Wikipedia:Bots/Requests for approval/QxzBot 2

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section. The result of the discussion was

Withdrawn by operator.

QxzBot[edit]

Operator: Qxz

Automatic or Manually Assisted: Automatic

Programming Language(s): Visual Basic .NET

Function Summary: Discussion page indexing

Edit period(s): A handful of edits (exact number varying, but small) every few hours or so

Edit rate requested: At most 2 edits per minute, while editing

Already has a bot flag: No

Function Details:

Recently I've been working on various things related to indexing discussion pages – taking the contents of such a page, separating out the individual sections, posts, users, timestamps and so on, and turning it into a 'machine-readable' format, processing it, then reformatting it nicely for display purposes. An example of the sort of application this has can be seen on my talk page. This presents discussions in a bulletin-board style, giving a preview of each (making it concise yet more informative than a table of contents) and allowing them to be sorted by most recent comment or something else. This is a good way to summarize archives and make them easier to sift through, see e.g. User talk:Qxz/Archive 1. The idea can be extended by tacking extra bits on; at the bottom (scroll down) of User talk:Qxz/TestIndex until I delete it is a snapshot of the administrators' noticeboard, showing the result of an extra bit of parsing which picks up on the presence of {{resolved}} and marks sections as such – giving an at-a-glance view of what's done and not done. Hiding resolved issues completely is simply a matter of changing a template parameter.

Anyway, on to the bot request. First of all to make one thing clear, I only created this to see if it could be done. I'm not trying to force this upon anyone, and I'm not expecting to see it used on WP:AN or anything like that, that was just a test. Several people, though, have asked me for a similar set-up on their talk pages, generally not realizing that (except for archive pages) it becomes worthless after a while because it goes out of date. Even ignoring these pleas and pretending for a moment that other people don't exist, it would be nice to keep my own talk page's incarnation up to date. So, what I'm proposing is this: every few hours, a bot scans a list of talk pages and sees if any of them have been modified. Those that have been modified, it parses, and outputs a blob of confusing but machine-readable junk to another page, in my userspace. Templates elsewhere then feed upon this data and use it as required. The once-every-few-hours is a minimum; if a page gets no comments for two days then obviously it doesn't need updating. If multiple pages have been modified, obviously the corresponding indexes will all be updated in succession, since there's no rush to do this it won't make more than two edits per minute in these cases. So just to clarify, it will read whatever pages it's given but all edits will be made within my userspace. Is this acceptable? – Qxz 08:31, 19 March 2007 (UTC)[reply]

Discussion[edit]

It would be nice if this were a software feature, but as it is a bot seems like a good idea, and it should be relatively harmless. --ais523 18:18, 19 March 2007 (U T C)

The demonsration pages certainly look very impressive (nice to give the "illusion" of forums :)) - I have one technical question - you say that a machine readable blob of data is output, which templates then feed on. What is the appearance of this blob of data, and how do the templates use it? Thanks, Mart inp23 18:25, 19 March 2007 (UTC)[reply]

Right now the output looks like this. It essentially consists of calls to an IndexRow template with varying parameters, wrapped in some parserfunctions that can be used to conditionally return some rows but not others (thus generating different tables from the same index). The IndexRow template takes the parameters passed to it, and with the help of some more parserfunctions formats them into a table row. Transcluding IndexData thus returns a table without column headers or a closing |}; this is combined with a few much simpler templates that provide the column headers and the frame for the whole thing.

If this is too much templating, then I could modify it to simply output a complete table to the IndexData page instead. This would remove the ability to generate different tables from the same index, which would mean more edits were necessary, though not many. So kind of a trade-off – Qxz 20:13, 19 March 2007 (UTC)[reply]

Are you aware of the template limits? Your TestIndex says 'Pre-expand include size: 1035821 bytes' in it's HTML source code, which is more than the 1024000 bytes that the developers wanted to set it to before they were persuaded to increase it; so I'll suggest that you should simplify the template coding and/or precalculate it all so as to stay within the server usage limits that the developers wanted to set. --ais523 13:41, 20 March 2007 (U T C)
This does look nice, but wouldn't it be better to work with the developers to get it implemented in Mediawiki? --kingboyk 14:00, 20 March 2007 (UTC)[reply]

Apparently "This could be integrated into MediaWiki itself" is now a reason not to run a bot. Never mind that the developers' time is already taken up with much more important things and I have zero PHP programming experience. I don't quite see why this applies in this case but not to, say, category renaming bots, but never mind; this isn't an important enough feature for it to be worth my time arguing to get it approved. Withdrawing request – Qxz 01:07, 21 March 2007 (UTC)[reply]

Withdrawn by operator.. This bot was withdrawn by the operator. -- RM 12:25, 22 March 2007 (UTC)[reply]

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.