the first dot and the time. I checked in both Chrome and Firefox on a PC so I don't think it's on my end, but I figured I'd ask before seeing how to report a bug like this. Woodroar (talk) 21:17, 28 March 2021 (UTC)
Woodroar, that looks similar to how the watchlist looks when "Group results by page" is enabled. If you click on the dropdown with the label "x changes, x days" try to uncheck "Group results by page" if it's checked. – BrandonXLF (talk)21:41, 28 March 2021 (UTC)
BrandonXLF, I was just going to reply that I found it in the Preferences. I wonder how it got enabled because I certainly haven't been messing around in there?! Anyways, thank you very much! Woodroar (talk) 22:01, 28 March 2021 (UTC)
It would be nice if preference changes were logged. More than once, I've noticed something is different from what it used to be, without quite knowing what's changed, only to have it end up being that some preference got changed. Which I assume I managed to do by accident somehow. There's a zillion preferences settings, the vast majority of which I have no idea what they do. Being able to get a chronological log of changes to my preferences would really help debug these incidents. -- RoySmith(talk)18:42, 31 March 2021 (UTC)
@RoySmith: while not exactly what you were asking for, you can dump your preferences to an output using this link. Please keep in mind that link will include some of your private information so you should not post it publicly (such as on a wiki page), but you could save it to your computer to refer to in the future. — xaosfluxTalk19:09, 31 March 2021 (UTC)
http vs. https
Wikipedia went to HTTPS quite some time ago, as did many, if not most,sites on the web. I'm wondering about the remaining HTTP urls we have embedded in the sourcing of all our content. It would be a gigantic, mostly impossible, undertaking to correct all that. I have noticed that when clicking on an old http url - which we don't know until we click on a source - our browsers are taken to the old less secure URLs. It makes Wikipedia's sourcing insecure for the reader. Has this been discussed before? — Maile (talk) 14:25, 30 March 2021 (UTC)
@Maile66: are you suggesting we validate external links in our article sources and update them from http to https when the destination is still the same content? — xaosfluxTalk14:28, 30 March 2021 (UTC)
Granted, the content is the same. Perhaps you can clarify for me what aspect of http is "insecure" for the end user? There must have been a reason everybody went to https, but when I click on one of those links, I always wonder how that affects me personally. — Maile (talk) 14:36, 30 March 2021 (UTC)
@Maile66: "in general" for https vs https: those with access to the connection between your computer and the website (e.g. your local network provider / employer network / internet service providers / hosting center for the webserver; and those that can gain access to those connections such as governments) can eavesdrop on your connection and see what you read, and possibly can change what is being read along the way. For example, if you had http and read our article on Elephants , someone could tell not just that you connected to Wikipedia, but that you read the article on Elephants, and they could possibly tamper with the content on its way to you, and change what you saw. — xaosfluxTalk15:07, 30 March 2021 (UTC)
@Bender235: has done a lot of https conversion work. Targeting specific domains. I think a general bot that web scrapped every http URL looking for https would be difficult for a couple reasons. There are also sites that convert from https to http as they stopped paying the fee. -- GreenC15:53, 30 March 2021 (UTC)
I get it now. Thanks everyone. I guess the only sites to worry about are the ones that contain our personal and private information, and all mine have the https. I don't otherwise connect at sites I'd be worried about, and Wikipedia has already gone to https. — Maile (talk) 16:40, 30 March 2021 (UTC)
@Maile66: some people have reason to be worried about others finding out what they read about. For example say this was the reference link - some UN Report about Liberian women (on a http only link), I can't imagine anyone would care that I read that -- but in some places of the world it might be harmful to the reader if someone knew what they were reading. — xaosfluxTalk16:58, 30 March 2021 (UTC)
You need to validate that an https link is valid, which is not true of all websites. If it isn't, then you'd accidentally lead to link rot. Bots can intelligently update common websites they know to be https friendly. Shushugah (talk) 18:07, 30 March 2021 (UTC)
Editing pages seems to only automatically watchlist pages for a limited period. How do I make these additions permanent without manual un/watchlisting? DaßWölf00:31, 1 April 2021 (UTC)
Why has the logo for MediaWiki (the one at the bottom right of Wikimedia projects) changed from a yellow sunflower to a red-pink flower? 54nd60x (talk) 09:48, 1 April 2021 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
Some very old web browsersdon’t work well with the Wikimedia wikis. Some old code for browsers that used to be supported is being removed. This could cause issues in those browsers. [52]
IRC recent changes feeds have been moved to a new server. Make sure all tools automatically reconnect to irc.wikimedia.org and not to the name of any specific server. Users should also consider switching to the more modern EventStreams. [53]
Problems
When you move a page that many editors have on their watchlist the history can be split. It might also not be possible to move it again for a while. This is because of a job queue problem. [54]
Some translatable pages on Meta could not be edited. This was because of a bug in the translation tool. The new MediaWiki version was delayed because of problems like this. [55][56]
Changes later this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from 30 March. It will be on non-Wikipedia wikis and some Wikipedias from 31 March. It will be on all wikis from 1 April (calendar).
Auto hyperlink for "RFC #" text format does not work
Hi, articles like Hypertext Transfer Protocol and URL in the Infobox section and in the "Base Standards" part have "RFC #" format text (e.g. RFC 3986) that previously have been auto hyperlinked, i.e. "RFC 3986" text automatically hyperlinked to https://tools.ietf.org/html/rfc3986 and it becomes RFC 3986 and this was done "automatically". But this capability is not working now (see HTTP article for example). I think this is a good and useful and correct capability and also in Deutsch language for example, this "text format" is still correctly auto hyperlinked to the "ieft site", but in English it does not work now. Please correct this bug. Thanks, Hooman Mallahzadeh (talk) 15:32, 1 April 2021 (UTC)
Request: Please could someone create a template to show the number of files on Commons which use structured data (like a template on en.wiki)
Hi all
Please excuse the Commons request here (also posted here, I want a template which functions very similarly to a en.wiki one. I'm working on improving the documentation for Commons:Structured data and want to find a way to show the current use of it. One thing that would really help is to be able to write documentation that doesn't need to be updated manually. There is a template to show number of articles on English Wikipedia, would there be a way of having a template that showed the number results of:
This is not possible in 'automatic' template form. You will need to create some sort of database query and update a template manually (or via bot, like is done for the template transclusion counts on enwiki). Izno (talk) 00:13, 2 April 2021 (UTC)
Does the MediaWiki software enable the possibility of decimal-time block durations? Intrigued to know since I recall an admin trying to block a user for pi seconds some time ago. --Heymid (contribs) 11:52, 1 April 2021 (UTC)
@Heymid: no: (a) the parser for that only accepts numerical values in positive integers; (b) the blocking database only has precision to 1 second for datetime values. — xaosfluxTalk15:22, 1 April 2021 (UTC)
(c) any admin that did that would be hauled straight to arbcom for misuse of admin rights. Not by me though. --Redrose64 🌹 (talk) 23:18, 1 April 2021 (UTC)
Guilty as charged ... as a bad April fools' day prank that I regret now, even though it didn't have any major consequences. I tried a similar test with 2.71828 seconds on my account before then. Both of those non-integer-second blocks were displayed with durations like "1 January 1970" for a while. When I originally made them they were displayed with the correct number of seconds. Graham8708:19, 2 April 2021 (UTC)
Any technical reason why "Show previews without reloading the page" isn't the default setting for new accounts?
Referring to the last checkbox in Preferences / Editing. To me, activating this option not only results in a shorter load time, but also has the added benefit of not resetting the scroll bar and cursor in the source editor. Since getting new editors to hit the preview button instead of accidentally making a mess and getting their Talk page templated by a well-meaning recent changes patroller is something we're somewhat struggling with, I feel that making this feature more convenient to use would have nothing but benefits. AngryHarpytalk09:35, 29 March 2021 (UTC)
No reason that I know of. Absolutely concur that this should be turned on by default. It also preserves the undo stack across previews so that Ctrl+Z undoes changes done before previewing. – SD0001 (talk) 10:06, 30 March 2021 (UTC)
If you enable it and preview a page which transcludes itself, typically a template with example calls in the documentation or in noinclude tags, then the transclusion uses the saved version instead of the previewed version. That is confusing and makes template development harder. For example, try previewing Template:X1 with this:
The documentation should show Foo but it doesn't. Toggling the feature doesn't affect edits which have already started so enable it before starting the test. I don't know whether the feature has other bad quirks but this is enough for me to not use it. Many users rarely or never edit templates. PrimeHunter (talk) 12:15, 1 April 2021 (UTC)
PrimeHunter, I see, that's certainly not ideal. An argument could of course be made that this is a fairly fringe issue, especially for new users, but it would certainly cause confusion for some of them down the road. Since I'm admittedly extremely clueless about the backend of Wikipedia, I'm just going to ask outright: Is getting this quirk simply ironed out a realistic option here? AngryHarpytalk08:40, 2 April 2021 (UTC)
It's phab:T184466 from 2018: "Live preview doesn’t use new versions of templates embedded in themselves". I don't know how hard it is to fix but there appears to have been no work in three years. PrimeHunter (talk) 09:29, 2 April 2021 (UTC)
We could just disable live preview in template ns of course... Not that strange, as it also doesn't support the TemplateSandbox extension for instance... Then again, at some point, we should probably also fix this for the 2017 WE, which I assume has the same problem. —TheDJ (talk • contribs) 13:52, 2 April 2021 (UTC)
Any article that uses cs1|2 templates ({{cite book}}, {{cite web}}, {{citation}}, etc) or uses {{sfn}} (and family) or {{harv}} (and family) templates transcludes itself. The cs1|2 templates transclude the articles that use them so that they can render publication, archive, and access in the form specified by the {{use dmy dates}} and {{use mdy dates}} templates. The {{sfn}} and {{harv}} families of templates tranclude the articles that use them so that they can do error detection (see Category:Harv and Sfn template errors). At present, that is about 4.7 million articles. There is a slow-boil move afoot to create some sort of automated mechanism to render all dates in an article according to an existing {{use xxx dates}} template.
I don't know if it's because I use ol' stable Monobook, but enabling this feature doesn't create a live preview or do anything for that matter. - Floydianτ¢16:36, 2 April 2021 (UTC)
@Floydian: it doesn't make a "live preview" as you type, but when you click "preview" it will soft refresh the preview area of the page without reloading the page. — xaosfluxTalk17:53, 2 April 2021 (UTC)
Image alt parameters
I've been adding image alt text occasionally and noticed that when I mouse-hover over an infobox image, the alt text is revealled as an overlay; when the image is in the article body no such annotation occurs. Recent examples of non-infobox articles are at History of the Jews in Hull and Matlock Bridge being building and bridge descriptions repectively. A recent example of infoboxed is at Bart Simpson which shows the text overlay on Bart's caricature-illustration.
Could someone please confirm if this (non-appearance) is the 'norm' and known, and preferably check with a screen-reader? For my own peace of mind. I'm using only IBM and almost-exclusively with Firefox 87.0. Thanks.--Rocknrollmancer (talk) 15:01, 27 March 2021 (UTC)
@Rocknrollmancer: We've recently lost our Grade A1 alt-text expert, so I'll try to answer. The alt text is present on the images at History of the Jews in Hull, Matlock Bridge and Bart Simpson, you can check this using your browser's "Inspect element" or "View source" features; it's the alt="..." attribute of the <img /> tag. In the case of Bart Simpson (but neither of the others), the alt text is copied to the title="..." attribute of the enclosing <a>...</a> element, and it is that title that some browsers are displaying as a tooltip.
The purpose of alt text is not to provide a tooltip (although some browsers do that) - it's to provide replacement text for use when images are not available, or, more fully, The image given by the src and srcset attributes, ... is the embedded content; the value of the alt attribute is the img element’s fallback content, and provides equivalent content for users and user agents who cannot process images or have image loading disabled. (see HTML 5.1 spec). So when the image can be displayed, the alt text need not be rendered; but if the image can't be displayed - such as if the upload.wikimedia.org server is flaky - you instead get a rectangle of the appropriate size, but instead of the image it contains text such as "Close up of natural stone arched bridge and tapered piers over river on a sunny day with blue sky". --Redrose64 🌹 (talk) 17:04, 27 March 2021 (UTC)
ThanQ Redrose64, BrandonXLF - I'm beginning to understand, I've been out for a few hours and realised when away from the keyboard that I'd not described adequately-enough the query. What I was needing to know was - for visually-impaired, is the alt text converted to a voice announcement? Or just the main caption? I know of a young lady who's lost her previous partial-sight totally about two years ago and is using Voice Over for iOS (?) on iphone, that's from memory as I'm unfamiliar with whatever that is (or if there's an equivalent on Android app or plug-in). Checking, she's taken it off her social media profile now, but the previous annotation stated not to send her images without a caption. That was the main gist of the enquiry, and I probably mis-stated when using the term screen reader. Hope that's clear-enough!--Rocknrollmancer (talk) 21:20, 27 March 2021 (UTC)
Yes, this is why WP:ALT#Basics says The alt text is intended to be read out by screen readers just before the caption, so avoid having the same details in both. --Redrose64 🌹 (talk) 22:58, 27 March 2021 (UTC)
Good-O, pleased to have that clarified - it's been a long time since I looked; think it was when I tried to flush-right a TOC and it was reverted as potentially causing problems, but I couldn't quite remember on what device(s) were mentioned and how the hard/software may have developed since. Thanks for all your comments.--Rocknrollmancer (talk) 00:12, 28 March 2021 (UTC)
Thank you, Graham87, comments noted. I have only just seen your message from 28 March today, 2 April, when I wanted to submit another query, and the browser landed at this section from past browsing history. I have now seen the blank image field displayed momentarily with text, before the image appears.--Rocknrollmancer (talk) 18:48, 2 April 2021 (UTC)
What non-trivial thing could cause 1 table row out of many to exist in the wikitext but not previews or after posting?
It was a missing < on a </ref> which was separated from a |- on the next line by only a line break which made everything disappear till the next proper |-. Sagittarian Milky Way (talk) 19:33, 2 April 2021 (UTC)
I have been seeing blue links with underscores, often. I do not understand why some are shown this way with underscores. The second example below is a by bot reverting a BLP addition. Both observed 2 April. What is the significance, if any? Under what circumstances would they (underscores) be appropriate. How would I achieve this, if I needed to? Thank you.--Rocknrollmancer (talk) 21:13, 2 April 2021 (UTC)
Underscores and spaces are interchangeable in wikilinks [[...]]. It's nearly always best to use spaces. Some tools use underscores for no good reason, maybe slightly easier coding or a programmer who didn't think about it. Underscores added by editors are probably copied from url's in most cases. Spaces are represented by underscores in url's. They are not interchangeable there. The only case where underscores is better in wikilinks is if the subject is written with underscores like static_cast and others in Category:Articles with underscores in the title. PrimeHunter (talk) 22:01, 2 April 2021 (UTC)
Hey Rocknrollmancer, underlines are purely optional. This commonly happens when copying/pasting from the browser bar instead of the visual text on the screen. I try to remove the underlines for visual reasons, but both are acceptable technically and lead to exact same page. When pasting a link outside of wikipedia, the underscores are important, so that your website knows it's one url and not multiple. Shushugah (talk) 22:06, 2 April 2021 (UTC)
I had a similar conundrum with a clerk announcing at approx 11.30PM that evidence phase would close at midnight "in approx two hours". I had to query this as I'm used to ebay time as unequivocally 23:59:59, not with a +90/120 minutes tolerance. Stated to be flexibility for whomever was online after midnight. BTW, the 1 minute disparity when signing above was how my digital clock was manually set.--Rocknrollmancer (talk) 23:42, 2 April 2021 (UTC)
Spaces are never permitted in pure URLs - if desired, they must be encoded in some way, but the manner of encoding varies between websites. If the URL has a query string, most sites will use a plus sign + to represent a space in that part of a URL. Spaces before the query string are usually percent encoded, becoming %20. Wikipedia (and other sites using the MediaWiki software) allow the underscore as an alternative to %20, but not always as a valid replacement for the plus sign + in a query string. --Redrose64 🌹 (talk) 07:50, 3 April 2021 (UTC)
My scripts are not working correctly
A bunch of my installed gadgets aren't working for me since yesterday, but I see through edit summaries on my watchlist that they are still operational. I'm not able to use the User:GregU/dashes.js, User:Ohconfucius/script/EngvarB.js, or User:Ohconfucius/script/MOSNUM dates.js scripts, and possibly others. Twinkle is partially working for me, but it wouldn't allow me to welcome an IP earlier today. There's also a weird thing on my notifications, where instead of displaying the time since the ping, it shows {{PLURAL:$1|$1h}} for all of them. Any ideas of what is happening? I haven't done anything that would have changed anything. – Muboshgu (talk) 23:00, 1 April 2021 (UTC)
Longer reply coming: This is common when one of your scripts breaks, usually due to a change in MediaWiki. Izno (talk) 23:29, 1 April 2021 (UTC)
Let's do some cleaning:
Use the Twinkle gadget in preferences. Remove the line in User:Muboshgu/vector.js which imports it there.
Remove everything in User:Muboshgu/monobook.js except for the last line. Move the last line to User:Muboshgu/common.js if you still want access to Lupin (I do not know if it works). Consider whether the options stored in monobook.js are still relevant to your WP:TWPREFS and add them there.
Izno, all done with those steps, it did not fix the problem but I was noticing how cluttered it was getting on those three subpages. – Muboshgu (talk) 23:55, 1 April 2021 (UTC)
@Muboshgu: in User:Muboshgu/common.js, turn off all the scripts, then make sure your gadget-powered things are working. Then turn then back on one at a time to try to find which one is causing you a problem -- then go follow up with that script's author. — xaosfluxTalk23:58, 1 April 2021 (UTC)
Okay, one more besides Xaos' comment: Remove the line for User:Kephir/gadgets/rater.js also, since Evad's version is maintained. Izno (talk) 00:10, 2 April 2021 (UTC)
With all of the scripts disabled, Twinkle is still not working right for me. And reenabling User:Evad37/rater.js back on (that script had been working for me before), it is now not working. :/ – Muboshgu (talk) 00:12, 2 April 2021 (UTC)
Nm that rater part, apparentlly the Evad rater shows up in a different spot then the Kephir gadget I'm used to. I'll just have to keep futzing around I guess. – Muboshgu (talk) 00:20, 2 April 2021 (UTC)
for now. I reset all of my preferences back to default and spent a bunch of time toggling scripts and gadgets on and off. I've gotten everything I need working, and found some interesting tools I hadn't tried before. Thank you everyone for your help! – Muboshgu (talk) 03:32, 2 April 2021 (UTC)
@Izno and Xaosflux:. Thanks for troubleshooting the problem. FYI, I would confirm that there have been no changes to my scripts of late from my end, and what you have done to solve the problem seems to prove it. Best, -- Ohc ¡digame!16:33, 3 April 2021 (UTC)
Creating Special:Diff links?
If I'm on a diff page, is there an easy way to generate a Special:Diff/xxxx wikilink to that page? As far as I can tell, you need to manually type "Special:Diff/" and then copy-paste the oid out of the browser url bar. That's enough of a pain that I don't usually bother, and just grab the whole url. Am I missing something obvious? -- RoySmith(talk)14:16, 3 April 2021 (UTC)
Registration is now open for a scientific study to test the efficacy of my new anti-promo tool. If you are interested in participating and you meet the inclusion criteria, please leave a message on my talk page. Inclusion criteria and other information about the study may be found on my talk page. Sam at Megaputer (talk) 04:26, 4 April 2021 (UTC)
Ultimately because the underlying bug hasn't been fixed yet, and while FlaggedRevs doesn't have an "owner" on the dev team it's unlikely to get fixed any time soon! ƒirefly ( t · c ) 19:01, 27 March 2021 (UTC)
Huh, I don't really know if edits using Twinkle count here, but apparently, my edits at Smilodon are automatically accepted now? Maybe this a sign that they're starting to fix the bug? Hopefully... JurassicClassic767 (talk | contribs) 17:48, 29 March 2021 (UTC)
I have the same issue from what I have been seeing, see Special:AdvancedReviewLog and there were many bot accepts of edits by autoconfirmed users that have not been auto accepted, and also I have been trying to accept edits by autoconfirmed users that have not been auto accepted. This issue is still present. User3749 (talk) 17:34, 4 April 2021 (UTC)
I wonder if someone could help me change the name of this place back to "Peyton". On March 5, 2020, an IP changed the name. I'm not sure if changing it back will make something go wonky. Thanks for your help! Magnolia677 (talk) 21:52, 4 April 2021 (UTC)
@Magnolia677: IPs cannot rename pages. The only IP edits dating from March 2020 are these two, which are simple text changes and should easily be revertable by any user. No special rights are required. --Redrose64 🌹 (talk) 22:36, 4 April 2021 (UTC)
@Redrose64: Sorry, I should have given more detail. Last year an IP went through the text of the article and changed all the "Peytons" to "Pentons". Then today User:Maxbmogs actually changed the name of the article to Penton. I changed all the text back to Peyton today, but the name of the article still needs to be changed back. Thanks! Magnolia677 (talk) 22:42, 4 April 2021 (UTC)
Hi, the links which hide and show a collapsible table are commonly "[Hide]" and "[Show]". Can the links have other names? "Collapse" and "Expand" for example. Is there a way to have the collapse link in two places; at the top of the table and at the bottom? Thanks, PeterEasthope (talk) 23:07, 4 April 2021 (UTC)
Collapse links cannot be placed at the bottom of a table. You cannot customize on per-page what the toggle displays, that can only be done in Javascript (it might be possible for you to personalize for yourself if you work hard enough, but I do not guarantee that it would look pretty). Collapse and expand are actually the default, but for historical reasons we preferred hide/show. Izno (talk) 01:49, 5 April 2021 (UTC)
In page histories, manually accepted revisions are now displayed with a #d5fdf4 background instead of the #eaf3ff background used for automatically accepted revisions. The text "accepted by Username" has been replaced by "approved by Username", a change which I have overriden at MediaWiki:Revreview-hist-quality-user to match our protection policy's wording. If we find a little consensus to do so, an interface administrator can also override the new color. That said, I personally don't mind seeing the distinction between automatically and manually accepted changes. Note: Twinkle's rollback manually accepts the resulting revision, resulting in brightly highlighted vandalism reverts. ~ ToBeFree (talk) 00:43, 5 April 2021 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
Editors can collapse part of an article so you have to click on it to see it. When you click a link to a section inside collapsed content it will now expand to show the section. The browser will scroll down to the section. Previously such links didn't work unless you manually expanded the content first. [57]
Changes later this week
The citoidAPI will use for example 2010-12-XX instead of 2010-12 for dates with a month but no days. This is because 2010-12 could be confused with 2010–2012 instead of December 2010. This is called level 1 instead of level 0 in the Extended Date/Time Format. [58]
The new version of MediaWiki will be on test wikis and MediaWiki.org from 6 April. It will be on non-Wikipedia wikis and some Wikipedias from 7 April. It will be on all wikis from 8 April (calendar).
Future changes
PAWS can now connect to the new Wiki Replicas. Cross-database JOINS will no longer work from 28 April. There is a new way to connect to the databases. Until 28 April both ways to connect to the databases will work. If you think this affects you and you need help you can post on Phabricator or on Wikitech.
Need a new template design for religious text's verses
Can someone please invent a reference template (that maybe includes the preview ability) that allows a Bible verse to pop up and be readable within the text of an article? If it could be designed to be used by plugging in any religious text, that would be awesome as well. It probably needs to go to one of the off-WP sites that enable reading, and I think that's been the hold-up with this. They are all about the same, so picking one is max-nix. Right now, I can't find a way to reference texts under discussion so our readers can easily see what's being talked about. I am not techno at all, and as a result, I have total faith in your ability to work this magic! :-) It's really needed and would add value to WP articles on religion. Thank you!Jenhawk777 (talk) 18:56, 3 April 2021 (UTC)
This is basically not possible in a template without hosting a copy of the Bible on Wikipedia somewhere. Which then gets into the reason we have one or two templates that give you a choice of Bible to read. Izno (talk) 19:06, 3 April 2021 (UTC)
Oof. The idea of the warring over which version of a religious text to use in these pop ups causes me to shudder.--Jorm (talk) 19:08, 3 April 2021 (UTC)
But Jorm dear, the current Bible verse template [[s:Bible (King James)/Psalms#Psalm 23|Psalms:23]] already allows the editor to choose which version of the Bible to reference. The current template is a link to those verses that are already available on wikipedia, just for that purpose, and it is already possible to link to the Revised Standard and a couple other versions as well. There's no reason to change any of that. My request isn't affected by that, and really, I think that's a non-issue. I've looked at several Bible sites and they all provide multiple versions of any verse you look for - automatically. You pick a version for what you are looking for, and voila, that's what pops up at the top of the page, but then there's also a list of the same verse in multiple other versions below it. Why can they do it and we can't? I have referencing envy.
But my actual complaint is that the current template isn't a pop up. A reader can't just hover and read the verse like in a preview function. That would be valuable - and would simply show the verse in whatever version the editor had chosen. It can be done Jorm, because it is being done elsewhere. We need that ability. It would better the encyclopedia by making it easier and more informative for our readers. Jenhawk777 (talk) 04:59, 4 April 2021 (UTC)
No one is really saying the idea is impossible, it is doable, but it is tricky. [[s:Bible (King James)/Psalms#Psalm 23|Psalms:23]] links to an different website, called Wikisource, hosted by the same company as Wikipedia and as such the link is internal (although we also have internal links to non-affiliated wikis, but that is beside the point). Fetching content from a different website basically is not possible. If you really want to be thorough then the exception is Wikidata, but they rejected hosting bible verses already, all other wikis cant get data from each other since mw:Manual:$wgEnableScaryTranscluding is off, and no it will not be turned on. Like has been said the only option is to have the verses locally and make the template work that way.--Snaevar (talk) 08:18, 5 April 2021 (UTC)
If the passage is relevant enough for someone to definitely want to read it, then simply quote it. If it is not quite that relevant, there are several methods of making Bible references of which {{Bibleverse}} is probably the best. – Finnusertop (talk ⋅ contribs) 09:08, 5 April 2021 (UTC)
I noticed that someone finally added preview cards for enwiki articles, showing the first image in tweets, on slack, etc. I thought that was declined by the tech team. I'd like to give a hearty thank you to whomever coded that. Anyone know how to find out? mw:Extension:TwitterCards is still listed as "experimental" and is not shown as installed on any Wikipedia. 2601:647:4D00:2C40:6130:2EC6:6215:83F0 (talk) 23:44, 5 April 2021 (UTC)
I've come across a cluster of Wiki articles that are loaded with unsourced statements. What is the best way to get them on a list of articles that need improvement that everyone can see? I've been adding the Ref Improve template to the top of such articles. Does that automatically put it on some "articles needing improvement" list?
In some cases, the entire article is unsourced. Some contain line after line of equations in math-major notation, with no supporting footnotes. The average reader has no way to ascertain whether that "stuff" is supported by experts in the field, generally accepted, or some anonymous Wikipedia editor's original work. I've been trying to find reliable sources, but it's difficult because that requires access to specialized textbooks that aren't readily available online or in my local library. The articles I'm talking about concern finance. For example:
I just added a comment at the last thread Talk:macOS High Sierra, and the edit window blanked all the existing text. Consequently, this diff has removed a bunch of comments (when I didn't want it to!) and ended up with a reply to the wrong thread. I know talk page communication is difficult, but if I'm struggling, what hope have new users got? Ritchie333(talk)(cont)14:58, 6 April 2021 (UTC)
Same issue here, I was only able to get to edit here by doing it in mw:safemode. It doesn't appear to affect me while logged out (I popped into an incognito window and appeared to be able to edit fine). ‑‑ElHef (Meep?) 15:20, 6 April 2021 (UTC)
Yeah, I just realized what that was referring to and I do have that enabled in prefs. Thanks for the quick fix. ‑‑ElHef (Meep?) 15:42, 6 April 2021 (UTC)
I have the same problem as Ritchie above. I tried to add this comment in the previous section, but was unable to do so. RolandR (talk)15:05, 6 April 2021 (UTC)
Apologies if this is the wrong place, but I don't know where else to ask I regularly search for links in articles to User pages, as these are contrary to MOS:DRAFTNOLINK and are often vandalism - usually linking to unacceptable content - or people mistakenly signing their additions to articles To find these I search for insource:User insource:/\[\[ *User *:/ sorted by edit date - as here Today, User:polbot has auto-created 150-160 articles, each of which includes User:polbot and this; along with User:GreenC bot/Job 18 and User:Kvng/RTH are cluttering up the search results, making it difficult to find the links to "real" users pages. However, I don't know enough about the coding to omit those three "users" from the search results - can anyone help me out ? Thanks - Arjayay (talk) 15:14, 6 April 2021 (UTC)
Wow PrimeHunter thanks a lot. I started off with 3250 search results this morning, SD0001 reduced that to 750, and you've got it down to 40 - My only regret is not asking sooner - thanks again to both of you - Arjayay (talk) 17:19, 6 April 2021 (UTC)
I've discovered that some pages like Doctor Who (film) are being transcluded in other mainspace pages. I'm not sure I agree with this practice. This is a job that I think should be done only with templates. Most editors, especially if you're using VisualEditor, probably don't realize that editing the table at Doctor Who (film)#Production effects over 340 other articles. With that in mind, I have semi-protected the page indefinitely as high-risk, which is something that happens automatically in the templatespace.
What's even worse I think is using parser functions like #section and #section-h. This kind of selective transclusion offers useful functionality but it is fragile. #section requires special markup on the target page, and #section-h will break all transclusions if an editor changes or removes the target section heading.
Is there any technical reason not to use templates for these use-cases? If not, should we propose a policy/guideline to disallow mainspace transclusion? The practice seems to be widespread (~3,500 articles have a transclusion count of 5 or more). It would take a lot of work to convert everything to use templates, but we could at least deprecate and/or formally discourage this practice.
I have for quite some time considered making a bot to deal with transcluded sections changing names, but haven't got around to it. I think it's actually quite a good idea to tranclude sections to perhaps a handful other pages but when the count gets to the double digits it should probably be a template. I think in any case MusikBot probably should be enabled in mainspace, just in case. --Trialpears (talk) 16:24, 6 April 2021 (UTC)
The people wanted section transclusion and we gave it to them. This is what happens. I agree with Trial that while we support it, MB should probably operate in the mainspace. There should probably be a report for "large number of main space transclusions". --Izno (talk) 18:27, 6 April 2021 (UTC)
Exactly, and if you follow the first link that I provided in my last post, you'll see that I already mentioned that. --Redrose64 🌹 (talk) 18:57, 6 April 2021 (UTC)
I think it's OK to transclude part of an article in one place. A common example is a TV series transcluding an overview from a list of episodes. For example, 24 (TV series)#Overview transcludes List of 24 episodes#Series overview. It does sometimes fail when somebody removes onlyinclude tags and the whole list is transcluded. It could be nice with a tracking category of mainspace pages which transclude another mainspace page without onlyinclude tags. PrimeHunter (talk) 01:36, 7 April 2021 (UTC)
A special page or database report wouldn't be very useful. A category could be used in searches like incategory:... -hastemplate:Excerpt -insource:... The search link could be placed on the category page. PrimeHunter (talk) 09:35, 7 April 2021 (UTC)
In that example, the transcluded part is literally just a call to a template. You could make a wrapper for that template and simply transclude it where needed, then you don't need the fragile <onlyinclude> tags, etc. Right? — MusikAnimaltalk03:42, 7 April 2021 (UTC)
There is code like <includeonly>List of 24 episodes</includeonly>#Season 1 (2001–02). It would be a little more complicated but could be done. It would also be harder to update a separate template and check that section links work. PrimeHunter (talk) 09:35, 7 April 2021 (UTC)
Agree that there's not a problem with doing this in general, although it should only be used when a significant amount of content is actually being transcluded and not in cases like the one that started this thread where the transcluded content is one purely static number. * Pppery *it has begun...03:34, 7 April 2021 (UTC)
Where is this totally incorrect media viewer caption coming from?
Because of maintenance on their primary database master, there will be a read-only period on English Wikipedia starting around 05:00 UTC on April 28. This means you can read, but not edit, the wiki. The window is 30 minutes long, but it will most likely be significantly shorter. See phab:T279505. /Johan (WMF) (talk) 11:36, 7 April 2021 (UTC)
Hi, A significant portion of the watchlist table in this wiki is just watchlist of bots. It's because "watch pages and files I create" is turned on by default and bots that create pages (like user talk pages) have them added to theirs. It distorts Special:UnwatchedPages and I doubt any user checks their bot's watchlist so I'm inclined to delete all of them here but let me know if you object to this. Also, if your bot created lots of pages here and you're okay with the rows being deleted, please let me know here and I'll delete it sooner. I already deleted mine. See phab:T258098 for more info. Ladsgroupoverleg13:54, 7 April 2021 (UTC)
I'd also be supportive of having a dev manually turn off that setting (one time run) for any of those same identified bots that are still creating pages. — xaosfluxTalk14:02, 7 April 2021 (UTC)
Oh lord, cluebot what have you done. I think it would be worthwhile to also purge the watchlists for these bots if not most/all bots. —moonythedwarf (Braden N.) 14:39, 7 April 2021 (UTC)
phab:T258108 was resolved in early January and was announced in Tech News, so these bots have already stopped auto-watching new pages. If they do need to watch pages, they have to explicitly set the 'watch' option when editing with the API. — MusikAnimaltalk16:07, 7 April 2021 (UTC)
With effect from 1st March 2021, chapter numbers are no longer used when citing Singaporean enactments. According to the Interpretation Act 1965, as amended by the Statute Law Reform Act 2021:
8.—(1) It is sufficient for all purposes to refer to a written law—
(a) in the case of an Act—
...
(ii) where the Act, as enacted or revised, provides that it may be cited by a short title—by that short title; and
...
— section 8(1)(a)(ii), Interpretation Act 1965
Examples of using the new citation method can be seen at new bills like this and this.
I therefore propose to make updates on {{Singapore legislation}} to bring conformity with the new citation method.
Also, I would be grateful if the updated Singapore legislation template can direct references to individual sections, like that in {{Cite Hong Kong ordinance}}. This helps standardization of various citations.
As I have no experience on dealing with these complex templates, and are clueless about how to ensure that enactments are redirected to the correct external pages, I hope anyone who are familiar with templates can give a helping hand on these problems, many thanks.廣九直通車 (talk) 03:24, 8 April 2021 (UTC)
OK, I can look into it. Whether linking to specific sections, etc., is possible depends on whether the website actually has links to them. — SGconlaw (talk) 04:53, 9 April 2021 (UTC)
Perhaps it will be better to create another new template for more specific section linking? Anyways, I also find that adding "#pr*-" (where "*" is a number) at the end of the principal URL at Singapore Statues Online directs you to the specific section.廣九直通車 (talk) 09:17, 9 April 2021 (UTC)
Plainlinks are not displaying correctly in timeless skin
All the edit links are now showing an arrow indicating an external link. I do not know if this is a bug or what, but the class=plainlinks should stop it, and yet it is not. It looks dreadful. Can this be fixed? Aasim (talk) 01:18, 9 April 2021 (UTC)
BTW this is happening for other plainlinks, including links in AFD templates, links on shortcuts templates, even on some interface messages. Aasim (talk) 01:20, 9 April 2021 (UTC)
@Awesome Aasim: just want to verify that this is not actually all the edit links on a page - for example on this random page the "edit links" on: sections, the literal labeled section "edit links" in the languages box, and the "edit this page" control at the top of the page are fine. You are only seeing this on full url's that contain things such as ?action=edit or ?redirect=no correct? — xaosfluxTalk11:01, 9 April 2021 (UTC)
The API seems to be interpreting line breaks in the wikitext as new lines even though they don't render that way to the reader. For example, in the article Hytracc Consulting the API call sends the sentence:"The company is \nheadquartered in Stavanger, Norway, and also has offices in \nAberdeen, Calgary, Groningen, Houston, Trondheim, and \nKuala Lumpur." Can someone link me to the right place to report this? Sam at Megaputer (talk) 14:16, 9 April 2021 (UTC)
The API is accurately reflecting the source wikitext and doesn't know anything about how HTML is rendered by a browser. This is not a bug. Izno (talk) 14:24, 9 April 2021 (UTC)
If the API is reflecting the source wikitext, then why do double linebreaks get reported as single line breaks (see the text after Kuala Lumpur.\n)? Sam at Megaputer (talk) 14:35, 9 April 2021 (UTC)
What you've linked to is the output of the mw:Extension:TextExtracts API, which uses certain heuristics to produce a page extract and so unlike mw:API:Revisions it will not necessarily reflect the source wikitext. If you think it's a bug, report it here, though high changes are that it's normal behaviour. The wikitext formatting of that page is weird anyway. – SD0001 (talk) 14:57, 9 April 2021 (UTC)
Sam at Megaputer, Sam, this is working as designed. See WP:LINEBREAK#Causing line breaks. Sadly, I don't know of any authoritative documentation for wiki syntax and semantics. The language is poorly designed and full of historical oddities.
If you're interested in parsing wikitext, the official answer is to use Parsoid. Parsoid is big, complicated, and confusing to use, but it is the final word on parsing wikitext. If you're using Python, mwparserfromhell is a reasonable alternative; it's much easier to use, but diverges from Parsoid in some minor details. If you're trying to parse wikitext with regexes, you're doing it wrong. -- RoySmith(talk)16:00, 9 April 2021 (UTC)
Thanks for the suggestions. I see from WP:LINEBREAK#Causing line breaks that "\n" in the wikitext is expected to register as " " in HTML, while "\n\n" registers as a line break in HTML. But the TextExtract API maps both values to "\n", which makes it impossible to tell if "\n" should be interpreted as a line break or not. If I don't get a response to my report, it might just be easier for me just to consider this as acceptable noise that to try to re-parse the wikitext. That's always an option, though. Sam at Megaputer (talk) 17:54, 9 April 2021 (UTC)
Sam at Megaputer, I have never used TextExtracts myself, but from reading the docs, it looks like this was intended as a way to grab just the text without any of the markup, as you might want if you were building a search index. If you want to get any of the markup semantics (i.e. "there's a paragraph break between these two words"), I suspect Parsoid is your friend. -- RoySmith(talk)20:32, 9 April 2021 (UTC)
User:Citation bot does some GB URL cleanup. But I agree we need a dedicated bot for Google Books because there are many issues that can only be determined/fixed with header and page checks. There are millions of Google Books links on enwiki, and many of them no longer work as originally intended: hard and soft 404, link has a page # but redirects to "About this book", book ID usurped and points to a different book, etc.. more at Wikipedia:Google Books and Wikipedia. -- GreenC03:04, 10 April 2021 (UTC)
TOC font size in modern skin
Has there been a change made to the modern skin css to reduce the TOC font size? I might be imagining it but the font size seems a lot smaller than it was. Nthep (talk) 14:34, 10 April 2021 (UTC)
This is one of those things that's been annoying me forever. On my Android phone viewing the desktop site with Chrome (Vector skin), references appear in either normal or reduced size text (see screenshot).
This seems to be related to whether I've visited a link or not, but it's not that simple. For example, if I tap on the first reference (Connor), the first thing that happens is the text shrinks to the small size. If I tap again, it takes me to the UPI page. If I go back to the wikipedia page and reload, it's back to being the normal size. -- RoySmith(talk)17:20, 9 April 2021 (UTC)
PrimeHunter, Hmm, yeah, that does sound like exactly what I'm seeing, right down to "often result in clicking an unwanted link which has now squirmed to under your finger". Just for fun, I put my desktop Chrome into mobile emulation mode, and couldn't reproduce it, not that I really expected it would. -- RoySmith(talk)21:20, 10 April 2021 (UTC)
You can set long press threshold to 0.3 seconds or so. The whole holding down a link then letting go then re-aiming then clicking without risk procedure doesn't have to take that long. Sagittarian Milky Way (talk) 22:06, 10 April 2021 (UTC)
Table of contents limit doesn't appear to work on QAnon page
A couple of times over the last few months, I've tried to set a Table of Contents limit (Template:TOC limit) on the QAnon page. There are currently 54 sections listed in the articles' ToC. Limiting it top-level headers only would bring the number down to a much more reasonable 15, making the page more readable for all users. However, each time I try to put the limit in place, it breaks something and the ToC either disappears, or the lead does. Does anyone know what's going wrong? Thanks. Ganesha811 (talk) 21:50, 10 April 2021 (UTC)
It looks like you inserted {{TOC limit|1}}, which does nothing. TOC level 1 is the page title. Level 2 are the main headings on the page, so {{TOC limit|2}} limits the TOC to a single level. A bit confusing, I know. – Jonesey95 (talk) 22:29, 10 April 2021 (UTC)
The universal template for sports events by month by country works fine for most countries (and is very useful!) but for the Philippines displays the category “2015 in Filipino sport” instead of “2015 in Philippine sport” (which has a temporary fix). Can this be fixed please? See Category:January 2015 sports events in the Philippines. Hugo999 (talk) 21:41, 7 April 2021 (UTC)
Hugo999 The problem is that the nationality isn't the same as the adjective for something pertaining to the country. Looks like a {{country2adjective}} template should be created to deal with these cases. I think Le Deluge may be intressted in dealing with this. --Trialpears (talk) 22:00, 7 April 2021 (UTC)
An adaptation of the (bicontinental) Russian (or Turkish) sports categories works fine for the Philippines (and does not require editing for specific months/years!); but other editors have to be aware that they cannot use the usual universal category for the Phillipines; see Category:January 2015 sports events in the Philippines. Hugo999 (talk) 11:12, 11 April 2021 (UTC)
This article, which I recently created, has rendered the article's title in italics. I believe it is because I have included an album infobox within the article, as per the guidance given here. Is there a way of getting the article's title back to normal type, without the disruption of deleting the album infobox altogether ? Thank you. - Derek R Bullamore (talk) 13:02, 11 April 2021 (UTC)
This is an issue in some skins (Timeless also has it occur) for images and certain responsive views. I don't understand what it is about the responsive views that causes it. (Presumably some 'upstream' CSS somewhere).
I think technically it could also occur in Minerva/mobile if we didn't have a line in MediaWiki:Mobile.css that prevents the width of images from shrinking beyond recognizability. Izno (talk) 20:10, 11 April 2021 (UTC)
Yeah intentional. As there is no good way to know the intended minimum size of an image inside a table. The best way is to set a min image size on the image via a template or something (if u want to keep it). Or to simply hide the image altogether if it’s decorative (like the mbox mobile styles in my narrow/wide screen gadget do). Author styling that handles the responsiveness. It’s just flagicons in tables are so prevalent that we deployed a fix for that. —TheDJ (talk • contribs) 06:49, 12 April 2021 (UTC)
Is there anyway of fool-proofing the notice so it won’t ever get deleted again? Perhaps make it smaller? More hidden comments? Move it to the page’s edit notice? --Heymid (contribs) 22:18, 5 April 2021 (UTC)
I know this. I haven't been living under a rock. So there’s no way of forcing the template to be there outside of putting it in the page’s wikitext along with everything else? A problem is that some users add the nobots template which disables bot editing of the page. And it’s not like the bot immediately restores the message once removed. --Heymid (contribs) 19:47, 6 April 2021 (UTC)
Edit notices are more and more irrelevant. They are not seen in mobile view because like navbox templates they provided little information and usually by one editor wishing to stop editing of a page.Moxy-22:38, 11 April 2021 (UTC)
@Moxy: I don't really agree with this. Editnotices should be rare - maybe 1/50 - 1/100 pages at most - but are very useful when done well. I feel like editnotices on talkpages should probably be used more than talkpage banners. Elli (talk | contribs) 22:58, 11 April 2021 (UTC)
...I phrased that quite poorly, I mean that they should often be used over talkpage banners - talkpage banners should still be more common. But people hardly ever saw {{Warning RS and OR}} for example, since you don't see talkpage banners when making edit requests. Whereas migrating that to an editnotice is far more effective. Elli (talk | contribs) 22:59, 11 April 2021 (UTC)
Would need to make then viable in mobile and the app etc. ...because as of now only 30% of readers....who may press edit will see them.Moxy-00:00, 12 April 2021 (UTC)
Is there an actual problem here? The sandbox is just for testing, and sometimes test edits remove the header...which is why we have a bot stick it back in. Legoktm (talk) 07:26, 12 April 2021 (UTC)
Problems accessing WMF sites
Hi - an editor I know and trust has contacted me by e-mail to tell me that she is unable to access WMF sites today - no commons, no EnWiki, no EsWiki. Not that she can't edit, but she can't connect at all, her browser is giving her "The server at en.wikipedia.org is taking too long to respond." She's checked with friends and family in the same locale with same ISP, they can get access, but despite multiple reboots, trying different browsers, even trying her husband's computer, she's not getting access. Other websites all work fine, just WMF sites won't load for some reason. Any thoughts about what to try? If it matters, she's in Mexico. Thanks in advance for any suggestions, which I will have to forward by e-mail (since she can't see this page!). GirthSummit (blether)18:15, 8 April 2021 (UTC)
SQL - she sent me a screenshot of what happens when she types 'en.wikipedia.org' into her address bar. The address bar changes to https://en.wikipedia.org/wiki/Main_Page, but she's just seeing the failure screen: The connection has timed out. The server at en.wikipeida.org took too long to respond, etc. Any more thoughts?
Volteer1 - I'm afraid that might be excessively technical for her. She's retired, a first-rate writer, researcher and overall contributor, but her technical skills are not too slick, she's daunted at the thought of attempting that, having had one look at what comes up when she Googles 'Google Public DNS' (I confess, I would be too). GirthSummit (blether)19:11, 8 April 2021 (UTC)
I would suggest looking in the javascript console and/or network tabs of the developer's tools, but from what you say, I'm guessing that's beyond her technical reach. -- RoySmith(talk)22:31, 8 April 2021 (UTC)
I woke up to an email telling me she's back on - whatever the problem was, it seems to have gone away (for now at least) - thanks all for your suggestions. GirthSummit (blether)05:43, 9 April 2021 (UTC)
There is a picture of the opening of the Stockton and Darlington Railway (File:Opening of the Stockton and Darlington Railway.jpg). There is also a cropped version (File:Opening of Stocking and Darlington Railway (crop).jpg) whose title is obviously mis-spelled (Stocking for Stockton). The latter is only linked once in en.wikipedia (from Stockton and Darlington Railway), but it is also linked from the es.-, ka.-, no.-, and zh.wikipedias.
Can I move this page to a correct spelling? Will this leave a redirect in place, and will that redirect work for the other wikipedias? I have never moved an article, and this is a file: rather than a mainspace article, so I don't want to make a mess through ignorance.
Also, how can I put a link to the file into this talk that shows the title (informative) rather than pulling in the picture (pretty, but not so germane)?--Verbarson (talk) 20:43, 10 April 2021 (UTC)
1. You don't have the requisite permissions to move the file, but you can request someone who does move to file by visiting the file's page on Commons (c:File:Opening of Stocking and Darlington Railway (crop).jpg) and clicking the "Move" link (which may be in a "More" menu). I've now done that for you. Once a Commons file mover processes my request, the move will leave a redirect, which will work on other Wikipedias. Nevertheless, it's customary for them to use a script to automatically replace all usages on all wikis, although I'm not sure why.