Wikipedia talk:Requests for arbitration/Betacommand/Workshop

"Sentence first - verdict afterwards"
I removed the two proposed findings related to bad username blocks because even their proposer has acknowledged there is no evidence (at least, not yet) for them. I don't mean to disrespect anyone's participation by quoting the Red Queen, but this is an example of a problem common to many arbitration cases when editors focus on the outcome (people proposing and defending against remedies) before figuring out exactly what happened. (In fact, I have considered proposing that the clerks in future cases not create the workshop page until after evidence has been placed on the evidence page, although this creates other problems and does not need to be debated here.) There is no predjudice against proposing new findings related to Betacommand's blocks (which did raise concerns on the noticeboard), but I suggest that someone first add evidence of the blocks to the evidence page, at which point findings of fact can be proposed that state clearly and accurately what he did, and why it was a problem. Thatcher131 18:22, 31 March 2007 (UTC)
 * No disrespect taken on my part, I had no idea it would be such a big deal, but now I know ^_^ Milto LOL pia 21:21, 31 March 2007 (UTC)

This is becoming a damp squib
I wonder if the arbitrators have made a mistake in accepting this case (or whether they ought now to close it). The likely outcome is censure, which the community is more than capable of doing it by itself. There seems to be little in the way of recent substantive problematic activity which wasn't dealt with by WP:BAG, and our removal of Betacommand from the group, removal of his bot flag (now reinstated), and the censure he has already received, was, I think, action enough. I appreciate that other activities of Betacommand's, such as username blocking, is up for review here, but I'm not seeing much of a case being made or remedies that require ArbCom to be involved. --kingboyk 15:44, 11 April 2007 (UTC)
 * There is a principle I'd really like arbcom to express, which is that users shouldn't interpret wp policy mechanistically. I wrote something to that effect in the ANI thread during Betacommand's link deletion rampage I'll try to put it here during the weekend and see if I can make it into a workshop proposal.  I've been too busy in RL to spend much time here during the week. 64.160.39.153 06:17, 13 April 2007 (UTC)

Lack of community confidence
I don't feel like hunting down diffs. But, it was clear on AN/I in the conversations that proceeded this case that several different editors had no confidence in his ability to use the admin tools reasonably. I was one of them; I don't remember who the others were but I remember seeing experienced editors agreeing. The ongoing pattern of poor judgment and an inability to explain his actions became apparent very soon after he was given the admin tools. In my opinion, this unfortunately slow and unwieldy process drives home the point that we ought to have a community process for desysopping. The community showed consensus for giving him the tools. Then, consensus changed. What could have been very simple turned into an Arbcom case. Friday (talk) 16:37, 14 April 2007 (UTC)


 * Don't want to get into much threaded discussion on the other page. I agree that Tony's objections are things that could happen.  But these are problems we already have.  We already disregard unreasonable arguments when determining consensus.   We already let "the mob" participate in RFA and the world hasn't fallen apart.  Friday (talk) 16:41, 14 April 2007 (UTC)

We don't allow "the mob" to participate in RFA. Insomuch as it does, this is without community consensus. We have serious problems in RFA. Adding to them by holding lynch-mob votes (which I'll remind you were strongly opposed at RFA when they were ordered by the arbitration committee in the Stevertigo case) would make Wikipedia a more ugly place, a haven for trolls. --Tony Sidaway 17:44, 14 April 2007 (UTC)
 * Nonsense. A "haven for trolls"? Like, trolls would come, set up accounts, make enough edits to be taken semi-seriously and then go get strangers desysopped as their primary goal?  So much trouble, and for what?  The internet is not festering with legions of trolls just waiting for the chance to take down Wikipedia.  The great majority of Wikipedia editors are good people, and any input on innovations or proposals that come from another mindset are worthless.  Milto LOL pia 21:45, 14 April 2007 (UTC)
 * One word: Quickpolls. --Tony Sidaway 12:23, 15 April 2007 (UTC)
 * Tony, Quickpolls were before many of our time (I've heard of them in passing only), so more than one word might be helpful in this instance. Newyorkbrad 12:44, 15 April 2007 (UTC)
 * Quickpolls was a one-month experiment in summary mob justice during the middle period of Wikipedia. --Tony Sidaway 12:59, 15 April 2007 (UTC)
 * Thanks. Interesting reading and background for us (relative) newbies. Newyorkbrad 14:32, 15 April 2007 (UTC)

Adminbots
At present there is really no guidance from the bot group on use of automated and semi-automated tools on admin accounts, even though we know it happens. There are a number of possible policy variations which I would like to suggest.


 * 1) Admins may not use bots or scripts on their accounts to perform automated or semi-automated blocks or deletions without approval.
 * 2) Admins may not use fully automated bots to perform admin actions on their admin accounts. They may use semi-automated bots and scripts provided that each action requires manual confirmation.  (Admins who use automated tools are responsible for all edits made by their tools.)
 * 3) Admins may not use fully automated bots to perform admin actions on their admin accounts. They may use semi-automated bots and scripts with the approval of the bot approvals group.
 * 4) Admins may use bots and scripts to assist with deletions provided that each action requires manual confirmation and that the edit rate does not exceed X edits per minute. Faster editing requires a bot-flagged admin account.

And there may be more variations. There are definitely already admins using scripts to make mass deletions easier. So I don't see why the Bot group can't step in an excercise some oversight. If there are going to be admin bots or scripts, there should be an approvals process to make sure they work properly and a mechanism to withdraw approval. Assuming Betacommand is telling the truth that his image deletion bot (November 2006) and his link removal tool (March 2007) were scripts and not fully automated bots, then nothing he did violates the bot policy as currently written. His edits were very fast but there is no specific guidance on when script-made edits need a bot flag and/or a separate account due to their speed, and there is nothing in the bot policy to prohibit using a buggy script (as opposed to a buggy bot). I really think the BAG needs to step in and make some rules here, and I don't see any major opposition to approval of at least scripts to perform deletions since that happens every day already. They need to be regulated. Thatcher131 16:16, 16 April 2007 (UTC)
 * Rules ought to be made by the community, and then implemented by BAG in a fair and consistent manner. We have no problem with doing the latter, and I personally would welcome the opportunity for us to approve admin bots. --kingboyk 12:53, 17 April 2007 (UTC) (BAG member)


 * I can see two problems- first, there should not be a strict ban on fully automated bots- as I said on the evidence page, there is sometimes a good use for an automated bot, such as deleting pages where only one user has edited, and has requested its deletion via db-author. Second, I don't believe it's technically possible for an admin account to also be a bot account.  Ral315 » 20:07, 17 April 2007 (UTC)
 * I have assigned privs on private wikis to give the sysop + bot flags to an account in Mediawiki - I'm sure it would work here too. As a member of the BAG, I feel that proposal 4 of Thatcher's proposals probably represent how the BAG would act in such a situation.  The standard is that, if a user edits too fast, and repeatedly they will be required to create a seperate account - I see no reason why admins should be treated differently, but clearly a policy needs to be formed by the community, and implemented by the BAG. Mart inp23  20:17, 17 April 2007 (UTC)


 * Just how often do db-author requests actually happen? The collateral damage from debugging a bot to service them, to say nothing of the development effort that could be put into more useful things, are probably worse than several years' worth of handling those requests manually. 75.62.7.22 05:08, 24 April 2007 (UTC)