Wikipedia:Bots/Requests for approval/ReferenceBot

The following discussion is an archived debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA. The result of the discussion was

Approved.

ReferenceBot

Operator: A930913 (talk · contribs · SUL · edit count · logs · page moves · block log · rights log · ANI search)

Time filed: 22:10, Thursday November 7, 2013 (UTC)

Automatic, Supervised, or Manual: Automatic

Programming language(s): Python (pywikibot for editing)

Source code available: On request

Function overview: BracketBot/DPL bot for references.

Links to relevant discussions (where appropriate): Wikipedia:Bot_requests#New_REFBot

Edit period(s): Daily

Estimated number of pages affected: 20-50 user talk pages/day

Exclusion compliant (Yes/No): Yes (pywikibot)

Already has a bot flag (Yes/No): No

Function details: Daily, check the categories, Pages with broken reference names, Pages with incorrect ref formatting and Pages with missing references list, for new pages, and messages the user that placed it there, asking them to fix it. Notification style like DPL bot, using User:ReferenceBot/inform/top, User:ReferenceBot/inform/middle * n and User:ReferenceBot/inform/bottom. (See the relevant discussion link for examples.)

Discussion

How does the bot make sure it is not recent vandalism that caused the page to break? — HELLKNOWZ ▎TALK 22:37, 7 November 2013 (UTC)[reply]

It will be run daily, so all [run of the mill reverted quickly] vandalism will have been reverted by the time the bot comes round. Also, previous hashes are compared, so reverted vandalism won't trigger the bot. 930913(Congratulate) 00:50, 8 November 2013 (UTC)[reply]

Approved for trial (30 edits, but at least one day with as many as they come). Please provide a link to the relevant contributions and/or diffs when the trial is complete. — HELLKNOWZ ▎TALK 16:48, 8 November 2013 (UTC)[reply]

Trial complete. Internally, it keeps crashing; externally the edits look fine

The only "problem" I found was when it tried to warn AnomieBOT when it made things worse. I'll exclude bots from being watched. 930913(Congratulate) 04:58, 10 November 2013 (UTC)[reply]

There seems to have been no discussion as to the effectiveness of notifying IPs; about half of the notices are to IPs. Could you please work up a table of how long it took the anons, and the logged in editors, to go back to the notified articles and edit (doesn't have to be to repair the damage, just a subsequent edit)? Can you also classify the subsequent re-edits as positive or negative (I suspect talk-noticing the IPs will only produce a WP:BEANS response).

I know this is only intended as a talk page bot. In this edit the user added an inline ref to a page with a references section. Should some bot have just thrown a {{reflist}} into the article?

Is it possible to classify the error - I know that it's common to put a reference below the {{reflist}}, as in this edit, and I think that's because the error message is misleading - "There are <ref> tags on this page, but the references will not show without a {{reflist}} template" ...after it.

In summary: good identification of the editors and edits. Josh Parris 08:40, 10 November 2013 (UTC)[reply]

@Josh Parris: Not quite a quantitative summary, but a qualitative one. All bar one of registered users who have edited since the notification have fixed it. The one still checked it, but failed to find the error. Of the IP editors, only one has edited since, though not to fix it. Either it is too soon for IPs, or it is likely the effect of these weekend, off-the-cuff editors. As such, there has been no incident involving legumes. Vandalism is represented at just over 10%, which is slightly raised from the background noise. 930913(Congratulate) 17:45, 10 November 2013 (UTC)[reply]

I'm pleased and surprised. Messaging talk pages regarding breaking edits (brakets and disambiguation) has precedent, so I imagine that the community wouldn't object to this going live as a useful bot. BRfAs are conducted in public so other editors can comment on them, and I haven't seen any objection here or in the botreq (linked). Your reported bot crashes don't affect the output of the bot, so I'll leave resolution of those crashes as a matter for you. As such at this time I intend to approve the bot.

As for the source code: do you have plans to release it? Have you considered the license it will be available under? Editors (and botops) come and go, if your sources are freely available someone else will be able to pick up the mantle if you retire. Josh Parris 06:02, 11 November 2013 (UTC)[reply]

@Josh Parris: It's mainly for lack of effort that I don't release it. (Also, I don't get on well with git, and my code is not designed for reuse (read messy.)) If I did, it would be under something fairly open such as GNU/MIT/CC/copyleft. As it stands, it is on the labs, and therefore were I to disappear, the tool could be taken over by another person. 930913(Congratulate) 14:03, 11 November 2013 (UTC)[reply]

I strongly encourage you to make the source code publicly available in the near future. Cleaning it up from a messy state to one that is fit for public scrutiny is a lower priority, yet also important in that it will increase both confidence in the code and your ability to fix any problems that you find in the future. Remember that you will be held responsible for any misbehaviour by your bot.

I made had a cursory examination of your history and it appears that - baring some early question marks - for a few years you have been an editor in good standing.

As such, given the bot has demonstrated itself to be useful and well behaved,

Approved. Josh Parris 22:41, 11 November 2013 (UTC)[reply]

The above discussion is preserved as an archive of the debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA.