Wikipedia talk:Bots/Requests for approval/RjwilmsiBot 6

This is an example of a poor decision to approve a bot where there is no community consensus for the task being sought. Most ISBN-13's are unhyphenated (apart from one hyphen after 978 etc.), and the addition of hyphens lengthens the text, leads to ugly line breaks, and causes problems with searches. Geometry guy 21:14, 23 February 2011 (UTC)
 * I understand what you are saying, but the request was open for threes weeks and several outsiders commented, all supporting. There wouldn't have been any grounds for denying. -- MBisanz  talk 23:37, 24 February 2011 (UTC)
 * My above comment does not adequately express my concern. How many articles are likely to be affected by your decision to approve this bot, would you estimate? Geometry guy 23:56, 24 February 2011 (UTC)
 * Based on my good faith reliance on the bot operator's estimate at the time of approval, several thousand articles. -- MBisanz  talk 00:05, 25 February 2011 (UTC)
 * If it were instead several hundred thousand articles, would that change your view on approving the bot? Geometry guy 00:43, 25 February 2011 (UTC)
 * I might have asked him to post a link on the village pump, but no, it wouldn't change my view on approving the bot as it functioned as indicated in the trial and all commenting parties supported it.  MBisanz  talk 15:20, 25 February 2011 (UTC)
 * Do you believe bureaucrats have a responsibility to the encyclopedia? At least 142,000 articles will be affected, and you actually approved changes (converting ISBN-10s to 13s) affecting another 800,000 articles. And on the basis of just one notification (to WT:ISBN) and just three editors commenting, you say "there wouldn't have been any grounds for denying". This is totally backwards: you need grounds to approve, not grounds to deny. Have you read WP:Bot policy? It still says what is said in June 2008, when a BAG editor approved the creation of over a million micro-stubs. Geometry guy 22:22, 25 February 2011 (UTC)
 * Yes, I believe BAG members and bureaucrats have a responsibility to the encyclopedia. Yes, I have read the bot policy, I helped write some of it in fact and I even invented the means for challenging a bot approval at WP:RFC/BOT. Literally thousands of bot tasks have been approved and given that another BAG member had trialed the task and agreed it worked properly and two other people had agreed with it and no one else had commented in the three weeks it was open, there was consensus by the means with which bots are approved and why I approved it for functioning correctly.  MBisanz  talk 02:36, 1 March 2011 (UTC)
 * You approved it for "functioning correctly": that is utterly insufficient. Can I remind you, as I did in 2008, that bot policy states:
 * "The decision to approve a request should take into account the requirements above, relevant policies and guidelines, and discussion of the request." The "requirements above" include: "performs only tasks for which there is consensus".
 * Current discussions demonstrate that there was no consensus for an encyclopedia-wide reformatting of ISBNs along the lines you approved. Approving bots is not purely about checking they are "functioning correctly": thousands or millions of articles can be affected by a decision taken here; if you do not investigate the consensus, or even the number of article which might be affected, then you are not doing your job. Please learn from this. Geometry guy 00:35, 2 March 2011 (UTC)

Ive been doing the exact same thing when I edit the 10/13 conversion and have had zero issues raised, and Ive made over 5,000 edits where I converted isbns along with other edits. ΔT The only constant 22:24, 25 February 2011 (UTC)
 * I also convert ISBN-10s to 13s where the latter are available (and I check first). However, I bet you didn't make those 5,000 edits in one day, nor do you plan on changing nearly a million articles. WP:BRD is a concept which does not apply to bots. Geometry guy 22:54, 25 February 2011 (UTC)
 * Whether the bot edits @ 5,000 / 500 / 50 / 5 per day is inconsequential. The bot's task has consensus, the bot request stayed opened for weeks, no one objected. You don't even object to the task, you seem to object on purely procedural grounds. If the task is indeed controversial / doesn't have consensus, then it's a simple thing to revoke approval. Headbomb {talk / contribs / physics / books} 09:00, 2 March 2011 (UTC)
 * Please read before you write, as your comments are ill-informed and/or wrong in several respects. The bot did not change ISBN-10s to 13s, because the bot operator (independently) thought that this might not have consensus and might cause problems (yet this task was still approved). Instead the bot hyphenated ISBNs according to ISO rules, which are not used by the National Library of Congress, Amazon, or Google books. The bot operator does not appear to be convinced any more that this was a good idea either. Whether I object to the task is irrelevant, but your characterization of my opinion is wrong, as you will easily find on related pages (e.g. WT:MOS and WT:ISBN).
 * These discussions also show that there is not yet consensus for the task. It did not have consensus at the time either: there is no consensus without consultation, and the only visible communication was with the c. 70 watchlisters of WP:ISBN, which is a pitiful level of consultation for a task potentially affecting 1 million articles. The frequency of bot edits is very consequential if the task is disputed and needs to be revoked, because the frequency contributes significantly to the amount of damage done before problems are raised.
 * My interest in prolonging this discussion is not in "objecting on procedural grounds", but in ensuring that such failures, which have happened before, do not happen again. The response that "The bot is technically okay so there were no grounds to deny - anyway, if anyone objects, we can revoke the approval" is utterly unacceptable and contradicts Bot policy.
 * How about instead "Hmmm... maybe, in accordance with bot policy, we should tighten our notification procedures a bit and ensure that decisions are not based purely on technical grounds"? Geometry guy 00:21, 3 March 2011 (UTC)