Wikipedia talk:Bot Approvals Group/Archive 3

Latest comment: 16 years ago by Nichalp in topic Cocoabot
Archive 1 Archive 2 Archive 3 Archive 4 Archive 5 Archive 9

New Member

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section.

2 days, no more input, all supports - I'll declare the result now -- Tawker 17:24, 24 March 2007 (UTC)

Since there has been no comment on the request by User:Martinp23 to join BAG, I'm just going to appoint him shortly if there is no discussion. I've looked at his contributions and they seem great to me. I wanted to repost here just to make sure no one missed this if they had any issues. -- RM 22:56, 21 March 2007 (UTC)

  • It's not a particular prominent request though is it? Very easily missed. That said, I support anyway. --kingboyk 23:03, 21 March 2007 (UTC)
    That's why I reposted it here. :) I was concerned that people missed it, but posting it here seemed like the appropriate thing to do. -- RM 23:12, 21 March 2007 (UTC)
  • I support as well. —METS501 (talk) 00:33, 22 March 2007 (UTC)
  • Support, but would like to see this thread open for 1 week closing on 22:56 28 March 2007 (UTC) to leave room for community input. — xaosflux Talk 00:45, 22 March 2007 (UTC)
  • Support (as the nominator). Now that we have the attention of persons interested, we'll leave this open as requested above. -- RM 02:00, 22 March 2007 (UTC)
  • Support from me too. — Ocolon 07:17, 22 March 2007 (UTC)
  • Support, but I'd like to remind Martinp23 of Wikipedia:Bots/Requests for approval/Spelian and why a large amount of care is needed in approving bots. --ais523 09:46, 22 March 2007 (UTC)
    Yes, I agree, hence my stated objections to such a bot running :). I think that the best way for something like this to work would be to output a list of pages which appear to contain typos, and leave a human to fix them (or report to IRC?), hence my interest in getting a list of pages potentially affected. Of course, for this to work there would need to be a communtiy proposal and peole willing to work on the output :) Martinp23 12:25, 22 March 2007 (UTC)
  • Support, he is a fine user (and admin) with a good bot-related experience. Snowolf (talk) CON COI - 22:39, 22 March 2007 (UTC)
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.

Betacommand


Closure

I'm pleased to see Betacommand accepting, clearly, some responsibility for his actions. I think it might be a nice gesture were he to make a similar statement (on AN/I) that he understands the gravity of his actions and personally apologizes for the difficulty it has caused, but that's clearly his own decision - it's not for me to request or demand (this was just a general observation and suggestion). Personally, I think that BC's sixty day suggested "lapse" period is sufficient, if minimal, but I also think that's not a debate I choose to enter. However, I want to be sure that neither he nor anyone else thinks that's a punitive action - I think that it could be an incredible opportunity for him to take that time to learn how to handle similar situations in the future. I encourage Betacommand to discuss any large batch edits that he makes with someone he trusts who can evaluate the situation both technically and politically.
I would also like to say that I hope that this is settled very quickly (I would personally like to see this all closed up in the next 24 hours). One of the things that I love about wiki is the incredible warmth and caring for other people that is often demonstrated, and I can't imagine how difficult all of this must be, emotionally, for Betacommand. I think, then, that the proper action is to begin to wrap this up so that he can stop worrying about it and we can all put this energy into other things. For the record, I think the idea of a community sanction/block/ban, which some have nervously brought up is well outside the scope of reasonable response to these actions. Blocking/banning is our most radical steps and although I don't want to minimize the actions, I do not believe they rise to the necessary level for invoking that "nuclear option."
I also think that if Betacommand takes this time to work on rebuilding the community's trust, I think it behooves all involved to consider that when he reapplies (obviously, "applies" isn't the correct word, but I can't think of the right one at the moment) for BAG. And one final thought: I want to reinforce, as have so many before me, that Betacommand's work on wiki has been 95% brilliant and 5% challenging. It would be a terrible shame if people did not continue to recognize the good work he does, as well as comment on areas that need development. This community "forgives and forgets" more than any other I have ever seen (I am not including "trolls" in my definition of community, only those reliable and stable editors) and I know that if Betacommand works as hard at rebuilding the community's trust as he does building tools, he will be forgiven and all will be forgotten. I wish him nothing but the very best, and I trust that's true for all of you as well. Philippe Beaudette 07:00, 24 March 2007 (UTC)
Again, a very thought out and great response from Philippe. I agree with the proposal and response above except for one point. I think that BetacommandBot's approvals should all be withdrawn and Betacommand should be encouraged to reapply immediately. This is solely so that all the approved tasks are clearly documented in the case of any future controversy. —METS501 (talk) 14:41, 24 March 2007 (UTC)
I agree with Mets that all should be withdrawn so we can have a permanent record of exactly what the approvals are used for. I also second the motion to close this. Before the above section was closed, I did make another comment regarding the "personal insult" feeling that BC had for me, and I encourage him to read it. That said, since BC feels personally insulted by me (for unjustified reasons I think), I ask that another BAG member execute the proposal so that it doesn't look like I harbor any resentment or bad feelings. -- RM 14:56, 24 March 2007 (UTC)

I have closed the discussion on the above debate. Here are my findings:

  1. Betacommand has ceased to be a BAG member but he is free to help out here. He cannot approve of any bot.
  2. Betacommandbot will remain unflagged for now. To regain the flag, Betacommand needs to put in a new request which needs to be approved.
  3. After 60 days, Betacommand can reapply for BAG membership.

=Nichalp «Talk»= 14:56, 24 March 2007 (UTC)

Agree. As of now, Betacommand is officially unapproved for all tasks. —METS501 (talk) 16:48, 24 March 2007 (UTC)

BAG second opinion needed

at Wikipedia:Bots/Requests for approval/Rschen7754bot. Is this too broad and vague an application or am I being narrow minded? --kingboyk 13:05, 23 March 2007 (UTC)

I'm not a bag member, but the tasks ought to be a bit clearer. —— Eagle101 Need help? 18:59, 23 March 2007 (UTC)
They are a little ambiguous, it really depends on how much traffic it is going to be using for the other tasks in my opinion. Small single-runs that would be capable at human-AWB speed don't really need heave approvals here. (Note, I've approved ONE task for trial on this request). — xaosflux Talk 04:32, 24 March 2007 (UTC)

Another second opinion needed

I've been in quite a long discussion that with Calton at User_talk:Calton#High_speed_editing, regarding high-speedy editing (>6-8/min) sustained. This is mostly in regard to mass tagging talk-pages that they think should be speedy deleted as {{db-talk}}. These are mostly in the Image_talk: namespace, and are causing log flooding of the Image_talk recent changes log. (consuming >75% of the filtered log). I've suggested to them that they register another accont for this, so that we can bot flag it. They seem pretty adamant about not doing this. Most of their edits are valid nominations, and are being deleted, so history and recent chagnes log will be hard to review. You can see an example at User:Xaosflux/Sandbox13. I was going to block this user for running an authorized bot, but not quite sure, any opinions? — xaosflux Talk 03:15, 25 March 2007 (UTC)

Looks like other users are suggesting they run this under a bot as well now. — xaosflux Talk 03:18, 25 March 2007 (UTC)
Warn to slowdown to 2-4emp and if they dont stop block Betacommand (talkcontribsBot) 03:20, 25 March 2007 (UTC)
This is awkward. I've never been faced with the situation where a bot flag was recommended to someone who was not a bot. I'm not comfortable extending bot policy to real humans, although there is always a grey area separating a "bot" vs. a "browser". Alas I'm too tired right now to think about this. -- RM 03:38, 25 March 2007 (UTC)
I thaught of that at first, but we do flag bots that are manually assisted, and basically all this editor is doing is mass-tagging based on a list that they are generating in preview mode. I'd be happy to flag an alternant account of their's if they showed that they would meat the other criteria. — xaosflux Talk 03:43, 25 March 2007 (UTC)
Man. He seems pretty hostile. I can't see why he decided to take offense at your suggestions. Mike Dillon 04:54, 25 March 2007 (UTC)
I'd be concerned at flooding the CSD backlog too much..... perhaps 50 pages at a time tops / per day. Those backlogs can be brutal.... -- Tawker 05:49, 25 March 2007 (UTC)
I'd have thought it better to avoid CAT:CSD altogether, they are usually a lower priority than the article pages. Why not batch up a set of links onto a page and include a reference on the CAT:CSD page. Deleting those is usually a no-brainer, look at the article tab, see it's red, delete. A list page would enable however many need to be added in a single edit, and a small amount of maintenance to remove the red-links once deleted. --pgk 08:25, 25 March 2007 (UTC)
I agree with pgk, and with the suggestions offered by xaosflux on Calton's talk page - a single list shuld be created with a custion speedy deletion tag on it, explaining that admins should check and delete items on the list. If he won't do this, then we have the unattractive option of a bot flag and a flooded CAT:CSD, or, failing that, we may need to consider a block. This all depends on Calton becoming more open to suggestions, which could take a while to start happenning. He should, in any case, stop tagging images at such a rate until these concerns are resolved. Martinp23 10:44, 25 March 2007 (UTC)

Old requests

Hi! I just removed 5 requests that had been closed but not removed from WP:BRFA, so please try to remove requests when they are approved, denied, withdrawn, or expired. Thanks! ST47Talk 21:46, 25 March 2007 (UTC)

Regarding this did you archive these or just untransclude them? — xaosflux Talk 01:02, 26 March 2007 (UTC)

Betacommand RFAr

Just stumbled upon this: Wikipedia:Requests_for_arbitration#Betacommand.

As is often the way on Wikipedia, nobody bothered to inform BAG and there's hardly any mention of the fact that the bot issues have to an extent been dealt with. Betacommand has already been sanctioned by being removed from the approvals group. (However, the issues being raised at the RFAR go beyond what we dealt with, and the arbitrators have voted to accept the case).

Anyway, I think one of us should post a link to the discussion which was held here and perhaps a synopsis. A draft could perhaps be written on this page first.

I have no intention of getting involved in the case personally unless called as a witness, but the BAG events need to be reported as they're important. --kingboyk 14:08, 27 March 2007 (UTC)

I left a message for User talk:AnonEMouse regarding this, and will probably find time later today to post directly to the arbitration page, but until then that will have to do as far as I am concerned. -- RM 17:38, 27 March 2007 (UTC)
I've posted to the arbitration page with a statement. I think most or all of what I've said matches our opinion on this. -- RM 18:19, 27 March 2007 (UTC)
Very nice piece of writing. Meets with my approval. --kingboyk 18:34, 27 March 2007 (UTC)

BAGBot downtime

Just so you know, User:BAGBot will be down from Noon UTC tomorrow to midnight UTC monday morning. If anyone has a linux computer, let me know and I can send you a binary. ST47Talk 19:35, 6 April 2007 (UTC)

I don't know if this is possible, but have you considered running the bot on the toolserver? —METS501 (talk) 19:46, 6 April 2007 (UTC)
Im going to try and get it up there. (I have the code and there are no objections?) Betacommand (talkcontribsBot) 19:48, 6 April 2007 (UTC)
Beta can run it on toolserv - I'm in the process of requesting a toolserv account for this bot and others. ST47Talk 19:49, 6 April 2007 (UTC)
Awesome. —METS501 (talk) 19:50, 6 April 2007 (UTC)

Heads Up

there is some users that are not happy with BAG see [9] Betacommand (talkcontribsBot) 01:56, 13 April 2007 (UTC)

Thanks, just replied there. — xaosflux Talk 02:22, 13 April 2007 (UTC)
He's what one calls a forum shopper. If he doesn't get his way he moans on user talk, IRC, village pump, and just about anywhere. Check his contribs and tell me you want him to have a bot flag... --kingboyk 11:55, 13 April 2007 (UTC) Not a helpful comment in retrospect. Struck. --kingboyk 12:04, 17 April 2007 (UTC)

I had a word (well, four hours of words :P ) on IRC, firstly with Vishwin60, who invited me to the USRD's channel. Basically, I explained our position on each of the cases that Rschen has found issue with (and, in some issues, the whole of USRD were unhappy), and eventually we were able to amicably close the issue :). It seems that communication breakdown on both sides lead to this "dispute", but hopefully the issue has been resolved now. Rschen did have a couple of suggestions for the BAG, which I have invited him to bring forward here or at WT:BRFA. Thanks Martinp23 10:06, 15 April 2007 (UTC)

Trying to recall what they were :)... I know one was higher publicity for BAG nominations. Possiblt at the top of WP:BRFA? --Rschen7754 (talk - contribs) 04:44, 17 April 2007 (UTC)
Village pump, perhaps. --kingboyk 12:03, 17 April 2007 (UTC)

Javascript

Anybody fancy knocking up a monobook scriptlet to help with closing and archiving approval requests? --kingboyk 13:57, 13 April 2007 (UTC)

VoA has one. Betacommand (talkcontribsBot) 15:35, 13 April 2007 (UTC)
Cheers Beta. --kingboyk 15:41, 13 April 2007 (UTC)
User:Voice_of_All/UsefulJS. Installing now. --kingboyk 16:18, 13 April 2007 (UTC)
Didn't work for me, alas. --kingboyk 00:19, 15 April 2007 (UTC)

BetacommandBot case closed

I've closed and archived Wikipedia:Bots/Requests for approval/BetacommandBot. I think we'll all be glad to put the case behind us, and look forward to seeing many good and useful edits from BetacommandBot (talk · contribs). --kingboyk 00:18, 15 April 2007 (UTC)

Abstentee Operators

A discussion regarding bots operating without respondant operators is in progress regarding Hangermanbot on WP:BOWN. Thanks, — xaosflux Talk 01:17, 17 April 2007 (UTC)

Backlog

Well, the "Backlog" code I put in is finally in effect - if 3 or more BRFAs have the {{BAGAssistanceNeeded}} tag, it adds that note. Perhaps additional BAG people should be searched for? if we look at the history, we can see that the amount has been rising from 21 up to a peak of 29, of these 25, one hasn't even been edited this year, and 10 haven't been edited this week. ST47Talk 10:33, 17 April 2007 (UTC)

A lot of them are waiting in trials, I believe - I'll take a look at the BAGAssisstance... ones now. Martinp23 11:12, 17 April 2007 (UTC)
I think we could do with more people, yes. We're all volunteers, and being techie types we have plenty else to do (in my case, AWB, plugin, writing articles, adminship). --kingboyk 11:31, 17 April 2007 (UTC)
Indeed - at the moment, a lot of the BAG are inactive (wikibreak, working on other things...). More people may help. On the issue of the old requests - things aren't as bad as they look! Most requests are awaiting responses form the operators, and the one year old request is probably caused by me putting it in the trial section wrongly, and confusing the bot :) Martinp23 11:52, 17 April 2007 (UTC)
Heh! You all didn't let that backlog notice stay for too long! Well, all the active BAG members are admins, but there is no technical reason for it to be so, admins can't flag bots anyway. I know that as of maybe a month ago there wasn't an official process for additions to the group, but if there is now, can I be pointed towards it? ST47Talk 18:20, 17 April 2007 (UTC)
There is no "official" process other than someone nominates themselves or someone else and then there is a "vote". Typically anyone who receives unanimous or near unamimous approval after a week is appointed to the approals group. See the new member discussion at the top of this page. It is obvious to me that we could use more members, as I have been mostly inactive for some time now. I only do an approval here and there because I've recently been very busy. Generally the main requirement is that a person has been active contributing to bot approvals (just not approving) for some period of time, to demonstrate their understanding of bot policy and other applicable policies. -- RM 20:38, 17 April 2007 (UTC)

Nomination/Request/Vote/!Vote/Discussion/!Discussion/(\w+) for ST47


Result: User:ST47 is appointed to the Bot Approvals Group.

The support was unanimous so I feel that closing this and announcing ST47 as a member of the Bot Approvals Group is uncontentious and within my remit as an admin. If anybody wants to ask a crat to drop by and confirm the result, feel free. --kingboyk 11:31, 26 April 2007 (UTC)

Backlog

Again I'm sorry to nag, but it's been three days now with no response whatsoever from BAG people. --Selket Talk 23:35, 20 April 2007 (UTC)

  Done Martinp23 23:44, 20 April 2007 (UTC)

Additonal opinions requested

I've blocked Cydebot for editing in excess of 50e/min, please see User_talk:Xaosflux#Cydebot_Block for several things going on with this bot. Thanks, — xaosflux Talk 03:41, 22 April 2007 (UTC)

Moved to centralized discussion at Wikipedia:Bot_owners'_noticeboard#Cydebot_Block. — xaosflux Talk 05:24, 22 April 2007 (UTC)
The discussion on WP:BOWN has gotten kinda heated, and I'm stepping away from it for at least a few hours, please comment there as you see fit. Thanks, — xaosflux Talk 20:37, 22 April 2007 (UTC)

Inclusion in "Bot Approvals Group"

Hello. I intend to become a part of the Bot Approval Group. I understand I should post a message here indicating this intention, rather than just adding myself to the member list on the project page. - Mark 10:47, 22 April 2007 (UTC)

Strong Oppose I havent seen you in the BFRA discussions before. Betacommand (talkcontribsBot) 12:19, 22 April 2007 (UTC)
I think you misread what I said. I wasn't asking for your permission. I was merely stating my intention to become a member. - Mark 15:19, 22 April 2007 (UTC)
That's not how it works. If you're not capable of scanning a talk page (hint: how to become a member is dealt with just a few threads up) how does the community know you should be trusted with examining bot applications?
Is there some political context here? I see you removed reference to "authorised". --kingboyk 15:42, 22 April 2007 (UTC)
I saw the thread a bit above of someone starting a vote for themselves to become a member of the group. Given that they could not even decide whether it was a vote, a request, a nomination or discussion, I took that as an indicator that there is no formal procedure for admission for the group, as was highlighted on the project page before you edited it just now. Since this group is effectively a mere advisory WikiProject, aiding the bureaucrats with performing the task for which they have been properly elected (via WP:RfB), and with no real unique function of its own, I assumed that this group would be open to new members willing to put in time and effort on the project. Rather than being insular and self-determining, I mean.
As for the 'authorised' edit to the project page, I had someone complaining on IRC the other day about this group, demanding to know from where it got its authorisation. Because I could find no evidence of any such express 'authorisation', I thought it would be wise to remove the reference to avoid confusing such users in future. Besides, I feel it is worded better as it is now because it emphasises the role of the group as a whole rather than individual members of the group. I may be wrong; I was under the impression that the Group acts collectively rather than as individuals. - Mark 16:04, 22 April 2007 (UTC)
Ah OK, a misunderstanding then; sorry. The attached page was a little out of date; we generally now have a discussion for 5 days or a week, in which all community members are invited to participate. We then ask a crat to close it. This gives a slightly more formal air to proceedings. I've now, as you saw, updated the page to reflect current practice. Also, you are quite right, we need more members and welcome fresh blood. If you wish to join please nominate yourself here, and announce it at the bot owners noticeboard and village pump (technical). I know you have an impressive resume, but I think folks would like to see some evidence of bot usage or expertise, if you have such experience and/or knowledge it would be worth noting it in your self-nom.
With regards to "authoritah", I have no problem with your rewording, but would note that a current ArbCom case currently appears to be affirming this group's mandate over bot activity. --kingboyk 16:13, 22 April 2007 (UTC) PS I can probably guess who was complaining on IRC; we've heard a lot about it and the complaint was resolved to everyone's satisfaction, I believe.
Fair enough. I've had a look back over the past archives of this page and see that people generally have to have been active commenters on discussions about bot requests for approval. I'm not sure I entirely understand this. People are meant to give their opinions about whether a bureaucrat should flag a bot as lay editors, so that they can become members of the Group which does exactly the same thing? Okay, I'll give that a whirl before asking for membership again.
Just in relation to the current practice of having votes and having a bureaucrat close the vote, I only see one discussion in which a bureaucrat closed the vote, and only a few actual votes in this page and the archives. Are you sure that counts as 'current practice', when it happens so infrequently? - Mark 16:21, 22 April 2007 (UTC)
WRT to your first point, that's how it's mostly worked, yes, but I wouldn't personally require a prospective member to have frequented the bot approvals page before. I mean, if Brion Vibber came along and asked for membership I'd support! It's certainly not a rule, rather a !voting issue folks have often dwelt on. You're quite welcome to try your luck now but you're also quite right in your observation and would probably be best advised to chip into a few discussions before seeking membership. Unless you're Brion Vibber. Or Jimbo Wales.
WRT to second point, you may be right there. Perhaps I introduced a new "rule" :) Revert me if you disagree, but whether or not it's current practice I think it ought to be :) --kingboyk 16:29, 22 April 2007 (UTC)
(outdent). An important factor on ANY bot request is the opportunity for community input. Please note, that every editor is welcome to take part in the disucssions, and yours (and anyone elses) input in to the request would be welcome. — xaosflux Talk 20:01, 22 April 2007 (UTC)

Resignation

Over recent weeks, I've become less and less convinced that BAG enjoys the full support of the community and has a clear mandate. Today it would seem that we don't; in particular it would appear to be quite alright for decisions to be made elsewhere (on IRC, I'd wager) without informing us, and then for me to be personally attacked for attempting to uphold what I believed to be policy. I'm a volunteer here, and there are other things I can be doing with my time which are more productive than taking part in a process which doesn't have full community support.

This isn't one of those "meatball goodbyes"; nor do I want any fuss made about it. However, I do hope this will stimulate reasoned debate about BAG's remit and composition, a debate I look forward to taking part in. If BAG is reconstituted I may yet stand again, but I will not be seeking readmittance without going through some "official" procedure first. Yours, with love and respect, the pompous, idiotic empire builder also known as --kingboyk 22:11, 22 April 2007 (UTC)

An idea, I'm sure there are others but, lets make this back to what it used to be, a simple noticeboard where people who want to run a bot can post a message. If ther are no complaints in say a day or two, just run the bot. —— Eagle101 Need help? 22:24, 22 April 2007 (UTC)
I agree that the BAG definitely does not have the support of the community at the moment, and I respect your decision to resign (temporarily at least). I hope you're not too upset about the whole thing; you did well in the discussion, and many of the attacks were uncalled for. Let's try to get a real discussion going instead of arguments and then perhaps things will be different. Eagle's suggestion has merit; you might want to bring that up here. —METS501 (talk) 22:26, 22 April 2007 (UTC)
Cross posted. —— Eagle101 Need help? 22:28, 22 April 2007 (UTC)
Just to be clear, I think this resigning is rediculous and the whole idea that approvals do not have community consensus to be equally rediculous. That said, I have no objection to improving the process. -- RM 13:00, 23 April 2007 (UTC)
Given the string of recent events, maybe it would be worthwhile to re-propose the BAG as it now stands -- and re-elect members (I know this is asking a lot for the current members) through an official process as if it were a new proposal. If we do decide that this is the way to go, I think it essential that the BAG continue to operate as it does at present while the discussion is ongoing. --Selket Talk 15:14, 23 April 2007 (UTC)
Waste of time. We have a system that works. There is nothing wrong with it. If you want to make a bot, and you know what you're doing, if you have source code available if asked, if you watchlist the approval and pay attention to it, you can easily get approved in 3-4 days. Are you going to go to DMV and ask for a license but leave your identification at home? The main complaint that started this all was they Cyde was too impatient to go through a procedure. He didn't even have to, a simple note would have saved us all this trouble. There is nothing wrong with BAG - one admin blocked a bot that was editing faster than BAG authorized it to. At the time, he didn't know a dev authorized it. That's what any admin would do. We don't need to spend weeks replacing this system, we just need, as myself and many others have said, more community input into bot approvals. ST47Talk 15:24, 23 April 2007 (UTC)
Of course it's ridiculous RM, but how else am I supposed to prove I'm not interested in empire building? How else am I supposed to clarify whether or not the community consider me competent other than resigning and restanding when the fuss has died down?
I think the recent discussion shows that most people do support the BAG, which is encouraging, but a vocal minority don't. They were just a little too vocal for me personally to handle, and I'd rather take a back seat for however long it takes (be it a few days or a few months). I hope you can understand why I did this and appreciate it that I did it for honourable reasons and with a heavy heart. Besides, the resultant debate is good. --kingboyk 15:45, 23 April 2007 (UTC)
You must have posted this while I was composing my comment below, because I didn't notice your comment and was replying to ST47. As for your comment, I appreciate everything you have done and don't think YOU are rediculous, an important distinction. I have nothing against you, only good thoughts, and I understand why you did it. That doesn't change my opinion that I don't like it! -- RM 16:02, 23 April 2007 (UTC)
I understand, thank you for the kind words. --kingboyk 16:06, 23 April 2007 (UTC)
I agree that the system works and that I don't want to go through the headache and hassle of revamping a system that has the chief problem of not having enough members. Heck, I might not even stand to be included in a new system because going through that is just time consuming for a process that doesn't have enough interest anyway. It is blatently clear that the community does not want to increase its oversight of bot approvals. Every time we try to get more people involved, it doesn't much happen because it isn't all that interesting to "outsiders". They like the fact that a specialized group takes care of these issues and they only get involved when issues push the limits of current policy, as I think it should be. Those people complaining about the process would be better suited to just help out. That would be real help. Changing the process would not increase its participation, thus avoiding fixing the core problem. If you don't believe me, take time to look at the various archives to see how community input has been sought from time to time with only limited response. It is really easy to monitor bot approvals and get involved if you want to. Most people do not want to. We've had requests to join BAG which have been rejected. We ask them to participate in the process before joining BAG, but they often choose not to. It's as if people want the selective power of BAG without having to go through the work of being useful. It's terribly frustrating because all this effort could be geared towards actually helping out. -- RM 15:55, 23 April 2007 (UTC)
Indeed, and that's what hurts me the most. One moment I think I'm helping out, the next I'm an empire builder and a tick and goodness knows what else. WP:AGF went right out of the window. If folks want volunteers like us to do a job, they have to be a bit more supportive and appreciative and not undermine us when it suits them. --kingboyk 16:06, 23 April 2007 (UTC)

In that case we need to wait on Cyde to reply here as for his reasoning on not making a request. —— Eagle101 Need help? 16:08, 23 April 2007 (UTC)

KISS

I've been in the hospital for 3 days and had to catch up to some things. Also, I've been busy working on server-side script. At any rate, having read up on the cydebot issue I have to say I am dissappointed.

This started a while ago, as things became more needlessly bureacratic. The helper bot has at helped the formatting and such at least. However, I do see some issues with the purpose of the group. AFAIK, it makes sure:

  • Bots are in accordance with standard policy (MOS/XFd/disruption and such).
  • Bots are capable of being watched if they are perfoming a new or complex task.
  • Bots are not server hogs.

If cydebot is useful, following CfD consensus on relevant discussions, is tested, and does not cause server issues then it's OK by me. Unfortunetely, BAG sometimes becomes such a slowdown, some people may rather just ignore it. In think that one issue is that bots are supposedly not supposed to be ran much at all until approve, which seems to be bullocks. I do believe that any high edit bot should be tested first and mentioned to BAG (for the sake of watching it/preventing disruption), have a good user page description, and keeping track of all high edit bots (as they start to add up on load). However, users should be able to run swaths of 100 edits or so and then come here, with that as an example of their final testing, which should get quickly approved, unless some real policy is violated.

The good thing about an edit limit is that as the number of high edit bots increases, the load is much less. However, generally few high-edit rate bots run at the same time as there are not that many and their tasks tend to be one-time. It is useful to set a developer recommended edit rate limit as policy, which would be much higher than the current one. The current one is to be able to "track" changes, which should only be done for bots that are in testing or doing somthing very complex. So the factors that decide a bot's max edit rate are:

  • How much it needs to be watched
  • Server use

The bot task was within image policy, simple, and the bot was by a veteran bot operator. I'd assume that Cyde would have tested it anyway, so I don't see what the big deal is. Also, if the developers thought it OK, then there is NO technical issue here. When I saw "Note, that these concerns are purley from the technical nautre" and "You still haven't answered the simple question of what position gives you the authority to approve of it." I wanted to puke. None of the BAG members there even are MediaWiki developers, while Gregory actually was. In fact only 2 BAG members are developers, none of which had an issue with the bot.

...*Sigh*...as for community support, I am not sure how exactly that can be regained at this time. Voice-of-All 23:48, 22 April 2007 (UTC)

How many times do I have to say that I DIDN'T KNOW HE WAS A DEVELOPER?! All he had to do was say "I approve it because I am a developer" and I would have effectively said "OK, boss" (which I did, when I knew it). Which part of "You still haven't answered the simple question of what position gives you the authority to approve of it." isn't plain English?
As for Cyde being an experienced bot op, he's also an experienced bot op with quite a block log.
Finally, the edits themselves seem to be trivial and speaking solely as an editor (I quit BAG anyway so can't speak as a BAG member), a waste of time - mere semantics. Unless, in fact, the idea is to mark out "victims for the kill" and the next step will be to delete all album covers?! --kingboyk 23:53, 22 April 2007 (UTC)
Oh, let me say one other thing. Since when was the ability to write a bit of PHP a ticket to being rude, abusive and all powerful on wiki? --kingboyk 23:54, 22 April 2007 (UTC)
The issue wasn't that Greg was clearly known to be in a position to know what he was talking about. The problem was that the BAG members were not in a position to question him even without knowing he was a dev, and ironically, he was. So you didn't have to "know he was a developer", and it's OK not to have, thats not the issue. The issue is that being on the group does not grant technical competence, whatever you had before is what you have. And being a dev is more that PHP, such MySQL, PostgreSQL, Apache, JS, CSS, XML, invoking browser plugins, I/O and overhead issues...and not all the devs here are good at all of those yet (myself included). In fact PHP isn't even Gregs strong area and the edit load issues are about the database usage much more so than the PHP CPU/memory use. I am not trying to be rude, but you must understand that other users feel that BAG itself is sometimes moderately disruptive and questioning users authority without having it, even if in good faith is still condescending, like what happened to Greg. Voice-of-All 00:12, 23 April 2007 (UTC)
So, Greg says he approves the task and I should accept that at face value. Why? As far as I was concerned at the time, only BAG has explicit authority to approve bots. If some random chap comes along and says "I've approved it", and I don't know he's a dev, is it not my duty to ask him on what basis he has approved it? What about if I go close an RFA now, isn't a crat entitled to ask what the hell is going on?
And what makes you so sure that I don't know PHP, MySQL, Apache, etc or don't have technical competence?! I was using Linux in 1993 for Gawd's sake. I'm an MCP and an AWB developer, and my AWB plugin has done hundreds of thousands of edits. Technical competence is supposedly a requirenment for BAG membership.
And, again, these "other users" would be the IRC cartel, wouldn't they? I've not seen any other serious complaints. --kingboyk 00:18, 23 April 2007 (UTC)
I am well aware that you have techinical competence in those areas. When I said "technical competence", I was refering to the relevant knowledge about this site and the what it uses. And IRC is not to be dismissed so lightly. A few days ago I had a user (who does a lot of spam work on meta) asking about whether a bot be need be approved, when I said yes, he moaned at the idea and I didn't see his request later. Voice-of-All 00:31, 23 April 2007 (UTC)
That's hardly my fault. I didn't invent the process and have been doing the job conscientiously and with good faith. Indeed, since clearly the process is broken, why am I not being commended for stepping back from it? Why haven't you resigned too? I really think you should (not you personally so much, but all BAG members. Why belong to a rejected process?)
Also, as I just said on David's talk page, this mess could have been avoided if we'd been notified in advance. Would xaosflux have blocked if we'd had advance notification that a dev had approved the task? No. Would I have voiced an opinion endorsing the block (which I must reiterate yet again, is all I did) if we'd had that advance notification again? No. Presumably since you didn't contradict my claim this has come from IRC, I'd like to ask the IRC crowd to let us Wikipedians know in advance the next time they plan to pull a stunt like this. --kingboyk 00:37, 23 April 2007 (UTC)
Don't you know? Developers hold all the actual power on the wiki. Administrators are merely a larger species of tick by comparison - David Gerard 00:00, 23 April 2007 (UTC)
Heh. Anybody would think Mediawiki is well written, not the jumbled up old mess it really is.
Xaosflux blocked a bot which was running - unannounced - at 50ppm. I had the audacity to endorse the block (whilst unblocking because Cyde had returned) and to ask gmaxwell on what basis he had approved it, and VoA wants to puke? I want to puke that my apparent inability to write PHP (which is incorrect, I've read and understood much of the Mediawiki code, albeit some time ago, and have written Mediawiki extensions for my own use) evidently makes it OK to talk to me like I'm a piece of dog sh*t.
Where were these "BAG members or are also developers" when the storm was raging? Not here that's for sure. We held the fort as best we could, yet apparently are criminals for trying to enforce what we believed (and which I no longer believe) to be policy. --kingboyk 00:07, 23 April 2007 (UTC)
Not criminals. No one things such, well, at least not me. I agree that we should have at least been informed, and a lot of bot user pages are very shitty at that too. However, there was no need for it to be/stay blocked. I am not saying that you deserved David's remarks for that. I can except users having made the block a lot more than the attitude that has built up in BAG that starts putting titles over reason. As Kelly said "If you don't know what the reason for the edit rate guidelines, I suggest you refrain from enforcing (or even arguing) them." Voice-of-All 00:50, 23 April 2007 (UTC)
Of course I know why there are limits; what I don't know is why it's set at 15ppm (and it now seems apparent that it's a totally arbitrary limit). Also, I obviously need to point out again that it was me who unblocked CydeBot! :) Look, I appreciate that you're not trying to make me out to be the villain, but there does seem to be an awful lot of scapegoating over what was a breakdown in communication and more the fault of the IRC crowd than of BAG. If they'd told us in advance, there would not have been a problem. Period. I've suddenly lost a lot of respect for several Wikipedians (not you) based on the way they conducted themselves over this.
Now, what about BAG? If it doesn't have community support, why aren't you resigning too? --kingboyk 01:02, 23 April 2007 (UTC)
David, this is the most un-wiki comment I've ever read here. I hope you're joking and even then it's not appropriate. Mike Dillon 00:14, 23 April 2007 (UTC)
You think that's bad? See Wikipedia:Bot_owners'_noticeboard#Endorse_block. All this because some BAG members dared to do their job, without checking with the IRC cartel first. --kingboyk 00:20, 23 April 2007 (UTC)
If so, DG has been "joking" quite a lot recently. Must be quite hilarious, in fact. Alai 00:24, 23 April 2007 (UTC)
Note that I'm one of said larger ticks. It's "un-wiki" to point out that hierarchy comes about naturally, and sometimes it's not what you expect? Read The Tyranny Of Structurelessness, one of my favourite essays on emergent hierarchies. Precis: if you pretend there's no hierarchy, one will emerge out of your sight and bite you in the backside. A hierarchy will happen because humans are involved; politics starts with two or more people in a space. When there's 45,000 editors in any given month, it's sometimes amazing anything but politics happens. Claiming that noticing this is "un-wiki" is not only burying your own head in the sand but insisting others do too - David Gerard 07:09, 23 April 2007 (UTC)
Cross post from WP:BOWN (VoA we must be on the same wavelength I was writing a very similar idea when you posted)

I was part of BAG and I think that BAG or some form of simi-tight control is needed. I have seen some completely ass-backward bot request for Approval/Bot request. On the other hand it is a real pain in the ass as a bot op to come back for every task. I propose a idea. Bot op's MUST apply for their first approval (and get bot flag). once a bot and operator is trusted let them do their work and any other similar work. if a task is a major change in what it does have the bot op leave a note on the BRFA talk and if there are any issue file a BrfA otherwise let the bot op work. (Keep It Simple, Stupid) the motto of all bot writers Betacommand (talkcontribsBot) 00:04, 23 April 2007 (UTC) (I will crosspost this)

Sounds like a good start. There are many users that we can trust with bot policy and the tech knowledge needed to let them do their own thing, unless it is very complicated or involves controversial policy issues. Voice-of-All 00:18, 23 April 2007 (UTC)
The trick is to come up with a process (however formal or informal) that reflects this understanding. An "only the first task needs formal approval" model could quickly come unstuck when task one is, let's say, interwiki, and task two turns out to be a welcomebot or a transatlantic English "correction" script (in either direction). I do agree there's a large element of "this is straightforward, hurry up and let me do it while I have the time and energy" frustration to the current format, though. Alai 00:32, 23 April 2007 (UTC)
Perhaps after first formal approval, the bot op can then post new tasks that he/she is going to run on a simple noticeboard like format that the bag members and others can watch? All I know is that we have a slight problem with existing bot operators not proposing new tasks under our current system, so we can a) Go after the bot ops (the system is fine and no need to change it), or b) Attempt to make a more friendly and simple way to go about things. —— Eagle101 Need help? 12:54, 23 April 2007 (UTC)

Philosophical question

Who granted a dozen or so editors the power to decide who can run a bot and who can't? Why can't it be a community consensual decision? Where does BAG authority come from? Wasn't it self granted? -- drini [meta:] [commons:] 00:29, 23 April 2007 (UTC)

It started kind of like FA director, a fairly small consensus descision to form it. Most new members are voted on by BAG. I've encouraged outside election attention before, and none came. Giving the whole community access would slow things down as this is a low attention area. It needs a few people moderatly dedicated to maintaining it. Voice-of-All 00:34, 23 April 2007 (UTC)
Seconded. It's meant to be based on community consensus, interpreted through the approvals group. Unfortunately, the elections - and the bot applications themselves - don't attract much outside interest. If anything good comes of this business (and I'm sure it will), it will be more community involvement in and awareness of these issues and decisions. --kingboyk 01:04, 23 April 2007 (UTC)
Totally agreed with kingboyk's summarization. As long as the community is allowed to discuss, and this doesn't become a vote like RfA ("Do you agree that a bot of this type would improve the project, and do you truse the operator to run it?" "No, wikiproject tagging is best done manually" "Yes, Jimbo is god and if he wants a welcoming bot, then it shall be so") comunity involvment is the way to go. ST47Talk 02:12, 23 April 2007 (UTC)
I've been on vacation for the last four days and I return to find people resigning and talk of how BAG is not run by community consensus and so forth. Sheesh. I suppose I couldn't agree less with what most people have been saying and don't understand what the problems are. Perhaps someone could explain things to me, since I am liable to misunderstand some aspects of the so-called problem. I'm not sure where the complaints are that BAG is not community consensus driven, but it's easy to find complaints that BAG is too slow because it has too few members or members that are not active enough. Those are different issues altogether. Let me address the question above.
My bot, rambot, was one of the very first serious bots. It was not the first bot, but it was the largest one. It predated this whole process and was the reason bot policy was formed. My bot caused all sorts of unforseen problems: server load issues, recent changes cluttering, random page muck, creation of large quantities of questionable articles, bugs in the wiki software, etc. It engendered questions of whether or not a bot should be given the same rights as any other user or whether bots should be banned outright. One of the original wikipedia guidelines was that bots were "frowned upon". (As an aside: if you claim that bot policy is not community driven, note it was that same process that determined that bots should be allowed at all and are no longer frowned upon) Here is an original version of bot policy mostly written by myself. It was written when the software developers were directly involved in the process. The original basic premise of bot policy was that bot operators should regulate themselves. They were users and their bots had reasonable rights and as long as they followed policy, they would be ignored. The policy at the time reflected the opinion of those who were involved, which included far fewer individuals than it does now. The impact of my bot was over the entire Wikipedia and its affects felt everywhere due to how it affected some of the core features. It was discussed on the village pump numerous times. It was those discussions that led to the formation of the policy.
Eventually a more formal process had to be created to grant the bot flag to a growing number of bots. Still, the anti-bot feeling of Wikipedians was strong (and is perhaps still strong), so while the opponents were a minority, their influence required a process that restricted what bots could do. Bot approvals grew out from that community consensus. To clarify, there were different perspectives on bots: those who didn't want them at all, and those on the other extreme who felt that no process was needed and they were just like any other users. The bot process has always been intended to ensure that bots are good (useful, not harmless, not a server hog). (See this old version of bot policy). With all that background, let me answer the questions directly:
  • Who granted a dozen or so editors the power to decide who can run a bot and who can't?
    The process grew out of the historical requirement to manage both the bot flag and restrict what bots could do because of their abnormal potential for abuse. The original editors were orginally granted power at a time because they stepped up and helped out where it was needed. It was a time when less oversight was required because it had a smaller scope and fewer people were involved with Wikipedia as well. The current BAG membership is mostly new, having grown out of the previous group or groups. The whole process has been evolutionary, growing as consensus has changed regarding the process. It has always been so, and for years has not been the subject of any major controversy. As far as I know, this is the first time anyone has made a stink about it. It's hard to think of how a process that has been going on for so long under such community scrutiny could be declared to not being following community consensus.
  • Why can't it be a community consensual decision?
    It is a community consensual decision. All those persons that care about it have input, not just BAG members. BAG members have the job of closing discussions, whereas bureaucrats have the job of setting bot flags and ensuring that the discussions were closed properly (a second layer of protection). It is no different in concept than, say, RfA, except that fewer people are involved. Just because fewer people are involved does not mean that it doesn't follow consensus, only that most people do not care or defer to those who are more knowledgeable about such things. Community members have commented on dozens and dozens of bots, sometimes swaying an approval.
  • Where does BAG authority come from?
    As stated above, it is based on policy enforcement and the other historical needs. Some process had to be made to decide who should be given a bot flag. The development of the process was evolutionary. BAG authority is to ensure that bots do not harm Wikipedia and that they contribute in a constructive fashion. The process has been hugely successful, but has perhaps not scaled well in recent times. It should be noted that bot policy is an official policy of Wikipedia, something that it was not when I wrote it, despite the fact that it was only a guideline at the time. The requirement for "someone to approve a bot" was a VERY early feature (See here). It became policy when it was recognized that the process had wide community support.
  • Wasn't it self granted?
    This is a shortsighted and oversimplified way to understand the issue. Would someone who self appoints following community consensus be considered self granted? Perhaps, but it is just bickering over technicalities as far as i'm concerned. There was a community need and BAG formed to address that need and it was accepted by the community. If anything, it was formed over the objections that bots should be totally unregulated. Bot policy has been considered a Wikipedia policy for quite some time with the community staunchly supporting it. I was not involved with various stages of bot approvals, so I would have to research it to find out more details as to whether it was technically self granted or not, but it clearly had community consensus.
As for adding new members, I proposed a more bueaucratic approach, but no one wants to overcomplicate things in the name of perceived "consensus". A simple "up/down vote" has always worked with those receiving unanimous support being appointed. Community members always contribute to such votes. -- RM 12:40, 23 April 2007 (UTC)
I did some quick research into the history of the bot and here are the long existing instructions for bot owners:
Before running a bot, you must get approval on Wikipedia talk:Bots. State there precisely what the bot will do. Get a rough consensus on the talk page that it is a good idea. Wait a week to see if there are any objections, and if there aren't, go ahead and run it for a short period so it can be monitored. After this period, you should ask that the user be marked as a bot at m:requests for bot status.
AzaToth on March 13, 2006 formalized the process at Wikipedia:Bots/Requests_for_approval. As can be seen it was always a relatively informal process, but it was always approved by the community as far as I can tell. -- RM 12:53, 23 April 2007 (UTC)

Ok thats doable, I'd be ok with this if we can find a way to fix the issue that this current Cydebot thing brought up. Namely why did Cyde not request approval by the existing method to run his bot? Was it too complicated? Whats the deal with this bit? —— Eagle101 Need help? 12:46, 23 April 2007 (UTC)

(In reply to RM's second posting on research) Yeah, I was proposing possibly going back to that old standard... perhaps make that the standard for approved (figure out the meaning of this at a later time ) bot ops? Then they can just drop new tasks there without the hassle of getting approved for every new trivial task. —— Eagle101 Need help? 12:59, 23 April 2007 (UTC)
I'm not even really aware of the problems and it will be take me some time read and respond to the issue. I started reading on this page and responded accordingly. I just glanced at the whole Cydebot issue and it seems crazy, so I'll have a busy time for sure. -- RM 12:58, 23 April 2007 (UTC)
RM, most of it is on WP:BOWN. :) —— Eagle101 Need help? 12:59, 23 April 2007 (UTC)
Yeah I gathered that. I just came back from vacation and started on this page. -- RM 16:32, 23 April 2007 (UTC)
Sorry you had to come back to this, but I hope you had a nice holiday! --kingboyk 16:39, 23 April 2007 (UTC)

On the Cydebot blocking

Besides discussing whether or not BAG needs reform, and how to achieve it, I'd like to get some feedback on what happened, now that the dust is beginning to settle. Not least, I'd like to hear from the BAG members who weren't around when this incident occurred.

Was the original block correct or not, and why? What about endorsing the block? My unblocking? Was the behaviour of the advocates of Cydebot (not Cyde himself, who has been respectfully quiet on the issue) acceptable? Are BAG members competent? What role should IRC play? All and any comments on these and related issues are welcome. Thank you. --kingboyk 15:51, 23 April 2007 (UTC)

Ok, just to give you guys some background as to what happened. This bot was running from 15:40, April 19, 2007 (hist) (diff) m Image:1943 arcade.png (Replacing template name per Wikipedia:Non-free content/templates.). So its been going for about 4 days now, Xaosflux only noticed it on day 3 of the tasking. Of course nobody could have known the start time without digging it out like I did :), but this may or may not be useful. (See: for the starting contribs [10]) In otherwords From start of tasking to the actual block there was ~3 days, but the blocker really had no way of knowing this, as it took me a good 10-15 minutes to find where it actually started. —— Eagle101 Need help? 15:56, 23 April 2007 (UTC)
Thanks, but I'm looking more for opinions rather than data; things we can learn from. --kingboyk 15:58, 23 April 2007 (UTC)

Well, I'm not BAG, but I've been watching for a while and here's my opinion:

  • Cyde realized that his bot was going to take a long time to complete its task and needed a higher edir rate. It's against policy, so BAG would never have approved it, so he went to the devs. He was told to edit once per second or with a delay of one operation, whichever was slower. This was an acceptable way to proceed.
  • He ran the bot at this higher edit rate, and appearantly even above the dev's rate. Noone was aware of this authorization, which presumably happened on IRC, and in the opinion of most admins, the bot was violating policy. It was wrong, but excusable, to run the bot at this higher rate without posting on the BRFA or the bot's talk page, preferably both. It was also wrong to violate the dev's limit.
  • The bot was blocked by those tasked with protecting the wiki from rogue bots. This was a wrong action, in hindsight, but is acceptable given the circumstances.
  • Cyde returned and said he had approval to run at the higher rate. I do not believe he stated at first that the person who approved was a dev. The bot was unblocked, also acceptably.

All that happened was that there was a communications breakdown. IRC is an excellent tool for getting opinions, but is not a place to get approval for tasks unless logs are posted in a relevant location. ST47Talk 16:17, 23 April 2007 (UTC)

Thanks for the helpful post. Had I been involved in the process and had Cyde posted such a request, I would have probably approved a faster than 15 edits/minute request. Of all people I understand the need for such things. It would have been nice if the dev would have posted here at the time. Now that we know, we can update policy to match that information. IRC is a terrible place to discuss anything that isn't informal. The formal bot approvals process serves the community by relieving them from having to determine whether or not bots are ok. This simplification saves a lot of effort. Bypassing it isn't helpful. -- RM 16:41, 23 April 2007 (UTC)
That sounds about right ST47. Now should we try to contact Cyde and get his version of why he did not want to do a BRFA, rather then attempting to second guess this? —— Eagle101 Need help? 17:54, 23 April 2007 (UTC)
No need to contact me, I found the question myself. There are a variety of reasons why I didn't bother with BFRA, all of them playing some part. 1) I didn't feel that it was necessary; Cydebot is already bot-approved, and this task was about as foolproof as any task could ever be. 2) I know what I'm doing, both as pyWikipediaBot developer and operator of one of the bots on Wikipedia with the most edits. 3) Frankly, I should be in BAG, assuming such a thing as BAG should exist. 4) I dislike unnecessary bureaucracy and appreciate the attitude of "If it's not broken, don't make a big deal about it". 5) I'm lazy, and the BFRA process has become a rather large chore. I remember when you just had to add a section to a project page describing the bot's task. Now there's templates and template parameters and subpages and transclusion ... it's messy. I literally spent more time writing, figuring out, and filing the BRFA for the listification of categories task than I spent writing the actual bot itself. --Cyde Weys 19:56, 26 April 2007 (UTC)
Aside from the aspects that a developer was providing input and that was not stated here, I agree with everything that you said, for the most part. I've often worked that way in the past, so I understand it. When operating the rambot, my bot page just had a list of outstanding tasks to do, and everyone just trusted me to be useful because I was so useful in the past. Also in the past when I made mistakes, I fixed them all myself. There was no need for an approvals group. But Wikipedia has grown and at this point it's important for there to at least be a minor post even to a talk page indicating what happened. Look, there are many user such as myself or you that have bots that perform large numbers of useful edits and are trusted users. There are other users with approved tasks that are not well established and less trusted. It's complicated trying to manage them without playing favorites. Bureaucracy is useful but it hurts those that don't need it. Anyway, thanks for your comments. -- RM 20:10, 26 April 2007 (UTC)
I think that's all fairly reasonable except the bit that you should be in BAG. If you had wanted to be, you could have put your name forward. If you want to be in the future, if BAG is retained, you have to get some sort of community approval like everybody else! And I mean on wiki community approval. --kingboyk 20:13, 26 April 2007 (UTC) (e/c)

Conclusion

Debate seems to have dried up. We should probably try and reach a conclusion now. --kingboyk 12:10, 26 April 2007 (UTC)

Approvals for approvals group

Can someone please explain to me (yes I've read the above discussion) how having a bureaucracy to grant people the right to vote on whether a bot is allowable or not is a) a good thing and b) helps foster participation? --Durin 15:39, 26 April 2007 (UTC)

Sure, I'll take a stab at this.
The first question is much easier to answer. The reason is really very simple on the surface. Bots having bot flags can run "under the radar" by having their changes hidden from Recent Changes patrollers. This means that they can cause large amounts of damage before it is caught. In practice malfunctioning bots (or their operators) often do not get caught until 24 - 96 hours have gone by. Managing bot flags minimizes potential damage. This aspect is clearly required by community consensus, and is a good thing. The other aspect of approvals is enforcing policy. The benefit to Wikipedia is that tasks that seem innocent are not attempted. It's probably cliche at this point, but if there were no approvals, dozens of people would be running automated spell checking bots. These bots would be changing legitimate spelling mistakes all over the place and as a result introducing very difficult to detect errors. Community consensus, at least at the time of bot policy formation, was that all bots need to be approved for the reason that people do not want them running unchecked since they are prone to both normal human editing error and the additional error of programming coupled with a high speed of editing. Approvals also ensure that we try to enforce restrictions caused by server load.
The second question is a bit harder to answer because participation is second to the policy issues above. This process is designed to be as participatory as possible. Anyone can comment and their opinions given weight. It is much like RfA where average users can comment, but they can't determine consensus and perform the promotions. The fact is that many people are intimidated and no matter what we say or do, participation remains low. I think everyone wishes that it would be different, but that's just the way it is. It's a lack of interest more than anything, I think. -- RM 16:30, 26 April 2007 (UTC)
I'm not asking about the need for an approvals group. I appreciate that explanation, but I understand that. I am quite frankly very confused as to how creating a bureaucracy to approve people to being in the BAG can somehow improve participation. You don't add barriers to contributing in order to increase contributions. I'm at a loss here. If anyone can comment, and their voice be given weight, why do we need a group? Why can't a bureaucrat decide consensus? --Durin 16:40, 26 April 2007 (UTC)
I'm not sure they want to. Perhaps somebody can ask them. The crats seem happy to delegate this to BAG. Arbcom too. --kingboyk 16:43, 26 April 2007 (UTC)
Also, I'm a bit confused by your answer: "I'm not asking about the need for an approvals group"; "why do we need a group?". And, the question seems a bit loaded. The volunteers who perform this role for the community shouldn't have to justify it. If the community doesn't want a Bot Approvals Group, let the community say so, but until then, let's thank them for the job they're doing (and doing well imho). --kingboyk 16:46, 26 April 2007 (UTC)
Why does someone need to be approved to join the group? --Durin 17:01, 26 April 2007 (UTC)
Let me ask you a question? (not loaded or rhetorical, I truly want to know) :) Do you think that the "group" should be "anyone who wants to join"? --kingboyk 17:05, 26 April 2007 (UTC)
Anyone should be able to contribute towards consensus generation. It shouldn't require approvals in order to get voting rights. Wikipedia isn't a democracy; we work on consensus development. It seems very odd to me that we have a voting system (must be unanimous) in order to grant the right for 9 people to vote...when we're supposed to be about consensus. If I have an opinion on a bot, I shouldn't need to get approved to be in the BAG in order to voice that opinion and have it carry weight. If I don't need to be in the group to have that, then why does the group exist? --Durin 17:08, 26 April 2007 (UTC)
The BAG member makes a decision and closes the debate, the decision is then implemented by a crat (if a flag is required), usually without further question. --kingboyk 17:14, 26 April 2007 (UTC)
Why the bar to admission? Why can't the community do this on its own? What are the qualifications necessary to join BAG? --Durin 17:15, 26 April 2007 (UTC)
I'm not going to answer these questions, sorry, because I resigned from the group and questioned whether indeed it does have consensus. I'm sure somebody else will be along later to answer.
What I will say is that this is a pretty thankless task being performed by several very able people (experienced bot ops, AWB devs, one Mediawiki dev, most of them admins); without their work there probably wouldn't be a process for approving bots (that might be good or bad, I don't know). I respect your right to prod and probe but please do remember to assume good faith about the work that goes on here :) --kingboyk 17:20, 26 April 2007 (UTC)
  • de-indent Ok, to be clear; I am in no way questioning the good work of the people who delve into this area of Wikipedia. There are numerous thankless tasks on Wikipedia (Lord knows I've done some myself). Just because something is generally thankless doesn't mean it's bad, and I do not mean to cast any aspersion upon those that take on this work.
  • I will say that I've previously run into the thinking that because a job is thankless, needed, etc. that we need to have a group in order to support the operation. I've seen this torn down in a few places and what came after it ended up working fine because the community is the most powerful thing about this project. Things that work against the community; i.e. groups with bars to admission, actively quash that which is strongest about us.
  • So, I'm still at a loss as to understand how have a group necessitate an approvals process to even be in it can somehow encourage contribution to it. In the last three months, there's been 29 contributors to this particular talk page. Only 9 people are on the BAG, less than a third. Why aren't all people who want to help "allowed" to be on it and why do I even need permission in order to say "There's no way I could support this bot" and have a voice in it? It doesn't make sense.
  • I'm open to input on this; I'm not a member of BAG, and have barely contributed here in the past so I'm quite willing to listen to input. So far, I'm still at a loss as to understand how this bureaucracy is helping, especially any factor more than it is hurting. --Durin 17:50, 26 April 2007 (UTC)
Anyone CAN contribute to the discussions, I think that's been made clear. But BAG is needed because otherwise, who is in charge? If there is a conflict over a bot, such as birthdaybot, then we can't allow the first user who comes along to say "You're approved", instead with a group of 9, or however many, who do this every day, we can keep the system running smoothly, and those people are those who have been reading and contributing to bot approvals, not those who come across a page and have an opinion. ST47Talk 18:11, 26 April 2007 (UTC)
  • So why not allow anyone to have authority to close? If there's no consensus to close, then it can be prevented or undone. I just don't see the point to this bureaucracy. Just add a section to the end of each approval "Any objections to closing?" or some such. You don't need these barriers to membership. It certainly does not improve the possibility of people contributing. Instead, it raises barriers and actively discourages people from participating. Also, the requirement of unanimous approval to join BAG is extremely, highly cabal-ish. You've created an exclusive group with supposed authority to approve bots or not, and then restricted membership to only people whom all of your club agrees should be included. This is mind boggling. What point is there to all this bureaucracy? You can do it a heck of lot easier without all of this bureaucracy. --Durin 18:16, 26 April 2007 (UTC)
(ec)Anyone can contribute towards consensus, however, unlike RfA where people's opinions are largely arbitrary and given equal weight, bot approvals are largely a technical reason. If 10 people say a spellbot is a great idea, to be honest, it doesn't matter what they think. The ideas are almost always bad (although some good ideas have been proposed but not followed through to completion). Anyone can vote, regardless of BAG membership. The "unamimous vote" system is not a standard. It's just descriptive of what has happened in the past. Consensus has always been obvious. In cases where it was not unanimous a bureaucrat determined consensus. There is no problem here. As stated above, the group exists because of the technical background and understanding of policy. As far as I know, there have never been any serious problems with BAG members determining consensus on bot approvals. -- RM 18:18, 26 April 2007 (UTC)
  • So a qualification of BAG membership is technical background? If you lack coding background, can you become a member of BAG?
  • Two, I still don't see the need for bureaucracy here. Maybe I'm missing it, but I'm not seeing a reason why the community can't make a call on whether something has consensus or not. Why is this vested in a highly exclusive group that prevents membership to anyone whom they don't personally approve of? --Durin 18:20, 26 April 2007 (UTC)
That's not correct. Anybody can take part in the !election to BAG; at the ST47 !election, Betacommand (talk · contribs), Snowolf (talk · contribs), Extranet (talk · contribs), took part and aren't members. The application to join was advertised on the bot owners' noticeboard and at the village pump (technical) (I'll find diffs if you want them, but assume you will trust me on this :)), and people encouraged to participate. It's not BAG's fault that only 3 non members took up the invite. --kingboyk 18:38, 26 April 2007 (UTC)
Sorry,didn't answer you fully.
One. Anybody can become a member. You could nominate yourself now, go announce it at a few places (announce not canvass, of course ;)), and if you get support you're a member. I don't know if every member ever has been a coder, but they're generally techies (bot ops, coders, assorted geeks).
Two. BAG looks at the technical issues if there are any, and consider consensus. However, I would have to say that in my time at BAG not once was I asked to look at any code. I don't see any reason why any clued up user couldn't do it, tbh. I don't see any reason why you couldn't do it. So, you may well have a point. --kingboyk 18:48, 26 April 2007 (UTC)
I see no reason why it can't be a community vote. I've encourage outside participation during the first larger "election", closed by Redux. I have no idea why only BAG members would be allowed to vote. I do think that techical experience with bots and familarity with WP:BOT is a requirement. These elections should be held at WP:VP or WP:CN. Voice-of-All 18:29, 26 April 2007 (UTC)
  • I don't think there should be elections at all. If we can't trust our community with helping form consensus about bots, we shouldn't trust them with the mainspace either. No technical experience should be required. Lots of opinions regarding how a bot functions can be had without having the first inkling of how the code works. These opinions are every bit as important as technical opinions. Restricting input from purely technical sources quashes legitimate consensus forming. This process is broken. --Durin 18:33, 26 April 2007 (UTC)
  • Haha, yes consensus would be ideal. I am not sure how well that would work on a large site like this, but I suppose it's worth a try. Voice-of-All 18:35, 26 April 2007 (UTC)
  • (ec)Too much is being read into what I said about "technical requirements". Technical understanding is one aspect that is important. When bot policy was first formed, any admin had to and could approve of a bot. Let me ask the question: Why do are admins given extra buttons and not everyone? Why are bureaucrats given the responsibility of determining consensus in important decisions? The basic reasons are trust and technical expertise. We do not trust all users equally, and nor should we. People with more experience determine certain things (I'm speaking generally here). In the specific case of bots, someone with enough understanding must give the ok, just like we do in many other processes. -- RM 18:41, 26 April 2007 (UTC)
As long as some members have technical experience and the others have access to it or discuss together, I guess we could have some only check policy issues. Fair enough. Voice-of-All 18:40, 26 April 2007 (UTC)
If the process is broken, let's try to fix it. (However, please note that the discussion already began on that but died down; can we perhaps resume the old debate so we don't have to go over old ground? :)) --kingboyk 18:42, 26 April 2007 (UTC) (e/c)

(deindent) I don't feel like posting in a number of places only to have edit conflicts kill me, so let me add some general points: Anyone can contribute to approvals and anyone can vote on new BAG members. I'd say more people vote on BAG members than do on approvals. We always get a fair number of "outside" views when new members are added. You can look at the archives for examples. This seems to indicate that the community cares mostly about delegating this responsibility and is happy with that. As for contributing to approvals, we generally only get input from the community in general in specific instances. Someone from a WikiProject may come along because they care about this bot or that bot, but widely ignore the other bots. They don't care about bots in general, only about specific bots. This has worked well in the past. The community contributes if and when it cares to do so, which is only in certain situations. Some bots, like one of the proposed admin bots, got a huge community response. The system works when it needs to work. So far the only arguments against this process seem to be theoretical. I don't know of any practical real problems with the process. As for requirements for BAG membership, there are no set requirements, just like there are no set requirements for an RfA. You don't have to be a programmer or even run a bot, that is just helpful for understanding. Far more important, in my opinion, is general policy understanding. There are so many pitfalls that a bot can have that prevent its approval. Just look through all the withdrawn and failed requests, and you'll see what I mean. The bot approvals process is time consuming. It often requires users to look deeply at contributions, do research into the topic in question, and build on previous experience. It's not a matter of simply saying "I like this idea" or "I don't like this idea". If it was that simple, we wouldn't need an approvals group. I like a lot of ideas that are rejected! In the ideal world there would be so many comments on bot approvals from the community that the only purpose of BAG would be to tally up !votes and determine consensus. In that case, the reason to have a bot approvals group would then be to evaulate whether or not those individuals would be qualified to determine consensus. (Bureaucrats are small in number and very busy: there has so far been no problems with BAG members determining consensus, so there is no reason to escalate that to a "higher level", thus decreasing the amount of general participation). -- RM 18:56, 26 April 2007 (UTC)

I have some additional thoughts on adding BAG members. Part of the reason BAG members are voted in is because many many bot requests are handled by a single person. We often will speedily approve entries. Because of this, we need to ensure that the people doing this are qualified because they have to understand previous precedent and so forth. As far as I know there have been very few, if any problems, with this because we don't allow just anyone to do this. Requiring a bureaucrat to determine bot consensus would only limit community participation. Going back to a system where any admin can approve a bot would be an interesting choice, although I'm not sure I'd be comfortable with any admin granting a speedy approval. The requirements for being an admin are not the same as those for being in BAG. Admins focus on different areas of expertise. Let me be absolutely clear: everyone has a voice in approvals and the purpose of being a BAG member is primarily to determine consensus and act as a technical safeguard. What happens in practice is that BAG members assume all roles because the community in general refuses to take part in the most important aspect: approvals. -- RM 19:15, 26 April 2007 (UTC)

Durin, Let me try to make a clear statement about Bots and BAG. I have seen bot request for approval that had support but was clearly not appropriate. If my memory is correct we had a request for a bot that was supposed to bypass redirects. It had a good amount of support and would have been operating except for the fact that the policy about redirects is dont bypass them if they are not broken. I have also seen the re-accruing WelcomeBot, several groups of users want it, yet the community has said NO repeatedly. BAG exist to be what B'crats are to RfA's for Bots. I have no objection for you joining BAG except that you havent commented and been part of any approval's yet. Hell when I joined BAG I couldnt program. I used AWB for all my bot stuff. As long as a user knows bot policy and has some common sense I see no reason that they shouldn't be part of BAG. yes people run unapproved bots, that happens for a few reasons, A. they want to cause trouble (spambot/vandalbot) B. their is not a community consensus for the task (WelcomeBot) C. the community has a admin bot phobia for the most part (CydeBot's CfD deletions). But WP:BFRA exist to try and prevent damage that some people would do by accident, who are acting in good faith (spellchecking bots). As for admittance into BAG, all that is needed is a showing of community trust for the user and that they know Bot policy. I tried 6 months ago to get more people involved into the process, I did get a few, but for the most part the community gives BRFA to the BAG, BAG has and wants more community consensus. Betacommand (talkcontribsBot) 01:19, 27 April 2007 (UTC)

This group is already dead

People are running bots without approval already. :-/ --Kim Bruning 18:05, 26 April 2007 (UTC)

Then they should be blocked per policy. Who are we talking about? ST47Talk 18:11, 26 April 2007 (UTC)
No one has a flagged bot which has not, at some point, been approved (AFAIK). The bigger area where the approval policy is ignored is with regards to additional tasks, in qhich case I'm in favour of a quick note to the op's talk page asking them to stop and get approval. If the task could be contentious, or the bot is making errors, it should be blocked without further consideration. Martinp23 18:15, 26 April 2007 (UTC)
That's the theory yes.
However, I was told by one person who left wikipedia that they had been using bots for admin tasks without approval. Seeing that theoretical per-admin workload has increased by 100% or thereabouts in the past year, I'm thinking that this person probably isn't the only one. --Kim Bruning 18:27, 26 April 2007 (UTC)
Admin bots are a grey area of bot policy anyway. If I wasn't so busy, I'd be working on forming a better consensus for that sort of thing. The example is an exception to the rule, and I wouldn't consider it to be representative of the whole process. The point is that they are not using the bot flag and if their usage was detected (which is often very difficult), they would be blocked. Just because someone can get away with breaking the rules doesn't mean that they should! -- RM 18:32, 26 April 2007 (UTC)
Wikipedia:Requests for arbitration/Betacommand/Proposed decision is considering these issues at the moment, and nearly finalised. --kingboyk 18:33, 26 April 2007 (UTC)
I just finished reading the proposed decision, and I'm not sure that the proposed decision on admin bots is even going to pass anyway. If it does pass it, that would clear up some issues. I've been arguing that the community has previously spoken on this issue of admin bots, but it was still unclear. An ArbCom ruling would be good confirmation, but I'd be happier if it was unanimous. (seems ironic if it isn't!) -- RM 19:35, 26 April 2007 (UTC)

I'm concerned about whether there is so much bureaucracy and catch22 that people are starting to bypass the lot. Once they start doing so and lie about it, we'd basically have no feedback, and we'd be fulfilling both the requirements to be in a state somewhat akin to freefall. (as we'd have "no hard ground beneath our feet").

What I'm wondering is whether the en.wikipedia community is really not in any kind of trouble whatsoever, or if in fact we are already currently in the freefall state.

We can wait for a year or so and find out for sure by the hard stop at the end (or hopefully lack of it) , but perhaps it's a good idea to keep our eyes pealed to see if we can catch clues either way.

At any rate the person I talked with did actually admit to having lied at the time (they have since left wikipedia). Hence my worry.

--Kim Bruning 18:56, 26 April 2007 (UTC)

Have now read the RFAr on Betacommand. He was stupid and got caught. That's a 2nd datapoint. --Kim Bruning 19:00, 26 April 2007 (UTC) gulp.
I ran a bot on my main user account once without permission. Sure permission wasn't required at the time (predated BAG), but it isn't terribly difficult to do. I don't think bot policy is broken just because people can lie about it. That's always been the case. People can lie and they will continue to do so. The approvals process requires good faith on all sides. Bot operators have to realize that the goal of bot approvals is to make sure that bots are good and useful. If they choose to operate in bad faith, what are we to do? We are not left with many options but blocking and other forms of sanction. In theory I don't have a problem with a user bypassing approvals and running covert bot operations, if they are doing it in a constructive fashion just to avoid bureaucracy. The problem is that with a community this large we need (or want) to know what bots are doing so we can prevent intentional or unintentional vandalism. Vandalism is a real problem and the whole underlying reason behind what we do here. -- RM 19:22, 26 April 2007 (UTC)


What? Did I really just see that? Did you really just say that you don't have a problem with someone running a covert bot without approval? Let me clearly go on record: I do. Although I'm certain that some folks here can, *I* can not think of a situation where IAR would apply to that. BAG, and the bot approval process are here to protect the database, and that's one of my primary concerns here. I'm concerned that there are people turning a blind eye to that. Philippe 02:20, 27 April 2007 (UTC)
Let me be a little more clear: What I said regarded theory, not practice. Back before bot policy I used to run a bot without much of any oversight whatsoever. I acted in good faith and resolved problems as they were brought to my attention. I was able to work efficiently and quickly. In the ideal world (that is, in theory), bots could operate without having to go through the muck of approvals. Unfortunately that way of thinking doesn't scale. As Wikipedia grew, the need for more oversight increased. Alas, we can't operate in the theoretically ideal world, so we have in practice this whole approvals process. I joined the approvals process to make sure that things went as smoothly as possible without trying to restrict bot operators beyond reasonable amounts. Maybe this helps clear up my opnion. I'm not endorsing blanket ignoring of the rules, although there are certain instances where ignoring approvals is merited, but those examples are not common. -- RM 02:26, 27 April 2007 (UTC)
Given the theoretical caveat, I understand a touch better. I get nervous when members of BAG say something that could be construed as "ignore BAG"... I know that's not what you meant, but was hoping to draw out your statement a bit, and you obliged me, for which I thank you. Philippe 02:31, 27 April 2007 (UTC)
I'm a strong advocate of IAR, but as RM says it doesn't scale well with regards to bots, and think that bot policy is one set of rules best not ignored. Just my 2c. --kingboyk 12:30, 27 April 2007 (UTC)
If you have to ignore the rules, at least have the courtesy take a minute and post to Wikipedia talk:Bots/Requests for approval explaining what you are about to do. Whenever a bot runs covertly and people figure it out, it generates a flurry of activity and confusion that could have been avoided with a little bit of openness. In the case of betacommand running "unapproved" bot tasks, people didn't have a BRFA to reference showing where certain tasks were or were not approved. It isn't about BAG, it's about the community: They want to know what tasks are approved so they can find out quickly. Ignoring the rules usually just causes a lot of administrative trouble. -- RM 12:45, 27 April 2007 (UTC)

Complaining vs. Helping

I've been neglecting approvals. I should probably be removed as a BAG member because I've been busy doing other things, some of which is defending BAG instead of getting useful work done. 90% of the approvals process is open to everyone. Everyone. About 10% of the job is determining consensus and having the technical and policy understanding to recognize if a bot is a bad idea even if it has approval otherwise. People seem to be ignoring the fact that no one is helping with the most important aspect and instead complaining about the 10% that is working out pretty well. It's the 90% that is broken: we don't have enough people commenting so BAG members get bogged down. BAG members are really good at the 10% part: technically knowledgeable and really good at determining consensus. As far as I know, we've had no complaints about what we've actually done. If everyone, including myself, would just shutup and help, there wouldn't be a problem. Anyone can join BAG if they prove themselves competant with the other 90% of the work. Why they won't do that is beyond me. -- RM 20:55, 26 April 2007 (UTC)

Round of applause. --kingboyk 21:01, 26 April 2007 (UTC)

The proposed solution, getting rid of BAG or changing it radically, does not help. It won't bring in more helpers and it will only eliminate the 10% that was working and make it worse because now instead of BAG doing that 10%, a busy bureaucrat will have to do it. And let me tell you, becoming a bureacrat is extremely difficult and almost impossible: I've tried. So the rest of the 90% will be far worse if we go that route, because delays will increase. The other proposed solution, letting anyone do approvals (the 10% part), is not feasible because it violates community consensus. The community for many years has spoken: it wants there to be restrictions on what bots can do and it wants there to be strong oversight. Letting anyone approve means that we'll let through bots that might otherwise have been rejected. Errors will increase in a process that had previously solved this. -- RM 21:02, 26 April 2007 (UTC)

I was wondering if the solution might be to appoint "bot crats", who close bot approval requests and flag bot accounts. Of course, this would require either feature seperation in Mediawiki, or trust that these "bot crats" wouldn't go around sysopping people. --kingboyk 21:06, 26 April 2007 (UTC)
That's basically what we already are. It would just be a change in name only and not satisfy the BAG detractors. Afterall, if these members had special priveledges, who would determine who was a member? It's the same as BAG, but membership would likely be even more difficult and/or elitist than it is (perceived to be) now. When I tried to become a bureaucrat to perform bot flagging, it became clear that *most* people opposed a bureaucrat with a single mission. It would have to be implemented as a separate feature. In the end, any "solution" will not solve any problems but create more problems because no proposed solution has addressed the issues of participation and speed of approvals (while satisfying community consensus for oversight). -- RM 21:44, 26 April 2007 (UTC)
This sounds like an attempt to preserve the fiefdom, and as such is not acceptable. I will not accept any resolution that grants special authority to members of this group. Kelly Martin (talk) 12:39, 27 April 2007 (UTC)
Fiefdom? This is just crazy. Most of the current members of BAG are new to this and the turnover rate for BAG membership is fairly high. Just look at who is contributing the most to approvals: those BAG members that are the newest. This is a not only a thankless job, it's also a hostile and stressful environment. Fiefdom! Even "ousted" member Betacommand is still involved with bots, despite being removed from BAG, so how is there a fiefdom? Check out his recent comment above. Look if there is a problem, consider proposing a realistic solution. But before you propose a solution, show specifically where the problem is. Find me some concrete examples where there are real problems with this process, other than speed of approvals and lack of community approvals caused by lack of interest. Your response shows a lack of understanding about where the consensus is on this issue. Just look at some of the comments generated in the miscellany for deletion discussion. Throw out the comments by current BAG members and you still have members of the community expressing that they strongly approve of this process. Those comments confirm what I've been saying all along: this process is not a covert one and has been operating for years. It is just not possible for it to be operating without consensus as it has visibility at almost every level. At the very minimum bots have always required approval from a specialized body, such as administrators. This has been codified as official Wikipedia policy and is being confirmed by ArbCom as we speak. That the administrators were replaced by BAG is no mistake: it was the natural evolution of the process that the community wanted stricter control over as Wikipedia grew in size. Adminship is no longer even a requirement, simplifying membership. -- RM 13:08, 27 April 2007 (UTC)
When you said "I will not accept any resolution" I presume you meant "I will oppose"? I'm not aware of any wiki policy which says that Kelly Martin has final veto over decisions made by the community... --kingboyk 13:16, 27 April 2007 (UTC)
That's what consensus means: every member of the community has a veto over any proposal made. Are you claiming that I am not a member of the community? Kelly Martin (talk) 14:44, 27 April 2007 (UTC)
That's an interesting interpretation of consensus. So you mean that even if you were the only person who disagreed with something, you would have a veto over it? We'd never make a decision if that were true. Danny's RFA had over 100 opposers, and yet the crats found consensus to promote. --kingboyk 14:49, 27 April 2007 (UTC)
It's the usual definition of consensus. What Wikipedia calls "consensus" is not consensus, but instead some sort of weird supermajority, basically along the lines of "move forward only if we feel that the objectors can be ignored". I think the proper term for this is ochlocracy. In the Danny RFA case, the b'crats decided to ignore the objections from the minority because they felt they could get away with it. I happen to agree with their decision in that case, but let's not pretend for a moment that that decision represents a "consensus of Wikipedians"; to call it that perverts the meaning of the word "consensus". Kelly Martin (talk) 15:01, 27 April 2007 (UTC)
I believe you're absolutely right, in terms of semantics. I'm still not seeing how you get a veto though. Wikipedia has changed; in particular, it's grown, and it needs processes which scale. One person can't throw a spanner in the works (unless they're Jimbo) and the old order don't get to run things any more just because they were here first. Likewise, a laissez faire approach to bots doesn't scale well either.
From what I can tell at the (unconcluded) MFD, the community does broadly support having a Bot Approvals Group. Arbcom is also in the process of delivering a judgement which confirms the BAG's mandate over bot approvals. Might we, then, accept that this group exists, is going to continue to exist for now, and has a small amount of power, and look instead at how we recruit to the group? --kingboyk 15:07, 27 April 2007 (UTC)
Wikipedia:Consensus#Consensus_vs._supermajority disagrees with the assertion that "What Wikipedia calls "consensus" is not consensus, but instead some sort of weird supermajority." Which interpretation is correct? Also, Wikipedia:Supermajority is listed as a rejected policy. --Elkman (Elkspeak) 15:20, 27 April 2007 (UTC)
Consensus does not mean everyone agrees, it means every wikipedian accepts it. Of course those Wikipedians that do not accept consensus are not Wikipedians for too long. HighInBC(Need help? Ask me) 14:53, 27 April 2007 (UTC)
Okay folks, this is not a power struggle. It makes sense that a certain group of people decides consensus with bots because the vast majority of Wikipedians to no understand how bots work or what the repercussions are. Fiefdom? I am sorry, but that has got to be one of the most silly statements I have heard in a while. This is a specialized field that needs people with specialized knowledge. HighInBC(Need help? Ask me) 13:20, 27 April 2007 (UTC)
If you think this isn't a power struggle, then either you're not paying attention or you are one of the people struggling to get or hold power. Kelly Martin (talk) 15:27, 27 April 2007 (UTC)
Granting them full bcrats rights is sloppy and unexceptable. However, giving them only the "makebot" right would require a "botOp" group. However, only stewards could grant that unless a "makebotOP" UI was made for bcrats. Crats already have the final say on flags, so requiring BAG approval does make an additional bureaucratic layer. Such can be removed by either giving crats the authority or BAG the authority alone. At any rate, if BAG is to have software hard-rights then "elections" need to be open to the community and needs to have significant input to pass (not just 5 comments by BAG or something). Voice-of-All 14:52, 27 April 2007 (UTC)
As I said before, the crats seem happy to delegate this role to BAG. (I wish some of them would come comment, btw). I'm personally far from adverse to making appointment to BAG a more rigorous and more widely advertised process (perhaps at WP:RFA) and giving BAG members makebot priveledges, if the powers that be would be willing to implement the software change. What, exactly, are your thoughts on this? You don't really appear to be offering much of an opinion at the moment :) --kingboyk 15:10, 27 April 2007 (UTC)
The current flagging system works fine. Flags are almost always set by bureaucrats within 24 hours and within an hour or two. It would be convienant if BAG had the power to do it, but it's not high priority considering all the other issues. It's a source of delay, but not the biggest one. In this case the additional bureaucratic layer adds accountability. A bureaucrat must at least check to make sure bot approvals are somewhat legitimate. I would consider any system where BAG members had a "makebot" right to be optional and it should be a subset of the RfA/RfB process. Not all BAG members need or should have the "makebot" ability, only those that are confirmed through the scrutiny provided by the RfA/RfB forum. Any other system creates a process where BAG members must be added through a more complicated or retricted vote that limits general inclusion. I'd say that it's easier to get into the so-called "fiefdom" now than it would be to go through a more formal RfA-type process. -- RM 15:50, 27 April 2007 (UTC)
I'd largely agree with that evaluation. The tradeoff, though, is that "botcrats"/mini-BCs might be seen as more clearly delegated and authorised by the community at large than the BAG is (at least in some quarters). Now of course, the irony is that in order to implement such a thing, it'd have to be run by the devs... and I wouldn't be in the least surprised if they transpired to be profoundly apathetic about the whole thing. Alai 18:23, 3 May 2007 (UTC)

Power struggle

I'm personally shocked that some people in this community think that the BAG is simply a power struggle to gain control over bots, which, as has been said before, is not true. Kelly Martin, aside from abolishing the group, which the community has basically voted against, is there anything we can do to affirm that we are completely not trying to gain or hold power? How about new elections, possibly on a wide-reaching forum such as RfA to confirm each member's entry and promote new entrants? How about two sections on bot approvals, clearly separated. One section would be ONLY for talking about technical aspects, such as edit rate, server load, etc, and in that section final say would be given to members of the BAG (although as always anyone is welcome to comment there), as it has already been decided that each of those members has the necessary technical knowledge for such a thing. Then there would be another section for talking about everything other than the technical aspects, mainly consensus for the task the bot plants to do. If EITHER section failed, then the entire approval would fail. The second section would be without prejudice to BAG members, and an uninvolved BAG member or any other administrator would be required to gauge consensus and close the debate, as technical knowledge is not necessary for that section. I would gladly agree to either or both, or pretty much anything else that would help us end this quarreling and get back to the encyclopedia. —METS501 (talk) 18:44, 27 April 2007 (UTC)

<sarcasm>You could redirect the page to User talk:Kelly Martin perhaps?</sarcasm> --kingboyk 18:47, 27 April 2007 (UTC)
Setting aside the useless and uncivil suggestion from kingboyk, I continue to advise that this group be renamed to the Bot Advisory Group and that it operate as a resource group for those who run or wish to run bots, with members free to join and leave at will; there is no need for elections to such a group. The decision whether or not to grant a flag to any given bot should be an open community process similar to the request for adminship process; input from members of the Advisory Group would likely carry significant weight (depending, of course, on the bureaucrat's evaluation of the competency of that particular member) but the final decision rests with the closing bureaucrat (as it always has), not with the Advisory Group. I absolute oppose the process-heavy, status-driven approach recommended above. Kelly Martin (talk) 19:51, 29 April 2007 (UTC)
Then I'm afraid that we're going to be unable to compromise for the time being. We've agreed (I think basically everyone except you) that open membership doesn't work. —METS501 (talk) 22:25, 29 April 2007 (UTC)

Policy rewrite proposal

I have started a (hopefully insightful) discussion on the possible rewrite of our bot policy here. Comments are welcome. Миша13 09:51, 28 April 2007 (UTC)

A note on MFD and reform

I have recently closed the BAG MFD as "reform", as that is generally what the community has called for. Some of the main concerns raised were:

  1. The name "Approvals group" gives the impression that this process is regulatory in nature, instead of the "Advisory group" as it should be. Several users have toyed with the idea of moving to Wikipedia:Bots/Advisory group.
  2. Several users noted that the regulation of BAG membership has made it into a "cabal". Nobody has proposed any method to fixing that, other than moving to a more "RFC style" Bot request for approval. Concerns were expressed regarding the control of bots on Wikipedia by a select few.
  3. People have requested that the process become "less formal, less bureaucratic" and into something more open to the community at large.

Of course, some people also expressed the sentiment that it isn't broken, and doesn't need to be fixed. I wish the community and BAG luck to whatever you plan on doing. This is a cross post to multiple venues. Sean William 14:00, 1 May 2007 (UTC)

A few generic "reform" messages are hardly grounds for a major change, especially if that change clearly violates the spirit of what everyone else said. I want to be clear that the MfD was quite useful for soliciting community opinion, but it is not a place for changing bot policy. The only changes that could be authorized by that process are changes to BAG itself, which technically falls somewhat on the fringe or outside of core policy.
  1. I did a quick head count and 6 out of 34 commenters supported a change to "Advisory Group", and it was not clear if they all thought it was necessary (a few were conditional) or whether they agree on whether it is a name change or also a procedural change. I don't support the changing of names just for the sake of changing names. If people are confused by "Approvals Group", they will be equally confused by "Advisory Group". It's a matter of semantics. You have to actually see what it is about before making any kind of judgement. The reason they get the impression that the approvals group is regulatory in nature is because it is. BAG has a community mandate to regulate bots based on bot policy, overall Wikipedia policy, and individual consensus for any given task. Trying to hide its regulatory nature by changing its name is spurious at best. I have nothing to hide, and changing the name is not helpful.
  2. You mean the cabal with the most active members being a couple weeks or months old? That's a rediculous allegation that doesn't match the facts.
  3. The process is already open to the community for 90% of the work. We do not need reform because commentors on the MfD didn't bother to research how the process actually works. The community, including most of those who commented on the MfD, does not participate actively by its own choice. I am unaware of any propoosal that combines "less formal, less bureaucratic" with the community regulatory mandate.
-- RM 15:12, 1 May 2007 (UTC)
One of the core issues surrounding this problem is that an "unelected" (which is untrue) body is determining community consensus. Any calls for reform based on claims that the community doesn't have strong input on approvals is patently rediculous. Perhaps the community doesn't like that BAG determines consensus. Would they rather increase the bureaucracy by making a bureaucrat do that? Would they rather decrease the bureaucracy by not requiring a determination of consensus at all or by letting anyone determine consensus? Everyone is talking about reform, but no one is really showing where any real problems are. What are the problems? -- RM 15:37, 1 May 2007 (UTC)

Congratulations and keep up the good work!

Bot operators, BAG members, and the community: I am most pleased to inform you that the WP:BRFA page has the fewest amount of requests it has had in at least 2 months, which means that not only do BAG members have more time to work on specific requests, but, assuming a steady amount of new requests, our turnaround time is also going down. To BAG: keep up the good work! To everyone else: to help to determine consensus, find problems, ask the questions that need to be asked, and speed up and simplify this process even further, you can watch a status report at User:ST47/BAG. ST47Talk 17:47, 12 May 2007 (UTC)

Cocoabot

I can't flag this account. See the error: Image:Cocoabot.png. Would like someone to take this up. =Nichalp «Talk»= 15:19, 17 May 2007 (UTC)

Noted, the account doesn't appear to exist. I'll speak to the op. --ST47Talk 18:14, 17 May 2007 (UTC)
Thanks :) =Nichalp «Talk»= 18:26, 19 May 2007 (UTC)

New template - Operator assistance needed

I've implemented a new template, {{OperatorAssistanceNeeded}}, that will show up as   A BAG member has requested the attention of the operator. Once the user has seen this message and replied, please remove this tag.. If you use this, the bot will send a message to the op and change the tag to {{OperatorAssistanceNeeded|D}}. Enjoy! BAGBot 13:59, 19 May 2007 (UTC)

Oops, meant to post as myself. --ST47Talk 14:03, 19 May 2007 (UTC)