Talk:Time series database/Archive 1

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Databases[edit]

It would be good to get a list of known time series database systems, open source and proprietary. —Preceding unsigned comment added by 203.206.187.157 (talk) 23:47, 15 June 2010 (UTC)[reply]


Eventual Consistency is not a database, it is a property of many distributed databases. Xorlev (talk) 06:23, 18 September 2012 (UTC)[reply]


Why RavenDB made it into the list of time series databases? It's a document/json database, these aren't very efficient at storing time series. Further, RavenDB uses Lucene full text indexing, which is totally wrong tool for time series. Is it some marketing effort to push RavenDB everywhere? — Preceding unsigned comment added by 92.63.32.143 (talk) 09:57, 9 February 2014 (UTC)[reply]

Not sure why Informix is called out in the commercial relational database section. Why? Many commercial DBs support time series including Teradata, Vertica, Oracle, etc. Is this IBM spam? Virtualelvis (talk) 14:46, 12 October 2017 (UTC)[reply]

@Virtualelvis: You all miss the point of what we are writing here. Not a replacement for Google for sure, an encyclopedia yes. We also seem to forget what the article is about. —Dirk Beetstra T C 17:04, 12 October 2017 (UTC)[reply]
@Beetstra: Agree completely. I'm not sure the relevancy of Informix at all here. I'm not suggesting that we list all the commercial databases that support time series, as the list would be long as subject to spam. Rather, I'm questioning why Informix is more important than all of those DBs.

Hi there, I saw a link to this page on Reddit. I'm working with two TSDBs at work that should be on the list.

Additional references to IRONdb; ACM article https://queue.acm.org/detail.cfm?id=3178371, and ACM video https://dl.acm.org/citation.cfm?doid=2742580.2742803 Sremaing (talk) 22:43, 14 September 2018 (UTC)[reply]

@Sremaing: those are the same sources as before. Do you have any truely independent references for these? --Dirk Beetstra T C 08:55, 15 September 2018 (UTC)[reply]
@Beetstra: the Gorilla vldb paper was cited in this talk page already, yes. I also cited an Association of Computing Machinery reference for the IRONdb TSDB. Are you saying these sources are not independent? Sremaing (talk) 17:37, 15 September 2018 (UTC)[reply]
@Sremaing: @Beetstra: The Gorilla stuff has been posted either on the talk page or in edits something like 4-5 times. It's a little amazing that we still don't have that entry added after about a year - especially when you look at the other entries in the table. My favorite part about this talk page the debate about reliable/independent/notable references, yet the very first reference on the page isn't even a link, it's simply the name of a company: "Canary Labs". Just incredible. @Sremaing: prepare yourself for a slog if you plan to try and contribute here...— Preceding unsigned comment added by Kamelkev (talkcontribs)
We are not writing a journal article here. Have you actually read what we are writing here, and what we base our content on? Despite years of requests for independent (i.e. secondary) reliable sources, all we get presented are primary reliable sources. --Dirk Beetstra T C 18:03, 15 September 2018 (UTC)[reply]
@Sremaing: An article that states 'we aggressively leverage compression tech- niques such as delta-of-delta timestamps ...' does not sound independant. --Dirk Beetstra T C 18:10, 15 September 2018 (UTC)[reply]
And re the ACM article, it is written by Theo Schlossnagle, who is working for Circonus. The same Circonus he mentions in the article: 'such as the first-class histogram storage in Circonus's IRONdb'. That article is hence a primary, not independent reference.
What you all are saying here is that ONLY people working for Gorilla write articles about Gorilla, and ONLY people of IRONdb write about IRONdb. There are NO people who have absolutely no affiliation with either piece of software that have everwritten about it. There are NO companies outside Facebook who implement Gorilla, and NO other companies outside Circonus that haveimplemented IRONdb. If that is true, then these TDSBs are rightfully excluded, becausethe outside world, as you repeatedly seem to show, does not care about Gorilla and IRONdb. --Dirk Beetstra T C 18:40, 15 September 2018 (UTC)[reply]

I worked with FAME in the late 1990's. At the time, the FAME TSDB was used extensively by many Central Banks for econometric analysis. I believe it is still around. Does it warrant an inclusion? https://en.wikipedia.org/wiki/FAME_(database) Davagh (talk) 14:21, 5 March 2020 (UTC)[reply]

Blacklisted Links Found on Time series database[edit]

Cyberbot II has detected links on Time series database which have been added to the blacklist, either globally or locally. Links tend to be blacklisted because they have a history of being spammed or are highly inappropriate for Wikipedia. The addition will be logged at one of these locations: local or global If you believe the specific link should be exempt from the blacklist, you may request that it is white-listed. Alternatively, you may request that the link is removed from or altered on the blacklist locally or globally. When requesting whitelisting, be sure to supply the link to be whitelisted and wrap the link in nowiki tags. Please do not remove the tag until the issue is resolved. You may set the invisible parameter to "true" whilst requests to white-list are being processed. Should you require any help with this process, please ask at the help desk.

Below is a list of links that were found on the main page:

  • https://www.timeseries.guru
    Triggered by \bguru\b on the local blacklist

If you would like me to provide more information on the talk page, contact User:Cyberpower678 and ask him to program me with more info.

From your friendly hard working bot.—cyberbot IITalk to my owner:Online 01:02, 14 August 2015 (UTC)[reply]

Please revert changes after 719858484 and re-apply additions from IPs[edit]

Ushkin N (talk) 07:55, 26 May 2016 (UTC)[reply]

External Links[edit]

The time series database domain is evolving and is non-standardized. The vast majority of modern TSDB implementations have been released within the last 2-3 years, most of them are open-source and are lacking articles on Wikipedia describing the product or the company behind it. In particular, open-source databases often have no official sponsors and cannot benefit from implied notability by linking them to company pages. Having a list of external links to TSDBs is a valuable content as evidenced by 44 additions from a wide range of editors (signed in and anon) to external links section since January 1st, 2016 alone. The list contains some 20+ entries. The list has been repeatedly deleted (19 times in total) by one particular editor although there have 1 more mass delete revision by another editor. This section is an attempt to develop a community consensus around the external links section. Other than yes/no approach, the solution may be in creating a section containing a table of TSDBs along with some common criteria as in Graph_database, for example. Rodionos (talk) 13:14, 5 August 2016 (UTC)[reply]

  • Rodionos, I'm not sure if you're talking about HighKing or maybe Beetstra. I need to point you to WP:EL, I think. Also, I am not sure what we are discussing here--I thought you were taking issue with this. Finally, an author who writes up something for Wikipedia does not have ownership of the content and cannot unilaterally decide to remove said content if other editors wish to keep it. This is what it says under your editing window: "you irrevocably agree to release your contribution under the CC BY-SA 3.0 License and the GFDL". Drmies (talk) 14:21, 5 August 2016 (UTC)[reply]
    • Drmies, no this is a separate issue. See below. The issue under this paragraph refers to External Links which are being continuously added by a variety of editors and subsequently deleted as non-notable, not as WP:EL, (19 deletion in total) by editor HighKing.Rodionos (talk) 14:28, 5 August 2016 (UTC)[reply]
      • On my talk page you asked "could you please tell why an author is not allowed to remove the section of the page that he/she contributed in the first place", and you were edit warring over this--and you said you'd take it up on the article talk page. That's not the EL section. Drmies (talk) 14:36, 5 August 2016 (UTC)[reply]
        • Drmies See section 'TS transformation' below on the article talk page. There are multiple issues with this article and I believe it would be better to keep issues separate. — Preceding unsigned comment added by Rodionos (talkcontribs) 14:39, 5 August 2016 (UTC)[reply]
Can you list the multiple issues please? Best to deal with as many as we can while this page has some attention. -- HighKing++ 20:17, 5 August 2016 (UTC)[reply]
Drmies, HighKing As a consensus solution, I'd like to refactor External Links section into the List of time series database, similar to https://en.wikipedia.org/wiki/Graph_database#List_of_graph_databases. Columns: Name, License, Language, Description. Would this work for everyone? — Preceding unsigned comment added by Rodionos (talkcontribs) 05:49, 6 August 2016 (UTC)[reply]
That might work well but be aware that the criteria for inclusion in the list should still be notability. If the list is large enough it may make sense to have a separate article like the one you've referenced. -- HighKing++ 12:20, 6 August 2016 (UTC)[reply]
WP:NNC specifically states that "The notability guidelines do not apply to article or list content". If the list evolves into being notable enough to be a stand-alone article, it will be subject to WP:NOTESAL. The individual items in the list are subject to WP:DUE and overall content policies. With this in mind, I would like to seed the list with entries that are currently present in article and those that were contributed to External Links as of January 1, 2015 and later. Rodionos (talk) 10:11, 7 August 2016 (UTC)[reply]

@Rodionos: I am not sure whether you are talking about this edit of mine, where I state "clearout list, this should only contain either wikilinked items, or independently referenced items". You reverted that edit ("no wikipedia guidelines referenced during mass delete"; while I think that the explanation should give enough suggestion that there is a guideline, or even policy, behind it). Anyway, I re-reverted that list pointing to ".. see WP:EL, WP:NOT, etc. etc.". The specific items there are WP:NOTDIR and WP:ELLIST respectively. Taking the latter explanation: "However, the lists themselves should not be composed of external links. These lists are primarily intended as providing direct information and internal navigation, not a directory of sites on the web." These lists did provide just external links with minimal, unsourced, information (note that Graph_database#List_of_graph_databases is a bit better, though many are still unsourced and one can ask whether the item exists in the first place, or whether they belong in the list at all). The latter of those two problems is that these lists tend to be utter WP:SPAMHOLEs. For many cases also a dedicated hobbyist can write his own interpretation, make it available for download, and add it to these lists - and many do that. Some method of filtering is needed on these, and the best is that if it is not independently (and not only primary) sourceable, then it is not of real importance and it has no place on this list.

You are of course free to set up a table, and include the items that do meet our inclusion standards. And maybe there I am wrong, that all these items are 'notable' enough to be included, but the burden is on the person who wants to include to show that. --Dirk Beetstra T C 11:24, 7 August 2016 (UTC)[reply]

Dirk BeetstraFirst of all, a proper Wikipedia table with the following columns Database Name, License, Language, Description is not a list of external links. As such as, there a specific set of policies that have been developed by the community, and WP:DUE is most directly applicable. I agree that table should not be a WP:SPAMHOLE. I don't think a hobbyist can develop a database, although some open source systems are indeed developed by a small group of people, at least initially. In your closing remarks you appear to suggest that you will enforce notability at the item level. Please review WP:NNC - "The notability guidelines do not apply to article or list content". The standard applied to items is WP:DUE. I think it's not fair to other editors to introduce new policies such as burden of proof and notability for individual table records in spite of WP:NNC? To give you an example, OpenTSDB database was part of external links list. It will be a record in the "List of Time Series Database" table with 4 columns. It's not notable since it has no article in Wikipedia. The Graph Database list contains a lot of records like this. OpenTSDB should be on the list because this database exists and has been around for 3+ years. What do you think? Rodionos (talk) 12:12, 7 August 2016 (UTC)[reply]
I reverted what was merely a loos list of external links - in a table format that is better.
I agree that information does not necessarily need to be notable to be included in a notable article, however, if something is a verifiable and true does also not mean that it needs to be included. Read the opposite of your statement - you mean that the article needs to be notable, but that you can include all non-notable information in the article.
Regarding database applications - you say that the noted examples are not developed by hobbyists, but by established programmers. That also means that the items can show that they are worthy of inclusion. Again, I agree that information does not need to be notable to be included, but if one can show that it is (and everything included needs to be referenced by reliable sources anyway), there is nothing wrong with showing notability. Otherwise you do run the risk that on other subjects, and maybe even here, you will include information based on 'I checked their website and they say that they wrote a TS database application so I am including it here'). WP:N is a guideline, but it is based on our pillars/policies WP:NOT, WP:V etc. --Dirk Beetstra T C 12:59, 7 August 2016 (UTC)[reply]
Note that as the article stands now (practically completely unsourced), you will have trouble showing that the subject itself is notable ... --Dirk Beetstra T C 13:01, 7 August 2016 (UTC)[reply]
To get started, I converted the list to a sortable table with additional columns for license and language. Case-insensitive sorting applied as on Graph Databases page. Rodionos
We now have several entries in the list which are of two types: databases with links to wikipedia articles and databases that have no links to wikipedia articles. Linked databases are either linked to company name such as Basho and to database such as Kdb. What would be the right common criteria to determine notability of a database? Does the entry in the list (the database, not the company) has to be notable or should WP:NNC apply? In simple terms, should the entry be excluded from the list if it doesn't have a product or company article? If not, what other criteria of notability should be in place? Rodionos

TS transformations[edit]

I am the author of the section on TS transformations which I compiled as a list of capabilities implemented by databases referenced in External Links. The External Links section was deleted. The TS transformations is no longer consistent with the remaining references. It provides wrong information to the reader that these TS transformations are implemented by the notable TSDBs. I would like to delete this section because it longer applies.Rodionos (talk) 14:35, 5 August 2016 (UTC)[reply]

I'd support removing these example transformations too, especially as they are totally unsourced. ghytred talk 17:03, 5 August 2016 (UTC)[reply]
Rodionos, are you saying that the transformations are only applicable to a product that is not currently listed? Such as Axibase only? Are you saying that the transformations are not supported by the products listed in the "Example TSDB systems" section? If not, where did you get the list from? -- HighKing++ 20:37, 5 August 2016 (UTC)[reply]
Ghytred, re. unsourced. I basically compiled a master list of all possible transformations and reviewed API documentation for databases that are currently not listed in the article to cross-check what's implemented. FOSS entries were easiest to work with since their API docs are public and doesn't require registration. I have no idea to what extent this list is supported by remaining databases.Rodionos (talk) 05:43, 6 August 2016 (UTC)[reply]
Rodionos, can you tell us for which databases you reviewed the API documentation please. -- HighKing++ 12:16, 6 August 2016 (UTC)[reply]
@Rodionos: that list is totally unsourced, and the rest of the article is also practically unsourced. What I get of this (".. which I compiled as a list of ..") seems to be original research. Can you cite independent sources for this compilation? --Dirk Beetstra T C 11:31, 7 August 2016 (UTC)[reply]
Dirk BeetstraI am actually advocating that this section be deleted. So far we have 3 editors in favor of deleting the section on TS transformations. Rodionos (talk) 12:13, 7 August 2016 (UTC)[reply]
Well, if there are no independent sources for it, it should indeed go. --Dirk Beetstra T C 13:01, 7 August 2016 (UTC)[reply]
I concur. It appears that the list was Original Research and that there are no standards for time series operations and transformations. I suspect the list was really the transformations that could be applied by one product (Axibase). I'll delete the list. -- HighKing++ 15:51, 7 August 2016 (UTC)[reply]

Addition of Axibase[edit]

I note that Rodionos has created a standalone list of TSDB previously and has now added Axibase back into this article. As pointed out previously by the same editor, WP:LISTN states that the contents of the list are not generally governed by notability guidelines but that the list topic is considered notable if it has been documented in (reliable) sources for notability. (note: As a small aside, I have not been able to locate such a source - can anybody else?) In WP:LSC it states that Selection criteria for inclusion in lists should be ... supported by reliable sources. WP:CSC states that Lists are commonly written to satisfy either the criteria that Every entry meets the notability criteria or Every entry does not meet the notability criteria or Short complete lists of every item that is verifiably a member of the group. The second and third criteria do not apply, therefore the first criteria applies. Before deleting the Axibase entry and other entries that fail notability criteria, would any editor care to comment or make a case for inclusion/exception in the case of non-notable entries? -- HighKing++ 16:33, 24 August 2016 (UTC)[reply]

Please see a discussion on External List above. We already discussed the notability of the list and the items there. In support of my position, I would like to invoke WP:NNC that states that "The notability guidelines do not apply to article or list content". The inclusion should be neutral and the individual items in the list are subject to WP:DUE and overall content policies. The list is small, it's not a link farm, it doesn't contain external links. My affiliation with one of the entries in the list is disclosed on my user page. The list is similar to https://en.wikipedia.org/wiki/Graph_database#List_of_graph_databases — Preceding unsigned comment added by Rodionos (talkcontribs) 16:48, 24 August 2016 (UTC)[reply]
Hi Rodionos, your quote above from WP:NNC is not the entire sentence. The entire paragraph reads:
The criteria applied to article creation/retention are not the same as those applied to article content. The notability guidelines do not apply to article or list content (with the exception that some lists restrict inclusion to notable items or people). Content coverage within a given article or list (i.e. whether something is noteworthy enough to be mentioned in the article or list) is governed by the principle of due weight and other content policies. For additional information about list articles, see Notability and lists and Lead and selection criteria.
If you read the "additional information about list articles", you come to WP:LSC and WP:CSC as I've pointed out above. Finally, the problem with you being the CEO of Axibase (and therefore WP:COI, etc) is that it just looks like you are trying every trick and wiki-lawyering to get Axibase mentioned in the article. -- HighKing++ 21:25, 24 August 2016 (UTC)[reply]
I strongly disagree with your inflammatory language and the personal attacks. You went as far as suggesting that I'm a paid editor on my talk page. Can we have a civilized discussion on the right criteria for notability of rows in the database list. If you agree, please contribute to the base paragraph where the external list is discussed. — Preceding unsigned comment added by Rodionos (talkcontribs) 22:03, 24 August 2016 (UTC)[reply]
I think you'll find that if you actually read carefully what has been said, there is no inflammatory language and certainly no personal attacks. Feel free to point out examples of one or the other. Your assertion that I suggested you were a "paid editor" is untrue. On your Talk page I was referring to the fact the you have declared that you are the CEO of Axibase and therefore it is prety obvious that you have a financial interest in Axibase. And also obvious that it is in *your* financial interests to promote your company and your product. And therefore it is not unreasonable for me and other editors to wonder and/or question (and possibly conclude) that your multiple attempts to contribute to this article are all driven by the sole motivation of promoting your company and your product. Your declared conflict of interest and your promotion of Axibase go hand in hand on this issue. If you've nothing to add in terms of a response based on policy of guidelines then lets wait to see if another editor wants to provide their opinion and perhaps we can reach a community consensus. -- HighKing++ 13:55, 25 August 2016 (UTC)[reply]
If calling my references to Wikipedia policies "tricks and wiki-lawyering" isn't inflammatory, I'm not sure what else is. At any rate, I appreciate your discontinuing this practice. 1) My affiliation was public ahead of any edits. 2) I don't have financial interest from the edits. 3) I contributed all kind of databases to the list that have no company/product articles and yet should be included based on due weight policy. — Preceding unsigned comment added by Rodionos (talkcontribs) 14:15, 25 August 2016 (UTC)[reply]
I used the word "tricks" as a summary of your previous behaviour. For examply, your edits to include a list of TS transformations in the article - only for us to discover that they were not industry standard transformations, but actually a list of transformations of your product. Leaving aside the disruption caused by this section before this was made clear, I view your original intention to include these transformations as a "trick" in order to promote your product's capabilities under the guise of adding generic information to this article. I use the words "wiki-lawyering" to describe your selective quoting of only those parts of sentences within the policies and guidelines that suits your argument and your current attempts to justify your inclusion of Axibase in a list in the article.
1) You say that your affiliation was public ahead of any edits. That is not true. Your affiliation was only placed on your User page on August 10th 2016 - after most of your edits including all those relating to TS transformations and your initial attempts to create an Axibase article and to include mention of Axibase into this article.
2) You state that you don't have financial interest from the edits. Yet you are the CEO of Axibase and will profit from sales of your product? Which would explain a motivation on your part to use Wikipeda to promote your product, wouldn't you agree?
3) Yes you did - which is why I created this section to ask the community their opinion on whether we should allow this list to be a list of every product under the sun, notable or not, despite guidelines such as WP:CSC recommending the opposite).
Under the circumstances and given that we appear to disagree, we should request a third party opinion. Again. -- HighKing++ 17:20, 25 August 2016 (UTC)[reply]
Let me just say that what you claim is not true. My affiliation was public ahead of edits that we're discussing. List of TS transformations is not related to Axibase products. I have no financial interest in the edits. All entries that I added existed before you deleted them, so your claim that this is a "list of every product under the sun" is wrong. I quote selectively and I provide links to policies. You have no right to call this "tricks" and "wiki-lawyering". If you continue attacks on me including creating dimunitive summaries of my behavior or using those derogatory terms, I am going to report you to arbcom and you can get blocked for this misconduct. Rodionos
I got 1) wrong above. You did have your affiliation with Axibase public on your User page before you made any edits so I hold my hand up. I still do not understand how you can say that you have no financial interest in the edits. Is it not true that you have a financial interest in Axibase? Is it not true that promoting Axibase is part of your function and that you have a financial interest in sales of Axibase? -- HighKing++ 12:07, 26 August 2016 (UTC)[reply]

I have pruned the list down, and requested references for the remainder that they are 'notable' for their support of Time series. Items in the list should only be in this list if it is independently verifyable (per WP:V/WP:RS that they support time series. Otherwise, this is just a WP:SPAMHOLE. --Dirk Beetstra T C 12:55, 26 August 2016 (UTC)[reply]

DB-engines ranking[edit]

DB-engines ranking in the #List_of_time_series_databases section is not OK. DB-engines is a commercial platform that accepts money from the database vendor (500 Eur for blog post and 9000 Eur for promotion as a featured product). The list is also not very representative. InfluxDB is ranked #1, Graphite is #3, and Prometheus is #9. This can only happen because InfluxDB is a featured product on DB-engines but Graphite and Prometheus are not. It also contains a lot of things that nobody uses, e.g. SiteWhere and SiriDB.

I can recommend using this survey as a starting point for better list - http://www2.informatik.uni-stuttgart.de/cgi-bin/NCSTRL/NCSTRL_view.pl?id=INPROC-2017-06&mod=0&engl=0&inst=IPVS or maybe some other good survey from academia. DB-engines is not a credible source. It's not clear why some DB ranked X and another ranked Y. There is no methodology behind it. At least it's not published. In it's current state this page is a marketing material. — Preceding unsigned comment added by 109.124.67.234 (talk) 07:02, 20 June 2017 (UTC)[reply]

I agree that a commercial site such as DB-Engines does not even remotely constitute an unbiased authoritative source. That being said, this is a listing of all TSDB, not just open-source or free ones. Removing legitimate TSDB from the list as "spam links" without regard for the discussion and intent communicated on this page is not constructive. I have restored the previous entries in line with the above conversation. Moving forward we are trying to create a comprehensive list of such databases, not necessarily a list of "the best" or "most well known".

— Preceding unsigned comment added by kamelkev (talk) 22:14, 07 August 2017 (UTC)[reply]

And I have removed them again. As long as there are no reasonable independent references any database, commercial, open source or hobby would fulfill the criteria as 'it is a TDSB', and hence could be listed. The notability criteria for lists are lower than for articles themselves, still they need to show a reasonable reason as to why we should list them - that is why we ask for independent secondary sources. If they are not provided (either directly, or through an independent article), they should not be listed. --Dirk Beetstra T C 19:46, 16 August 2017 (UTC)[reply]

"still they need to show a reasonable reason as to why we should list them" - This is fabricated criteria. Entries need only be identifiable as time series databases, which all of these entries are. I reject that innovative well known time series databases are somehow ill-suited or inappropriate for listing. This debate on this particular article has been going on for quite some time. I look forward to this listing being audited due to the number of conflicting edits. — Preceding unsigned comment added by kamelkev (talk) 20:17, 16 August 2017 (UTC)[reply]

No, it is a WP:SPAMHOLE, anything there can be listed - our criteria are independent sourcing for information. --Dirk Beetstra T C 02:22, 17 August 2017 (UTC)[reply]

I have yet to see any spam entries create a problem here. Instead I see you and one other person repeatedly taking down legitimate TSDB that meet the necessary criteria. In the case that a non-TSDB were added I would of course agree with you, but that is not what is happening here. — Preceding unsigned comment added by kamelkev (talk) 20:23, 16 August 2017 (UTC)[reply]

They are continuously spammed in, so we discuss inclusion. Do you have independent sourcing for the items that show that they are in fact eligible for inclusion? --Dirk Beetstra T C 02:29, 17 August 2017 (UTC)[reply]
And your guide is here: Wikipedia:LISTCOMPANY. The three items listed do NOT have an independent source, and have not had that for a long time (longer than what is in the citation needed parameters). I guess we'll be back to the same discussion as always, I'll bring it to a relevant noticeboard. --Dirk Beetstra T C 02:42, 17 August 2017 (UTC)[reply]

Reinsertion of unreliable information[edit]

And we are back at the same edit war. The material was challenged for months for not having sources at all, now it gets reinserted with unreliable sources. User:Kamelkev, you have now been challenged over and ovver, now first get consensus for inclusion on talkpages. See WP:V for more info. --Dirk Beetstra T C 04:39, 9 October 2017 (UTC)[reply]

And for those who want to be sure that the sources are suitable to show the reliability, there is WP:RS/N. Note that material that is challenged, and is not reliably sourced can be removed. --Dirk Beetstra T C 04:51, 9 October 2017 (UTC)[reply]

User:Beetstra No, we are not back to "the same old edit war". Instead you have continued to act in bad faith by reverting relevant content from more contributors, along with removing citations as well.

It is clear from the edit history that you have removed multiple authoritative references to notable TSDB. The best example of this behavior relates to Facebook's Gorilla TSDB. We can all agree that Facebook is not only notable, but a worldwide innovator on a variety of technologies. Facebook presented their research and development of an extremely novel and proprietary database known as Gorilla in September of 2014 at the 41st International Conference on Very Large DataBases, hosted by VLDB Endowment. It signaled a ground shift in TSDB performance, and paved the way for a large number of open source projects. It is one of the most frequently cited modern TSDB papers. VLDB itself is well respected and peer reviewed - pretty much as good as it gets in terms of reliability for DBs in general.

Yet you removed the entry without comment, along with a reference to the associated journal article which has been cited elsewhere dozens of times. You are unapologetic about it. You don't attempt to work with the community, to communicate intent, or even attempt to improve the entry yourself. Instead you just revert. This is counterproductive, and goes directly against the philosophies that underly this community.

I am not the first person to complain about your behavior with respect to this article either. A number of other contributors have made similar observations regarding your behavior. It is difficult to tell if you are trolling, or are instead unfamiliar with TSDBs as a whole. In either case it would make the most sense for you to take a step back and ask yourself whether you are helping to create a good encyclopedia entry, or if you are stroking your ego.

--User:Kamelkev 14:30, 10 October 2017 (MT)

Look better, 3 of the 5 are still the same as before, including one without any references. We can discuss the suitability of the references for OpenTSDB and Gorilla .. guess where we are. --Dirk Beetstra T C 22:25, 10 October 2017 (UTC)[reply]
@Kamelkev: (forgot to ping) --Dirk Beetstra T C 22:25, 10 October 2017 (UTC)[reply]

I am citing Wikipedia:NNC which states that something does not have to be notable to be mentioned in an article, only to have an article created about it. Citing Wikipedia:SPAMHOLE as a reason why entries should be removed from the list of time series database table is a non sequiter, as that list is not the external links section of the article. I am in consensus with User:Kamelkev that 'Entries need only be identifiable as time series databases', and as such am recommending the list of TSDBs that have been removed be restored. Camel gopher (talk) 21:43, 12 October 2017 (UTC)[reply]

@Camel gopher: And I have already cited Wikipedia:LISTCOMPANY. The subject does not need to be notable, it needs independent reliable sources to show they are worth mentioning in a list. Second to this, do we realize what we are writing here? Third, I am not the only one who cleans this list, there are at least 3 others who do that. What is the subject we are writing about? Why is this list so much more important than anything that it needs to be an extensive WP:SPAMHOLE. I’ve already started to block editors who keepinserting material without consensus and edit war to do so, Next step will be protection, so we are enforced to discuss and get consensus before inserting. You are well beyond WP:BRD. —Dirk Beetstra T C 06:38, 13 October 2017 (UTC)[reply]

So what does exactly mean "worth" here? Is it enough to show the github README or blog post to show how it is being used and capable as a time series database? @Beetstra: User:Umitanuki

Can we get a guide on what is considered an independent, verifiable source for having another database added to the list? TimescaleDB was removed despite linking to an independent website (aiven.io) who is offering a time series service using TimescaleDB. What exactly is the criteria for getting listed? There are TechCrunch articles about the company raising money, there are blog posts of others using it as a time series database, in addition to the Aiven source which offers it as a service. @Beetstra: RobAtticus (talk) 22:40, 11 July 2018 (UTC)[reply]

@RobAtticus: WP:LISTCOMPANY and WP:V are your guide. You need to show that the subject is worth mentioning in this list. ‘Oh, we do time series as well’ is not an inclusion criterion, and blogs are not reliable sources (with exceptions), github is self-published. —Dirk Beetstra T C 06:44, 12 July 2018 (UTC)[reply]
@Beetstra: Does a TechCrunch article announcing their raise of money meet the requirements? It certainly seems notable that a cloud platform provider is using TimescaleDB as their time series solution and thus put it in a press release (that happened to be in a blog format). As a counter point, Riak TS is listed despite the company backing it going into receivership and no update to Riak in almost a year. Anyway here is the TechCrunch article: https://techcrunch.com/2018/01/24/timescale-raises-12-4m-to-build-a-new-breed-of-time-series-database-software/ RobAtticus (talk) 13:56, 12 July 2018 (UTC)[reply]
@RobAtticus: I have been tempted, and still am, to remove Riak. Though it has an article of themselves.
My criteria are basically that either it is notable on its own terms, or there is a reliable source that lists the subject as a worthy member in such a list. Being a TDSB on its own is not sufficient. I know and believe that TimeScaleDB supports TDSB, as do many other DB solutions. —Dirk Beetstra T C 14:30, 12 July 2018 (UTC)[reply]
@Beetstra: Can you answer whether the article is sufficient? The list is made up of some strange choices for time-series databases. I am not trying to waste my time by adding it if the article is not sufficient support for notability. Your comment about TimescaleDB makes it seem as if time-series is a secondary support mode for TimescaleDB; it is the entire purpose of the database. It is not merely a database that can do time-series in addition to other things (like Mongo or Cassandra), but a time-series database exclusively. RobAtticus (talk) 14:50, 12 July 2018 (UTC)[reply]
@RobAtticus: does the subject have an own article on en.wiki? Does the blog post show that the subject belongs in a list of TSDBs .. Does the blog post independently and reliably show that it is worthy in the list. I doubt that a blog post shows that. —Dirk Beetstra T C 14:57, 12 July 2018 (UTC)[reply]
@Beetstra: In the time you wrote that comment, you could have skimmed the news article (not blog post) and answered the question. TimescaleDB does not have it's own wiki article, but that is a new criteria you are introducing. I think what a lot of us who are trying to contribute to this article and list are wondering is what explicitly is the criteria that is being looked for, because right now it seems to be arbitrarily applied. To elaborate -- I think a company raising money to build a time series database that people are using (and is being offered as a cloud offering by a separate company) makes it a notable TSDB, regardless of having its own Wiki page. But it seems like the bar is higher than that, and yet low enough to accommodate Riak TS (defunct) and other questionable choices. RobAtticus (talk) 15:03, 12 July 2018 (UTC)[reply]
No, it is not a new criterion. For the rest, see our policies and guidelines. A blog post does not suffice. —Dirk Beetstra T C 19:14, 12 July 2018 (UTC)[reply]
Okay. To be clear, I've been asking about a new article, but you keep talking about a blog post. As we've reached an impasse I will just let someone else try. RobAtticus (talk) 19:41, 12 July 2018 (UTC)[reply]
Sorry for the confusion (my confusion). But what you seem to be talking about is notability of the subject itself. If the subject has a Wikipedia article then it warrants listing here. Listing it here based on the techcrunch article alone seems too thin. —Dirk Beetstra T C 04:04, 13 July 2018 (UTC)[reply]

Page edit lockdown and reversions of TSDB listings[edit]

Hello. I'm a newish wikipedia user who is a computer science student focusing on TSDBs. I notice that there have been a lot of reversions of edits to this page about well known time series databases. And then this page was put into a ″semi-protected″ state a few days ago.

There's a number of factual errors on the existing page. Listing Graphite as a TSDB is incorrect; Whisper is the actual TSDB in the Graphite ecosystem. It looks like that correction was applied over a year ago, but was reverted.

RRDTool is also listed as a TSDB, but that's not a TSDB; it's a file format.

Several actual TSDBs were added, but then reverted. FB's Gorilla in particular was removed despite being documented in academia.

I'm trying to get an understanding of what the continuing issue is with listings of TSDBs on this page and why they keep getting taken down. I'm working on a research project hopefully leading to a thesis opportunity and the removal of TSDB listings here seems really odd. — Preceding unsigned comment added by Guntherhust (talkcontribs) 22:14, 4 September 2018 (UTC)[reply]

Wow. Only took almost a year for us to finally correct the Whisper issue. Glad to see we are finally making progress.

With regard to the protection status - Can someone remind me why we're allowing one person to dictate how and when the page gets updated? In spite of the notability guidelines?

To restate from Wikipedia:Notability: "There is no present consensus for how to assess the notability of more complex and cross-categorization lists (such as "Lists of X of Y") or what other criteria may justify the notability of stand-alone lists, although non-encyclopedic cross-categorizations are touched upon in Wikipedia:What Wikipedia is not. Lists that fulfill recognized informational, navigation, or development purposes often are kept regardless of any demonstrated notability."

The page history shows clear and consistent pattern of edits that fly in the face of this, to the great detriment to the subject matter at hand. There are no less than 25 TSDBs that are regularly presenting at industry conferences in 2018, with only a couple represented in our list here. Seems to me our current direction is broken. — Preceding unsigned comment added by Kamelkev (talkcontribs) 03:19, 5 September 2018 (UTC)[reply]

@Kamelkev: There have been requests for suitable references for a long, long time now. Yet, people keep adding the same databases over and over without providing suitable references (or even with 'citation needed' tags). I fully agree that this list has errors, I fully agree that this list is not complete, but our policies and guidelines prescribe what should be included in Wikipedia. Your reference to WP:N is inappropriate as I have explained before, that is for the whole of articles. We have, however, other inclusion standards as well (your argument boils down to 'if the subject is notable, then we can mention whatever in that article, as WP:N does not apply to content'). If that were true, Viagra could include all spam in the world, because the notability criteria do not apply to content, and .
(and do note, a) I am not the only person keeping an eye on this page, there are several. And the same criteria apply to other lists just the same, and that is what I consistently stand for.
@Kamelkev and Guntherhust: As new editors come over and over again adding the same material without references, now we are enforced to discuss. I have been asking for references for over 2 years, and per our policies, anything that is questionable and not reliably sourced can be removed without explanation. Wikipedia is NOT a soapbox. And mine, and others' argument is, that the information is not worth mentioning, and until now no-one has shown why the information is worth mentioning (again, over 2 years, no-one has shown us that others are notable and/or worth mentioning). If none of you can come up with reliable sources to why a certain TDSB needs to be included, then it should not be here.
So if we now for each database either can show that it is notable on a Wikipedia level (i.e. it is blue-linked), AND/OR show that said database is included in an independent, reliable source comparing databases or sufficiently discussing it (the latter including the TSDB capabilities), AND/OR that a reliable source shows proper application of said database by a worth-mentioning third party (don't come to me with two school-projects where the second uses the first), then it can be included. Existence is NOT an inclusion standard. Lists (or the whole of Wikipedia) are NOT just random 'everything that fits goes in', see WP:NOT. Show that it belongs in the list, and everything that does not have reliable sources to show that it should be in the list, should go. It is now a very, very long time that we ask for sources, and no-one has been able to show what is worth mentioning in this list. --Dirk Beetstra T C 07:15, 5 September 2018 (UTC)[reply]
@Guntherhust: (earlier ping failed). --Dirk Beetstra T C 07:16, 5 September 2018 (UTC)[reply]
So I included a reference to Gorilla TSDB in my previous comment (hosted on Very Large Databases Endowment). I think we can all agree that's a reference right? Saying that no references have been provided is a false statement.
But there should be no list rather than the current one which contains incorrect entries. I know this is a bit of a niche space in computer science, but that doesn't change the need for correctness. I'm happy to provide domain expertise if needed. This isn't a list of hundreds of entries, so there isn't a need for notability for inclusion according to the wikipedia guidelines. If it is a TSDB, it should be in the list! That is definitely a consensus from reading through this talk page, and the edit history. It's troubling to see the page locked down against the implicit consensus of the rest of the users here who have made edits.
Guntherhust (talk) 19:24, 5 September 2018 (UTC)[reply]
Errors in the current list should be fixed of course (I already implemented your comment about Whisper, thank you for pointing this out). And if editors can provide an independent reliable source for a specific database, such databases should be added. By the way: I'd recommend to move your Gorilla-related suggestion to a separate dedicated thread if it needs more discussion - just to keep these aspects easier to handle. But there is no consensus to reduce the actual list inclusion criteria: niche topics don't have reduced source requirements, information in such topics needs to be sourced just like any other more common topic. If 25 TSDBs have been presented in industry conferences as mentioned above, there should be atleast some coverage from independent non-promotional sources to verify this information. In short: some reasonable evidence is needed, that an added database is more than a personal pet project from 1-2 developers in their garage or a COI link addition to promote a niche product. GermanJoe (talk) 20:23, 5 September 2018 (UTC)[reply]
If in consensus you also include the !votes of COI editors and similar, and ignore the global consensus of our policies, then yes, there is consensus to make this a spamhole. But that is not how consensus works. --Dirk Beetstra T C 00:06, 6 September 2018 (UTC)[reply]
And a paper that speaks consistently about 'we' is hardly an independent, reliable source. --Dirk Beetstra T C 00:12, 6 September 2018 (UTC)[reply]

So, above we are talking about this addition, repaired to this version. It contains:

OpenTSDB GPLv3+ Java [1][2][3]
IRONdb Commercial C/C++ [4][non-primary source needed]
DalmatinerDB Commercial Erlang [5][non-primary source needed]
TimescaleDB (PostgreSQL extension) Apache 2 PL/pgSQL  [citation needed]
Gorilla Proprietary [3][6]

References

  1. ^ "Comparison of Time Series Databases" (PDF).
  2. ^ "Survey and Comparison of Open Source Time Series Databases" (PDF).
  3. ^ a b "Gorilla: A Fast, Scalable, In-Memory Time Series Database" (PDF).
  4. ^ "Introducing IRONdb".
  5. ^ "DalmatinerDB For Linux".
  6. ^ "Gorilla: A fast, scalable, in-memory time series database".

(taken out the ones with an own article).

Again, we are not questioning whether these are TSDBs, I believe that. We are questioning whether the references show that they belong in the table. TimeScaleDB is re-added without reference, hoping that it will be forthcoming while that has not come for years before. Refs 5 and 6 are (self serving / primary) blogs. Ref 3 is a very much self-serving primary document, 'we' all over. Certainly not independent. And reference 4 is also a self-published document (the domain redirects to irondb website).  References 1 and 2 are university published documents, not independently reviewed material.  Most of this material is self-serving material, not even published by an independent department of Facebook, or a completely independent source. And that someone in a university (and not the university itself) is using it is hardly a significant use.

I can agree that Gorilla seems appropriate, but there must be decent sources for this. But re-inserting material with vague blogs or even with a {{citation needed}} tag does not make a cause for inclusion. --Dirk Beetstra T C 10:51, 6 September 2018 (UTC)[reply]

Its a little frustrating to see this comment. I've seen multiple attempts to add Gorilla here, all with the same result - reversion and receipt of a snarky comment. It'd be fine if there was substantive conversation resulting in improvement, but from what I've seen that's not been the case.
I personally attempted to add Gorilla to this list, citing a conference journal. This contribution was reverted. I followed up on the talk page about this specific entry, and received more snarky commentary about why the content couldn't be added. Afterwards I more or less gave up attempting to improve this page, as the engagement here was far less productive than what I've seen elsewhere.
As noted prior, you're not going to find conventional "notable" sourcing for many significant TSDBs that the average person likely expects to find on a list like this. Instead you're likely find academic papers, journal articles, conference presentations, etc - as is the case with Gorilla. For this reason setting a high bar for list inclusion does the article a disservice, as that direction will ultimately exclude significant distributed TSDBs, operated by the largest companies in the world, that are less than well-known. At least in my case this is the exact type of software I had hoped to find when I originally visited this page. — Preceding unsigned comment added by Kamelkev (talkcontribs) 20:50, 11 September 2018 (UTC)[reply]
@Kamelkev: That is a blog post. --Dirk Beetstra T C 03:22, 12 September 2018 (UTC)[reply]
@Beetstra: would Time Series Management Systems: A Survey meet the requirement of a reliable AND independent source to establish the validity of including Gorilla in this list? It is published in the IEEE Transactions on Knowledge and Data Engineering, which is a peer reviewed journal. I believe this establishes it as a reliable source according to Wikipedia:Reliable sources. The authors appear to be related to Aalborg University, and the professor involved seems to have no (disclosed) financial dealings with timeseries companies. I believe this establishes it as independent according to Wikipedia:Identifying_and_using_independent_sources. The survey paper references Gorilla multiple times, I believe thus satisfying WP:LISTCOMPANY. Would citing this paper and the original publication of the Gorilla be acceptable for its inclusion into the list, or have I misinterpreted one of the above Wikipedia guidelines? Linearizable (talk) 07:39, 12 November 2018 (UTC)[reply]
@Linearizable: That sounds indeed like it is both reliable and independent (all authors are from the university, paper is 'financed' through a Danish innovation fund, etc. It appears that also others in the list on this page are mentioned there? I would however link to the original at https://ieeexplore.ieee.org/document/8012550 / http://dx.doi.org/10.1109/TKDE.2017.2740932, not (only) the arxiv version. --Dirk Beetstra T C 10:07, 12 November 2018 (UTC)[reply]

RfC: Does the list of TSDBs on this page require citations for entries?[edit]

The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion. A summary of the conclusions reached follows.
Issue rendered moot by independent reliable source. Guntherhust (talk) 17:14, 4 October 2018 (UTC)[reply]

Should entries in the TSDB list on this page require robust citations for inclusion, or is list notability relative to the subject matter? Guntherhust (talk) 21:42, 13 September 2018 (UTC)[reply]

Survey[edit]

  • Support, TSDB list entries should require a citation to simply be listed, as this prevents abuse. Guntherhust (talk) 21:42, 13 September 2018 (UTC)[reply]
  • Oppose, TSDBs are an emerging technology which may not always have the most robust citations, despite significant usage in industry or academia. Guntherhust (talk) 21:42, 13 September 2018 (UTC)[reply]
    • 'despite significant usage in industry andacademia' .. if that is true, there is independent usage, and therefore there are independent sources. --Dirk Beetstra T C 18:05, 15 September 2018 (UTC)[reply]
  • All lists on Wikipedia need some bar of what should be included, something that takes the item beyond a hobby project. Wikipedia does that through independent sourcing requirements. This list, like anylist, is not different fromt that. That the field of TDSB is emerging does not excuse inclusion of everything that is a TDSB. Either accept that ALL need sufficient sourcing, or accept that this list will fill up with spam (the latter being true for all lists that do not have any sourcing requirements applied). --Dirk Beetstra T C 22:22, 13 September 2018 (UTC)[reply]
    • So basically, I oppose the way items have been added to this list, the material needs to have independent (i.e. secondary, i.e. not material written by anyone who is related to the subject, even if it is published and peer reviewed) sourcing, as anything else that is in this encyclopedia and challenged (where here the challenge is whether the outside world, i.e. outside of the people developing, creating, building the subject have independently reported about the subject). We apply the same criteria everywhere (including other lists). See WP:V, see WP:RS. That has been asked for for 2 years, an no-one comes with it, always the same primary sourcing. --Dirk Beetstra T C 12:21, 18 September 2018 (UTC)[reply]
  • I oppose the way that the list content has been managed, and think we should try something different. The last year shows various attempts to add legitimate entries to the list of TSDBs (openTSDB, TimeScaleDB, IRONdb, Gorilla, etc), with and without conventional citations. These attempts have all been reverted, in near real time, preventing development of the subject matter. The current list now poorly represent the current TSDB landscape, and ironically includes some databases which are not TSDBs. The existing criteria is broken. I think we should remove all the entries in the current list, and then impose criteria that all new entries are a) purpose-built and b) have either one secondary source, or multiple primary sources. I think this criteria succinctly describes entries that a) are in-line with wikipedia policy and b) will help us move past this impasse and start improving this article. edited to refine my opinion further Kamelkev (talk) 23:33, 17 September 2018 (UTC)[reply]
  • Wikipedia is no free PR platform to popularize new "emerging" technologies or less-known products, but an encyclopedic project. The usage of independent reliable sources to build articles and lists is non-negotiable for an encyclopedia to provide accurate, relevant and unbiased information. I support a minimum requirement of atleast 1 independent reliable source with some coverage (not blogs, PR fluff or standard listings), but oppose any further weakening of our list standards. Also, disabling these standards against current guidelines would probably need broader discussion in a central forum (per WP:LOCALCONSENSUS). GermanJoe (talk) 17:37, 14 September 2018 (UTC)[reply]
  • Oppose, most of the entries in the TSDB list on this page do not have citations already, so why require them now? Noting a citation is needed is sufficient with a small list like this (and will likely lead to those citations being added).— Preceding unsigned comment added by Sremaing (talkcontribs)
    • All items either have references, or have them in their own article. Citations have been requested for 2 years, and (independant AND reliable) sources are not forthcoming. --Dirk Beetstra T C 21:47, 17 September 2018 (UTC)[reply]
  • I oppose how this page is administered. The vast majority of list additions are reverted by one particular super-user, and therefore it's fair to say that this page is not managed by a community anymore. The super-user who rejects all the edits under the protection of notability guidelines which are self-interpreted and imposed to require a link to an existing page on wikipedia. This essentially makes wikipedia a walled garden. For all practical purposes, the most up-to-date list is not available at https://db-engines.com/en/ranking/time+series+dbms.— Preceding unsigned comment added by 93.185.27.207 (talkcontribs)
  • Oppose. It's a list of "notable database systems". Items on the list should of course be notable, and non-notable items have quite properly been removed. There's no need for the items to be explicitly referenced in the list as well as being notable. Maproom (talk) 07:17, 20 September 2018 (UTC)[reply]
  • Support, ver WP:V if the DB in question is not described as TSDB in the lede of its wikipedia artcile. For example from article about eXtremeDB it is not a all evident is it optimized for TS handling.   Staszek Lem (talk) 18:16, 24 September 2018 (UTC)[reply]
  • Support, again per WP:V. No sources, no mention. Policies don't apply less strongly where good sourcing isn't available. Wikipedia's job isn't to summarise to describe the current bleeding edge of technology, it's to summarise all available reliable sources. If stuff is so obscure that no reliable sources mentioned it, then it doesn't warrant inclusion; if it's so new that no sources have written about it yet, then we wait until they have. GirthSummit (blether) 06:41, 26 September 2018 (UTC)[reply]

Threaded discussion[edit]

What do you mean with 'list notability'? That is what we requested all along. Can you show that X should be mentioned in the list through independent sources talking about it? If no-one outside the subject talks about it, it should not be in the list. Just like anything else on Wikipedia,per WP:V and, inthis case, WP:LISTCOMPANY. Note, your question should be whether a source shows sufficiently that an item belongs inthis list, not whether this article should follow global policies ornot.' And note that those sources exist (though thin), but that above discussions adhere too much too onlyprimary sources and blogs. --Dirk Beetstra T C 22:09, 13 September 2018 (UTC)[reply]

  • Direct editing appears no longer possible here? The same user who has rejected 90%+ of the contributions throughout the last year has “protected” the page. Claims page has become “WP:SPAMHOLE”. Other contributors, myself included, have referenced WP:NNC, noting that the subject matter is highly dynamic, and not always citable outside of websites such as highscalability or InfoQ. Guntherhust (talk) 18:54, 14 September 2018 (UTC)[reply]
    • That there are no other sources is no reason to corrode our policies and guidelines. It is simple, if there are no independent sources, then it is not worth mentioning. NNC has not much to say here, LISTCOMPANIES does (though that basically is a form of notability question). We have asked for years for independent sources, they do not come. Instead, the material is inserted over and over with the same sources, generally ignoring concerns. To avoid that the same material is again insertedwithout decent sourcing, we are now enforced to discuss. But we are still left with the same, no independent sourcing. --Dirk Beetstra T C 20:22, 14 September 2018 (UTC)[reply]
      • "That there are no other sources is no reason to corrode our policies and guidelines." This misrepresents prior dialog here. There ARE sources for several TSDBs, some discussed on this very page. Some discussed with you specifically. The information is simply ignored. To repeat myself again: Gorilla, from Facebook - Presented in VLDB in 2015. Source provided prior, but here it is again: https://www.vldb.org/pvldb/vol8/p1816-teller.pdf. This paper is cited by over 50 other TSDB papers, it's essentially the foundation for all TSDB innovation since 2015, is present in a discussion with you on this very page, and is evidenced here: https://scholar.google.com/scholar?cites=6833513786602939732&as_sdt=4005&sciodt=0,6&hl=en. Discussion around notability does seem relevant, as historical behavior here seems to violate generally accepted best practice. I think our best path forward is to open this page up for edits, requiring at least one single reasonably rigorous source, and to start removing entries on this page that aren't purpose built TSDB. I'd go one step further that in proposing we state criteria on this page explicitly declaring what is needed to be included: "one reasonably rigorous source demonstrating that the proposed entry is a purpose-built TSDB". That definition seemingly resolves all debate relating to every other proposed entry for the last two years. — Preceding unsigned comment added by Kamelkev (talkcontribs) 03:06, 15 September 2018 (UTC)[reply]
        • @Kamelkev: The requirements are simply ignored by you. what has been presented here are not reliable, and/or are not independant! 'To improve query efficiency, we aggressively leverage compression tech- niques ...',you see the 'we'? That is reliable, not independant. Those are papers you build hypothesis and academic research on.
        • Now you come with a google scholar search that shows 55 results. Most of those are reliable sources, most of those are truely independant. That are the sources we base an encylopedia on. I would now chose one or two who discuss the use of Gorilla in most detail. --Dirk Beetstra T C 04:42, 15 September 2018 (UTC)[reply]
          • @Beetstra: I'm glad you have clarified prestigious peer-reviewed journal articles and conference presentations from VLDB and ACM are "not reliable, and/or are not independant." The RfC seems appropriate given that stance. It's also nice to see you finally read the google scholar search which was provided to you on 10 October 2017, almost a full year ago. I'm not sure why you keep pretending that these materials haven't been provided for discussion before. They have been part of many edits, and have been discussed on this page. Any claims that "Now you come with a google scholar search" is a misrepresentation of the truth. Kamelkev (talk) 17:34, 15 September 2018 (UTC)[reply]
            • @Kamelkev: 1) the operative word being 'or'. It is not independent. 2) the Google scholar search is not a reference, and though the search was presented, no specific suitable references were presented, and certainly never used in the article. --Dirk Beetstra T C 17:47, 15 September 2018 (UTC)[reply]
            • @Kamelkev: Note that I have offered to discuss suitability of these references. None of them has ever been discussed seperately as particularly suiting (and being both independant AND reliable). --Dirk Beetstra T C 17:53, 15 September 2018 (UTC)[reply]
            • A source written by someone from Gorilla about their own product is not independent from the source, even if it is published by a third party. What we are asking for are articles written by other people than those who have a demonstrated link with the subject (so, for e.g. IRONdb, anyone who is NOT a (former) employee from Circonus writing about IRONdb). --Dirk Beetstra T C 18:50, 15 September 2018 (UTC)[reply]

Just a quick note that User:Robert_McClenon and User:Redrose64 kindly corrected my errors with the rfc tag, and that we should hopefully be seeing additional input from other editors. With regards to the votes submitted so far regarding this RfC, users ave provided opinions in the Survey section, but some have not stated 'support' or 'oppose'. Please kindly update your response to clarify your position on this matter. Thank you.— Preceding unsigned comment added by Guntherhust (talkcontribs)

You ask a should we do this or that, and expect a support or oppose? That does not make sense. And remember, it is a survey, not a strawpoll. --Dirk Beetstra T C 17:25, 17 September 2018 (UTC)[reply]
@Beetstra: please see the separate votes from discussion format which I have used for this Rfc. A vote of 'support' or 'oppose' is expected. There has been significant variation in the responses given for what criteria are required to list a TSDB in this list (even varying significantly within a single user), so I have created this RfC to have a straightforward discussion to reach a consensus on the question of 'Does the list of TSDBs on this page require citations for entries?'. That seems to be the crux of conflict on this page. Also, I would request that instead of replying to my thread on the Survey section, that you respond in Threaded Discussion, so that we can keep the discussion close to the guidelines suggested in the RfC example I referred to, rather than creating an ad hoc format. Guntherhust (talk) 22:50, 17 September 2018 (UTC)[reply]
@Guntherhust: Yes, but your question is not in an oppose/support form. Moreover, it will result in a local consensus that does not trump global consensus. ALL of Wikipedia must comply with the same sourcing policy. If something is not sufficiently sourced, it should be removed. Especially since we have been pointed to independent (and hopefully reliable) sources. --Dirk Beetstra T C 03:23, 18 September 2018 (UTC)[reply]
@Beetstra: would you please condense your multiple threads in the Survey section to a single response as shown at the format we are adhering to, and clarify your support/oppose response to the RfC question of 'Does the list of TSDBs on this page require citations for entries?'? You answered 'oppose', but that appears to be in direct contradiction of the content you posted along with the vote. All other users who have posted in Survey have only posted once in accordance with the format we are adhering to, I would kindly ask that you act as a team player so that we may reach a proper consensus on this issue. Thank you. Guntherhust (talk) 18:28, 20 September 2018 (UTC)[reply]
@Guntherhust: Again, you have not asked a oppose/support question, you ask a regular question. What you ask is to override global consensus with a local consensus. What I answer is fine, it is not a straw poll. --Dirk Beetstra T C 04:53, 21 September 2018 (UTC)[reply]

Currently the list of TSDBs has 7 entries; 4 of those don't have citations. One (eXtremeDB) has an 'unreliable source', and Informix has a reference by IBM (the creator of Informix). Meanwhile, several well cited TSDBs (Gorilla, Irondb, Timescale) have been removed from that list and the page has been 'protected against vandalism'. None of this makes any sense. Allow the list to be edited and note 'citation required' on any entry that needs one.— Preceding unsigned comment added by Sremaing (talkcontribs)

@Sremaing: Those items are linked to existing articles. For the first item it links to an article with 25 dependant and independant reliable sources and the article establishes that it is a member of the list.
ALL the other items that have been added are redlinked. Unless they have independant (i.e. secondary) and reliable sources (like everything else in Wikipedia) we can not establish whether it warrants inclusion in the list. --Dirk Beetstra T C 21:56, 17 September 2018 (UTC)[reply]
  • (RFC-Bot-summoned). Comments (1) The discussed list has a vague description: "provide support for time series data". Heck, my Lotus 1-2-3 spreadsheet already "supported" the time series of my expenses. I can imagine some marketroid can claim it supports time series just because you can have a field of "timestap" which allows addition of substraction of "time" format (eXtremeDB on the very top this list raises this kind of red flag for me). Therefore we have to have independent RS citeted that a particfular database has a significant TSDB functionality. (2) the lede is bloated with techical detail. Is thoud be really made in the summary of the article, and at least thta last two paragraphs mus be moved into appropriate sections. Staszek Lem (talk) 18:16, 24 September 2018 (UTC)[reply]
@Staszek: agreed that the description of the list has a description in conflict with the title 'List of time series databases'. eXtremeDB is a column oriented store, and RRDTool is not a TSDB either. I would think that removing the phrase 'The following notable database systems provide support for time series data:' would clarify the intent of the list as to only represent actual purpose time series databases. Regarding your mention that 'independent RS' need to be cited - several of those have been provided for TSDBs that have been rejected from that list prior to this page being 'protected against vandalism'. My interpretation of criteria for page content of this nature is that reliable sources need be cited per due weight, but that the items themselves need not be notable. To date, peer reviewed professional publications and peer reviewed scholarly articles have been rejected as unreliable, leading to this RfC. Guntherhust (talk) 19:13, 26 September 2018 (UTC)[reply]
@Guntherhust: they have NOT been rejected as unreliable. They are reliable sources. They have been rejected as NOT BEING INDEPENDENT. It is not a logical OR, is a logical AND. The published material has to be reliable AND independent. These sources are all NOT independent, they are primary sources. The remarks in this section state "I'm glad you have clarified prestigious peer-reviewed journal articles and conference presentations from VLDB and ACM are "not reliable, and/or are not independant." has been the whole point throughout. I answered that the operative word was 'OR', but that still did not get that message through: we need sources that are both independent AND reliable.
To clarify:
  • The ACM article, it is written by Theo Schlossnagle, who is working for Circonus. The same Circonus he mentions in the article: 'such as the first-class histogram storage in Circonus's IRONdb'. That article is hence a primary, and not an independent reference, for IRONdb. It IS reliable, but that alone is NOT enough. It is 'reliable OR independent', not 'reliable AND independent'
  • The VLDB article states "'we aggressively leverage compression techniques such as delta-of-delta timestamps", it is not an independent reference for Gorilla. It IS reliable, but that alone is NOT enough.  It is 'reliable OR independent', not 'reliable AND independent'
We have been pointed to the existence of independent, reliable sources (through a search), but there is NO discussion of sources that are both independent (secondary) AND reliable to show their suitability, nor attempts to use such as sources for inclusion in the article. The only thing we get, over and over, are dependent (primary), though reliable, sources. But dependent (primary) sources are NOT a reason to include material into this list. Again, we believe that these are TSDBs, we do NOT dispute that, but it is up to the editors who want to include these that independent (i.e. secondary) reliable sources show that. And whatever you think to get out of the RfC above does not trump that global consensus, see WP:V, WP:RS, WP:LISTCOMPANY. Stop with your misinterpretation of comments, and start to read and understand what we have been asking for the last 2 years.  --Dirk Beetstra T C 06:54, 27 September 2018 (UTC)[reply]
And by the way, removing the word 'notable' from the list-section does not notably change the situation - one could argue now that only TDSBs with an own article could be in the list, after removing the ones without an own article (which are hence not notable) still need independent, reliable sources. --Dirk Beetstra T C 10:37, 27 September 2018 (UTC)[reply]

Will you stop with this nitpicking when the article has bigger problems?[edit]

With the exception of this silly table the article HAS EXACTLY ZERO REFERENCES!!! All this time wasted on bickering about notability of particular bussinessess could have nad spent on article referencing and improving. Staszek Lem (talk) 20:33, 27 September 2018 (UTC)[reply]

That is because the only edits to the page over the last 3-4 years have been to include TDSBs (diff over three years, note that two paragraphs swapped position. NONE of the specialists here are interested to improve the text, only adding non-notable entries in the table. And that will continue, also after this useless, local consensus RfC). Anyway I would be worried about the sourcing that we then would get.
Apparently it is of utmost importance to have TDSBs in the list. --Dirk Beetstra T C 02:58, 28 September 2018 (UTC)(expanded Dirk Beetstra T C 03:37, 28 September 2018 (UTC))[reply]
The edits to this page have been limited because it has been continued to be policed in an authoritarian fashion as opposed to a collaborative effort. As has been stated in the threaded discussion of the RfC, several users are unhappy with the way that this page is being administered. Users who post relevant content to it are subject to immediate reversion, are often bullied by the administrator, and have been banned on occasion. It is no surprise that users are either scared to post, or feel that it is not worth wasting their time.
Yesterday I posted two pieces of independent content on time series databases, and they were promptly taken down. The first was a picture of a license plate referencing time series databases. The description on the image states 'TSDB California license plate. This is my license plate which stands for time-series database.'. Pretty cool, right? And definitely relevant.
TSDB License Plate
The second was a market analysis by the RedMonk developer analyst firm, The State of the Time Series Database Market. Another relevant source. People want to contribute to this page. Content like this begets more content of high quality, but as is evident, this page is not really open to contributions as it stands. — Preceding unsigned comment added by Guntherhust (talkcontribs) 17:07, 3 October 2018 (UTC)[reply]
@Guntherhust: that image is, by no stretch, a reliable source for anything, and you are by no means independent of that document. We are not writing a cool page here. Are you here to contribute facts supported by independent reliable sources?
The other is a (largely) independent document, but by no means used as a source. If you would have updated the list properly according to the source, I would not have reverted you.
This page is indeed not open to contributions until editors start properly sourcing the material. --Dirk Beetstra T C 18:50, 3 October 2018 (UTC)[reply]
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Redmonk review[edit]

This seems like a good independent reference for expanding the document. --Dirk Beetstra T C 04:12, 3 October 2018 (UTC)[reply]

I'll update this thread with the background for this entry. Yesterday I edited the page to add The State of Time Series Database Market. You can see that version of the page here. User:Beetstra reverted my change shortly after (and also removed another independent TSDB reference) without discussion, and created this talk page entry. The content for this page is still being policed, even for independent reliable sources.
This link is an independent reliable source which could be used as a citation for every item that has been removed from the list of databases which support time series data. Though as User:Staszek indicated previously, several of those entries are not TSDBs; I could claim that MS Notepad belongs there since it can persist time series data. I would recommend updating the description of the list to be only TSDBs, and add each TSDB referenced in the RedMonk document. Guntherhust (talk) 16:55, 3 October 2018 (UTC)[reply]
@Guntherhust: that's what I thought (according to the disclaimer not completely independent, but way better than what we have seen in the last years). That was not how it was used, can you please use this as a proper (wikipedia style) reference and update the list? --Dirk Beetstra T C 17:54, 3 October 2018 (UTC)[reply]
By the way, I did NOT 'also remove[d] another independent TSDB reference', I removed an inline external link, not anything used as a reference. You may want to familiarize yourself with WP:BRD. --Dirk Beetstra T C 18:42, 3 October 2018 (UTC)[reply]

@MrOllie: you seem to challenge this article as an independent reference to meet WP:LSC/WP:LISTCOMPANY (and I can agree that it is thin - it certainly is not enough to counter the claim by user:82.81.136.242 that "Timescale is a top player", and claim by user:Kigelim that "TimeScale and OpenTSDB are major timeseries dbs"). I guess it is time that we re-discuss this situation. --Dirk Beetstra T C 09:06, 11 August 2020 (UTC)[reply]

Adapted above post. Properly ping user:Kigelim who made a similar claim, also in a thread below. --Dirk Beetstra T C 09:10, 11 August 2020 (UTC)[reply]

This 'review' is essentially an attempt to establish relevance by counting social media likes (in this case github 'stars'). It doesn't cover any of the software in any real sense. - MrOllie (talk) 10:40, 11 August 2020 (UTC)[reply]
I have started another thread below. I agree this 'review' is not a great source and the current edit even remove DBs which are included in it. Kigelim (talk) 12:35, 11 August 2020 (UTC)[reply]

A question regarding links[edit]

Hey all, so when I was first learning about time series, I read this explanation from InfluxData, which I thought was a great introductory explanation, and which is still one of the best I've seen (though, that is personal opinion). I took a look at this page recently and saw the note at the top saying additional citations should be added, so, I figured it would be good to pass on this page. However, shortly after I added the link, it was taken down. Can someone more experienced in Wikipedia help me understand why, so that I don't make this mistake again? Thank you! Sukai Cerny (talk) 21:39, 16 July 2019 (UTC)Sukai Cerny[reply]

New TSDBs are being erased...[edit]

And there is no logical reason. This table is currently based on a very specific list as if no other sources are acceptable. One account doing the deletions had about 10 edits when it got involved. The phrase COI is thrown too quickly which raises a question, why is that the first assumption in complete contrast to WP:AGF. As soon as a new database gets some coverage, there is no reason to leave it out. If anyone suggests calling the list "Prominent TSDB", then the whole list needs to be reevaluated. Kigelim (talk) 17:59, 25 August 2019 (UTC)[reply]

Forbes contributor articles are usually not considered reliable sources (WP:RSP). See additional, more general info on your user talkpage (although the article-related aspects are best discussed here if needed). All lists on Wikipedia are implicitly focussing on notable or at least noteworthy information (see also WP:CSC about list inclusion criteria). Minor or secondary details are routinely omitted both in articles and lists. Lists of products, companies, organizations, and people are not meant to be fully comprehensive. GermanJoe (talk) 19:13, 25 August 2019 (UTC)[reply]
You set the bar too high IMO. I ran into CrateDB and thought "why not?". While I agree it isn't the strongest RS, it isn't quantum physics is referenced therefore the bar is lower. Whether Forbes info is from the developers or not, isn't relevant. It establishes it is a TSDB. Period. Kigelim (talk) 05:25, 26 August 2019 (UTC)[reply]
No, the bar is not lower. The bar is high to exclude spam, as we do on all lists in articles. Lower the bar and anything goes, including the hobby project of some retired programmer. (However, in this case, I should note that CrateDB seems notable in itself - then it might warrant inclusion if we can reliably establish whether it has significant TSDB features; on the other hand, these lists are not meant to exhaustively include all, just to provide the more important examples per WP:NOTDIRECTORY). --Dirk Beetstra T C 06:16, 26 August 2019 (UTC)[reply]
I accept this but the list right now is based on the "State of the Time Series Database Market" post that states "The below view of competitors in the time series space is a general survey, but importantly it is not an exhaustive list of every project". Isn't a company that made into a Forbes list as TSDB worthy of mentioning here? I think this is excessive. Kigelim (talk) 06:38, 26 August 2019 (UTC)[reply]
As I already stated, both companies seem to be notable in itself in which case the references do not need to necessarily show notability of the company, but just that they are a TSDB to a reasonable extent. --Dirk Beetstra T C 07:15, 26 August 2019 (UTC)[reply]

@Kigelim: comparison by another company does not fulfill our notability guidelines. Can you please remove Graphite again (also, it does not have an own article, you are pointing to the wrong subject). --Dirk Beetstra T C 12:21, 26 August 2019 (UTC)[reply]

Also, Graphite is NOT a TSDB. --Dirk Beetstra T C 12:22, 26 August 2019 (UTC)[reply]
@Beetstra: Yes, it is. [1] [2] The first isn't an RS but proves a point. The second is an RS. Kigelim (talk) 12:30, 26 August 2019 (UTC)[reply]
The latter has been shot down already earlier, as the fact that Graphite is a TSDB (and that is again proven by your first reference). Kigelim, this article has long time been protected against this type of editing as people kept adding stuff without first discussing, and without first supplying reliable sources. You are now trying to re-introduce that linkfarm and try to edge the reliability of the references. If your data was removed, keep it removed and discuss it here and do not re-insert the material again until consensus is reached. This is NOT a platform to promote TSDBs, we are writing an encyclopedia here. --Dirk Beetstra T C 12:46, 26 August 2019 (UTC)[reply]
@Beetstra: while I think the bar set is too high, I can see your point. But RedisTimeSeries was just removed and it appears in a book. Does anyone claim it is not an WP:RS? — Preceding unsigned comment added by Kigelim (talkcontribs)
Have you checked the release date? Especially vs. the date of the book? --Dirk Beetstra T C 09:02, 27 August 2019 (UTC)[reply]
@Beetstra: No. Would you enlighten me? Kigelim (talk) 17:34, 27 August 2019 (UTC)[reply]

I did not get any explanation of how is being mentioned in a book is not considered WP:RS. Does it mean everyone is in agreement? Kigelim (talk) 08:44, 29 August 2019 (UTC)[reply]

No, that is not necessarily the case. You have been reverted 3 times on that reference. The BRD cycle requires a discussion then to come to a consensus of inclusion. If the bold edit would have stood that would have been a consensus by silence, and equally if you first would have suggested it here and no-one would have objected. Now you have a revert, and a remark that I made regarding the inclusion.
The book is older than the announcement of the (rather recent) release of RedisTimeSeries. I fail to see how that book supports the inclusion of RedisTimeSeries.
Please, don't start this edit war again. We've had this page protected for a long time due to pushing of minority databases (hobby/school/university projects), primary sourced material and other items that were disputed. That something is a TSDB is not a reason to include it. (and actually, seen the last couple of months, this should be done already. Items on the list are NOT supported by the references, items are continuously introduced without coming to consensus after being removed). --Dirk Beetstra T C 09:17, 29 August 2019 (UTC)[reply]
There is no edit war. The editor who reverted it 3 times has 10 edits and gave creative reasons to delete it. A book can mention a product in its beta stage. This just highlights the fact it is notable. This isn't a highschool project and the source is as good as the one which supports all the other one. Kigelim (talk) 13:17, 29 August 2019 (UTC)[reply]
And there it is: you are the one to decide that that source is good enough. We deteriorate our sources, and below situation happens. The spamhole is open again. --Dirk Beetstra T C 03:37, 1 September 2019 (UTC)[reply]

apparently we are not harsh enough. WP:SPAMHOLE in action. Clear out this list. --Dirk Beetstra T C 19:11, 29 August 2019 (UTC)[reply]

That is not an answer. Since there is a book which mentions RedisTimeSeries and no one suggested anything about why it cannot be used, I am reinstating it. Kigelim (talk) 09:06, 31 August 2019 (UTC)[reply]
If it wasn't clear that was a new point of discussion to make clear what is happening to this page, then I don't know what is. Kigelim, you do not have consensus for inclusion, stop your edit war and revert yourself. --Dirk Beetstra T C 03:33, 1 September 2019 (UTC)[reply]
All the recent reverts were done by a user with 10 edits with surprisinly great knowledge about Wikipedia. A clear sign for a secondery account. GermanJoe didn't remove it when he removed another item So there is no war, just reinsertion after vandalism. If you feel strongly enough about removal, please do and we can keep this conversation going. Your only point was "I don't like this book" which isn't a great point IMHO. Kigelim (talk) 08:01, 4 September 2019 (UTC)[reply]
Who said that I didn't like the book. I just noted that the proper handling of time series was just introduced recently while the book is of an earlier time, and there are no independent (secondary), reliable sources from after the official release.
I am not going to discount an editor who is here maybe only for 10 edits. They may very well have made an account after hundreds of edits as an IP. They may just be someone who happened to come here new but has a wealth of knowledge regarding the subject. It is basically an ad hominim argument.
Point is, that the addition was challenged, and I am not convinced either. I think it is a deterioration of our inclusion standards already (which was promptly followed by another addition lower down the road. And I still insists that this is not supposed to be a full list of all possible TSDBs, we are not writing a linkfarm. Where is the notability besides being mentioned in a book. Where are the independent reviews? Why does RedisTimeSeries warrant to be listed (and no, being mentioned in a book is not necessarily a reason to be listed). And I note that the article starts to turn into a linkfarm: people are not here to write about TSDBs in an encyclopedic way, people are here to expand the table of TSDBs. I strongly suggests to cut down the list again to a bare minimum and protect the article. Per user:GermanJoe, this list is NOT supposed to be comprehensive and certainly not start to dwarf the article of the subject. We are not writing a directory service, we are not writing a promotional piece of text. --Dirk Beetstra T C 08:46, 4 September 2019 (UTC)[reply]
I'd rather stay out of a prolonged discussion (for now) - repeating the same viewpoints over and over is rarely helpful. Generally speaking, I agree with Beetstra's view that such lists should be handled restrictively. A lenient approach for any topic with COI-related issues (products, companies, organizations, celebrities) just leads to spam and we absolutely need to draw a line somewhere. Redis in particular is a borderline case: the sources have their flaws and the book is only a brief mention mostly focussing on an interface, not the DB itself. But at least there is some coverage. Frankly, I am not sure about Redis - consider me "neutral" in this specific case. GermanJoe (talk) 09:27, 4 September 2019 (UTC)[reply]
To just quote myself from just less than a year ago: "Apparently it is of utmost importance to have TDSBs in the list." (and consider the general thought in Talk:Time_series_database#Will_you_stop_with_this_nitpicking_when_the_article_has_bigger_problems? ..). RedisTimeSeries is new (released less than three months ago), and, but a fleeting mention in a book that was published now just a year ago (and the first beta of the publication of RedisTimeSeries is about 9 months ago!), not independently mentioned. Why is Redis so important that it needs to be in this list.
As usual, with all lists, the referencing is a slippery slope. There are some properly referenced (which was already a fight in itself), and slowly the next references become less strict and boom, you get the spammy ones back in (as we now just saw). I feel it is high time that we clear out this list again, and lock the article down so that we have a chance to discuss how to improve the article and keep the references to the high quality that they should be. --Dirk Beetstra T C 10:40, 4 September 2019 (UTC)[reply]

If you feel strongly enough a book cannot be used as a source, take it out. But the current list is based on one article that doesn’t claim to be comprehensive. The 10 edits editor who made it his mission to take Redis out and do nothing else on Wikipedia is not a benchmark for what should or shouldn’t be included. Thank you for the respectful discussion. Kigelim (talk) 06:21, 7 September 2019 (UTC)[reply]

Redis has for long been a notable database, but for the timeseries, released months ago, the book is not a suitable source, and the other references given (one for sure primary) make it rather thin for whether Redis earns a place yet in this list. I am sure that that situation will change in the next months and better sources will be published. --Dirk Beetstra T C 06:36, 7 September 2019 (UTC)[reply]

For future reference, since I am sure someone else will try this, the following URL contains screenshots of pages 350 and 351: [3]. Not that it much matters given the nature of the content, but the ISBN for the citation doesn't appear to line up with the intended reference by the poster. In any case, looks like we've achieved consensus here and are moving on. — Preceding unsigned comment added by Gaimrox (talkcontribs) 23:26, 12 September 2019 (UTC)[reply]

Once again you revert first and then write. Two other editors who participated in this conversation didn't remove it. A complaint was filed for edit-warring. Kigelim (talk) 06:11, 16 September 2019 (UTC)[reply]
@Kigelim: But that is the normal cycle: you inserted it, it was reverted and then you first discuss before re-adding it, see WP:BRD. --Dirk Beetstra T C 08:45, 16 September 2019 (UTC)[reply]

Removing databases from list w/o proper explanation[edit]

Many DBs were removed w/o any explanation. Seems like the standard was whether or not a wiki page exists for them. TimescaleDB is a top player and this question has been raised before. I believe we should either revert the last change or remove the list altogether. Kigelim (talk) 12:30, 11 August 2020 (UTC)[reply]

WP:WTAF is not a reason for removal of entry. Just remove the squared brackets. Kigelim (talk) 12:41, 11 August 2020 (UTC)[reply]

Kigelim I have been asking for independent references for a long, long time. I realized the Redmonks was not very strong, but at least it was an independent article talking about the different databases. But in the end, it is a counting of 'likes'. See this.

Removing the squared brackets does not make an item notable. See WP:LISTCOMPANY and WP:LSC. Existence is not a reason for inclusion into a list, topical relevance is. You again repeat 'TimescaleDB is a top player' (as was used by an IP), can you show us that it is a top player? Who is using it? Who is talking about it? If it is a 'Top player' many people must be talking about it. --Dirk Beetstra T C 13:25, 11 August 2020 (UTC)[reply]

Completely agree that existence is not a reason for inclusion. Maybe this or this can be considered. Kigelim (talk) 10:25, 12 August 2020 (UTC)[reply]
Kigelim, seriously, another count of 'internet likes' by DB-engines (and user:Kamelkev above defined it as "a commercial site such as DB-Engines does not even remotely constitute an unbiased authoritative source"), and a website that is generally considered a blog (see WP:RSP, entry for Medium)? And based on that you want to define that TimescaleDB is a 'Top player' (it was last year number 12 on that list of db-engines, now 10 ...). Nothing better? Dirk Beetstra T C 13:27, 12 August 2020 (UTC)[reply]
Beetstra, how about this paper? Kigelim (talk) 15:47, 16 August 2020 (UTC)[reply]
Kigelim, an analysis of public information in Studies in Informatics and Control (note the redlink), including regurgitation of public information , quotes: "The selection of the solutions to be analysed is based on DB-Engines Ranking of Time Series DBMS (https://db-engines.com/en/ranking/time+series+dbms). The most popular TSDB, according to their ranking from February 2019, have been chosen." and "For this we analysed several technical resources available online – GitHub (https://github.com), OpenHub (https://www.openhub.net/), StackOverFlow (https://stackoverflow.com) and Google Trends (https://trends.google.com)." That still does not make TimescaleDB 'a top player'. This is a bit more formal than db-engines, but basically the same: internet likes. Dirk Beetstra T C 05:11, 17 August 2020 (UTC)[reply]

I have protected the page again. Again the focus is that list of TSDBs, not the prose. Being in that list is the most important part of this page. Now please come with references or write the articles, it is clear that we are not getting anywhere otherwise. --Dirk Beetstra T C 04:14, 21 August 2020 (UTC)[reply]

The prose was terrible!! The list was the only thing redeemable on this crap article. Then you über editors are more concerned with exact Reference citations than you are with knowledgeable content. Perhaps if your could review and approve the draft articles with the same speed the you do with a simple missing reference, we would be in a better place. You WikiNazi's win, everybody else who comes for information and knowledge on different Time Series databases loses. Good work! MrFollie (talk) 20:22, 21 August 2020 (UTC)[reply]
MrFollie, I have looked at draft:TimescaleDB. 3 sources, the github and 2 of the product' website. Nothing independent, full of weasel statements. No mention of notability. It is the same problem as here. People exclaim that it is a top player but cannot show anything for that. I will leave it for independent review, but I am sure it is going to be declined in its current form. Dirk Beetstra T C 09:00, 23 August 2020 (UTC)[reply]
And regarding your 'The prose was terrible!!', but no-one here is trying to upgrade that, while that should be the focus if the prose is crap. --Dirk Beetstra T C 09:04, 23 August 2020 (UTC)[reply]

RfC on inclusion criteria[edit]

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.


This article has a list of time series databases. What should the main criterion for inclusion in this list be?

  • Option 1: Databases should have their own Wikipedia article to be included.
  • Option 2: Databases should be notable to be included.
  • Option 3: Databases should have coverage in at least one reliable secondary source.

Chess (talk) (please use {{ping|Chess}} on reply) 22:06, 31 August 2020 (UTC)[reply]

Addendum for newer editors[edit]

For the benefit of newer editors who are participating a lot in discussions here, our notability policy (and links to others) is at WP:NOTABILITY and our definition of reliable sources is at WP:Reliable sources. The main criteria for notability of databases would likely be the general notability criteria, which states that "significant coverage in reliable sources that are independent of the subject", which usually entails secondary sources, a definition of which is available at WP:SECONDARY. Chess (talk) (please use {{ping|Chess}} on reply) 22:06, 31 August 2020 (UTC)[reply]

Poll[edit]

  • Option 2 A good compromise that can serve as inspiration for genuinely notable new articles. (Summoned by bot) --I dream of horses (Contribs) Please notify me after replying off my talk page. Thank you. 00:28, 1 September 2020 (UTC)[reply]
  • Option 21, 3. As is our general inclusion standard (WP:LSC/WP:LISTCOMPANY), the subject has to reliably show topical relevance, not just plain existence. That means blue-linked entities where the wikipedia article reasonably shows that time series are a significant part of the operation of the software (option 21), and red-linked entities that have been mentioned in independent, reliable reviews of the software (option 3). As the latter has not been provided (and seems not available), this currently boils down to 21. --Dirk Beetstra T C 04:05, 1 September 2020 (UTC) (renumbered and some rewrite --Dirk Beetstra T C 07:51, 1 September 2020 (UTC)).[reply]
    • Option 2 does not require a Wikipedia article. There are many notable things that do not have Wikipedia articles. Anything with a Wikipedia article should have its notability demonstarted in that article. Meters (talk) 07:34, 1 September 2020 (UTC)[reply]
      Meters, I now notice that as well. I may rephrase my comment. Dirk Beetstra T C 07:43, 1 September 2020 (UTC)[reply]
      I do want to stress that I still do agree with the inclusion of redlinked items as long as references are given that a) show that something is a TDSB (which can be primary), AND b) they show 'topical relevance', i.e. that the outside world shows that it is a TDSB worth using (beyond having xxx github likes). Topical relevance is shown by references that are reliable AND are independent. --Dirk Beetstra T C 07:51, 1 September 2020 (UTC)[reply]
  • Opton 1, with the further restriction that the Wikipedia article should confirm that the system is a TSDB. This is an article about TSDBs, not a list of TSDBs. It should not be the duty of maintainers of this article to check whether each listed item could qualify as "notable" when it is not in fact "notable" in the usual sense. Maproom (talk) 07:10, 1 September 2020 (UTC)[reply]
  • Option 1, with reliable sources showing that the software is a TSDB (presumably in the software's article but in this article if for some reason it's needed). This should be a list of notable examples (in this case of TSDB), and we normally restrict such lists to items that have Wikipedia articles. Meters (talk) 07:34, 1 September 2020 (UTC)[reply]
    • Per WP:CSC it is acceptable to have ref-linked entries if the entry is a member of the group and it is reasonable to expect an article could be forthcoming in the future. In my experience, this is a recipe for disaster unless there are clear-cut rules to determine presumed notability for that group (e.g., mayor of any city with a population greater than X, any building on such and such a list of historic buildings, any NHL player, any recording with more then Y sales). Without such hard-coded rules (which do not and never will exist in this case) any red-linked entries become a potential point of disagreement over whether the supplied sources are sufficient to show that the entry would qualify for an article if one were written.And that does ot even touch on whether the article will ever actually get written.Meters (talk) 07:54, 1 September 2020 (UTC)[reply]
  • Option 1 - option 2 and 3 would be reasonable alternatives and I don't oppose these options (as long as Wikipedia's guidelines are used to define "notability", significant coverage and reliability). But option 1 is the easiest to maintain and avoids endless disputes about borderline cases. If a suggested entry is "notable" in Wikipedia's sense of the term, a stub with 2-3 reliable sources would be sufficient for inclusion. GermanJoe (talk) 10:07, 1 September 2020 (UTC)[reply]
  • Option 1 - Easiest to maintain and prevents potential arguments about notability outside the proper venues, which really should be AFC and/or AFD. - MrOllie (talk) 13:26, 1 September 2020 (UTC)[reply]

Discussion[edit]

For what it's worth, I think there should be a second criterion apart from the criteria listed above that any databases included should be characterized as a time series database by a reliable secondary source to prevent databases that aren't time series databases from being included on the list. I don't believe this is controversial and I didn't include it in the RfC for that reason, but we can have a second discussion about this if someone objects. I'm starting this formal RfC as an uninvolved editor because it's clear discussion has stalled on this topic and there should be more input. Chess (talk) (please use {{ping|Chess}} on reply) 22:06, 31 August 2020 (UTC)[reply]

  • I have asked for closure of this RfC on WP:ANRFC as there are several edit requests below that can only be reasonably judged by the outcome of this RfC. --Dirk Beetstra T C 05:22, 29 December 2020 (UTC)[reply]
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Confusing use of term "database"[edit]

The term "database" is used in two ways, as illustrated by the sentences "This company maintains a database of flavour molecules. It keeps it in an Oracle database". The article starts by preferring the former sense, using the phrase "software system" for what one might also call a "database engine". But the currently controversial section with the list is titled "List of time series databases". It would be less confusing if it were titled "List of time series database engines". Maproom (talk) 07:23, 1 September 2020 (UTC)[reply]

Add Apache Pinot to List of Time Series Databases[edit]

Request to add the following entry to the List of time series databases.

Name: Apache Pinot (https://pinot.apache.org/)
License: Apache License 2.0 (https://en.wikipedia.org/wiki/Apache_License#Version_2.0)
Language: Java (https://en.wikipedia.org/wiki/Java_(programming_language))
References: Included in the Apache Pinot article page
 Not done: See two sections up where inclusion criteria for this list were established. There is no Wikipedia article on Pinot, so it should not be listed. - MrOllie (talk) 00:03, 14 October 2020 (UTC)[reply]
@Nehapawar18: I suggest that you create Draft:Pinot (database) through WP:AFC and have it independently reviewed. If it makes it into a Wikipedia article then please come back here to request addition. --Dirk Beetstra T C 05:44, 31 December 2020 (UTC)[reply]
Updating this request now that there is an approved Apache Pinot article page. User:ForlornacornTalk 21:37, 14 July 2021 (UTC)[reply]
 Not done: According to the page's protection level you should be able to edit the page yourself. If you seem to be unable to, please reopen the request with further details. ScottishFinnishRadish (talk) 22:49, 14 July 2021 (UTC)[reply]
Thank you, User:ScottishFinnishRadish. You're correct. I've gone ahead and made the edit and updated reference sources. User:ForlornacornTalk 23:05, 14 July 2021 (UTC)[reply]
Forlornacorn, congrats on getting your article accepted too. Happy editing! ScottishFinnishRadish (talk) 00:32, 15 July 2021 (UTC)[reply]

Extended-confirmed-protected edit request on 6 November 2020[edit]

Please add this entry to the table listing existing Time-Series databases:

Name License Language Reference M3 Apache 2.0 Go Article Pi.meson.hadron (talk) 04:36, 6 November 2020 (UTC)[reply]

 Not done: See a few sections up where inclusion criteria for this list were established. There is no Wikipedia article, so this should not be listed. - MrOllie (talk) 00:03, 14 October 2020 (UTC)[reply]
@Pi.meson.hadron: I suggest that you create Draft:M3db through WP:AFC and have it independently reviewed. If it makes it into a Wikipedia article then please come back here. --Dirk Beetstra T C 05:45, 31 December 2020 (UTC)[reply]

Add RedisTimeSeries to List of time series databases[edit]

RedisTimeSeries is now offered on Azure which shows it's notability[1]. It does not have a Wiki page but Redis does. Two of the other databases on the list do not have independent page, Riak-TS and Whisper. 62.219.236.90 (talk) 06:06, 30 November 2020 (UTC)[reply]

@62.219.236.90: as has been discussed above, the article Redis has to discuss that the timeseries capabilities of Redis are noteworthy enough.
Riak and Whisper might need to go, those references are not sufficient. And I still insist, this article is about what a Time series database is, the focus is not a list of pieces of software that supports Time series. --Dirk Beetstra T C 06:49, 30 November 2020 (UTC)[reply]
Might need to go? Then let them go. I would have, but the article is protected.
Or remove the list altogether.
Not including Timescale or OpenTSDB in here make it hard toWP:AGF. Here is a list that should be considered more than the article currently used. [2] 2A02:14C:182:DE00:381A:8F8B:C911:FCDE (talk) 19:08, 15 December 2020 (UTC)[reply]
2A02:14C:182:DE00:381A:8F8B:C911:FCDE, that reference is just a listing of likes. Write articles for the two, showing they are notable and they will be included. Dirk Beetstra T C 04:12, 16 December 2020 (UTC)[reply]

Extended-confirmed-protected edit request on 17 December 2020[edit]

The list of time series software on this page does not reflect the current state of this space. It is very limited and only comes from a single source document which is highly opinionated itself.

Given the dynamics in the TSDB space it is urgent to update this list with a more informative set of options so people do not get incomplete information regarding the plurality of existing software.

Among the software missing from this list which have a rather important traction nowadays one might find:

@Mathias.Herberts: See Talk:Time_series_database#RfC_on_inclusion_criteria, although (still) not closed, the agreement seems to be that only databases with an own Wikipedia article, and which can show that they support time series will be included in the list. That means that the current list is likely going to be even shorter. That is very much in line with (or even slightly more restrictive than) our policies/guidelines, spec. WP:LSC and WP:LISTCOMPANY. --Dirk Beetstra T C 09:21, 17 December 2020 (UTC)[reply]
@Mathias.Herberts: Reconfirming, both QuestDB and Warp 10 (note that the bluelink for the latter does not go to an article about the database) do not have an own article, are hence not notable in Wikipedia terms, and hence will not be included in the article. I suggest to create Draft:QuestDB and Draft:Warp 10 (database) through WP:AFC. --Dirk Beetstra T C 05:37, 31 December 2020 (UTC)[reply]

Using VentureBeat as a reliable source for TSDB[edit]

VentureBeat just published an article reviewing the TSDB market. Before updating the data, can it be considered WP:RS? — Preceding unsigned comment added by Kigelim (talkcontribs) 07:29, 17 January 2021 (UTC)[reply]

Kigelim, what do you want to update? The list? For that, each item now needs an own article, where that article establishes that it is a time series database. See RfC somewhere higher on this page. Dirk Beetstra T C 07:42, 17 January 2021 (UTC)[reply]
In the discussions linked on WP:RSP regarding VentureBeat it is generally considered a reliable 'news blog'. --Dirk Beetstra T C 08:10, 17 January 2021 (UTC)[reply]

6 months[edit]

It is now 6 months ago since the closure of above RfC. This is what was improved in 19 edits. The only edits done are the continuous pushing of non-notable tsdbs to the list. No-one takes it to talk. I have protected for a year. — Dirk Beetstra T C 20:52, 12 June 2021 (UTC)[reply]

@Beetstra: Given how infrequently the article is edited, would you consider pending changes protection instead? Chess (talk) (please use {{reply to|Chess}} on reply) 20:51, 6 July 2021 (UTC)[reply]
@Chess: No, the only edits performed over the last 6 months were 'spamming'/pushing non-notable time series databases. I would only consider that if the article would actually have had actual content edits (or even just one). I do not see any use of pending changes if we anyway have to revert every edit then anyway. Dirk Beetstra T C 05:26, 7 July 2021 (UTC)[reply]
@Beetstra: I went ahead and made some improvements to the article. Would this justify PC at some point? Chess (talk) (please use {{reply to|Chess}} on reply) 06:20, 7 July 2021 (UTC)[reply]
@Chess: No - you can just make the changes anyway, and autoconfirmed is a pretty low bar. Pending changes would only be justified if there are several edits by editors that would (over the last 2-3 years in times it was not protected) that would not have been possible because it was protected. This is what changed in 6 months of not having it protected until recently - NOTHING. No edits by new editors who even did some copy-edit that would now have been disallowed. And the situation in the previous time of being unprotected is not dissimilar: 160 edits of which 2-3 rewrite of prose (by regulars: [4]) and ~75 edits with 75 reverts of additions of non-notable databases. The last edit that was performed by someone who would be blocked by protection was here, 2 years ago. And there were 0 edit requests regarding things people wanted to change to the text in the times that it was protected (though that is not the best of reference, but seen that there are edit requests to add non-notable TDSBs does suggest that editors who manage not to edit do know what they have to do). --Dirk Beetstra T C 09:02, 7 July 2021 (UTC)[reply]