Wikipedia:Requests for BAG membership/Ilmari Karonen


 * The following discussion is preserved as an archive of a request for bot approvals group membership that did not succeed. Please do not modify it.

I am withdrawing this request for BAG membership for the following reasons:
 * It's now been open for 10 days, twice the supposed length. It does not seem any bureaucrat is willing to close it (and I'm not blaming them for that).
 * Whether or not this method of applying for BAG membership actually enjoys consensus or not is unclear. As such, it's not surprising that 'crats don't want to touch it.  I was aware of this when I nominated myself, and stated my willingness to confirm my membership by other means even if this request was approved.
 * Given that the idea of discussing BAG membership at RfA was to increase visibility and community participation, I'm not all that convinced by the turnout. Had this nomination been closed when it was due, the tally would've been (14/5/4), including some !votes on both sides that I would've discounted if I were a 'crat.  I'm not sure that makes for much of a consensus.  An extra five days has brought 10 more comments, but it's still far short of e.g. Cobi's (43/0/0).

I thank everyone who commented, both in favor and in opposition to my application. Many of your comments have been very well thought out and informative, and have provided me (as well as, hopefully, anyone else who has followed the discussion) with a deeper insight into what the community expects of the BAG. Though a prolonged nomination ending in a withdrawal may seem somewhat disappointing, I do not feel this was by any means a waste of time, and I would like to hope neither do those of you who took your time to comment here.

One of the purposes I gave for nominating myself was to see what the !voter turnout in the "new system" would be, and what kind of discussion would be generated, for a non-obvious candidate, i.e. someone who wasn't already an established and well-respected member of the BAG and who had some controversial opinions and gaps in their experience. In that respect, I consider this nomination to have definitely succeeded. If the conclusion in other respects remains uncertain, then that only confirms what I rather suspected might be the case when I set out.

As I stated, I did intend this nomination as more than just an experiment, and I do intend to reapply for the BAG by more conventional means shortly. Rest assured that I will notify everyone who has commented on this nomination (yes, even the SPAs) when I do that. —Ilmari Karonen (talk) 11:05, 10 May 2008 (UTC)

Ilmari Karonen
(19/9/5); Withdrawn at 11:05, 10 May 2008 (UTC)

- As all the RfBAG nominations so far have been shoo-in reconfirmations of existing BAG members (even Coren's RfBAG, despite the opposition from people opposing the process itself, passed with 67% support), I though I'd stir up the pot a little and see how the process handles a real, even potentially controversial candidacy. So I'm nominating myself for membership in the Bot Approvals Group, just to see how it goes, and being genuinely curious about whether or not I might pass.

In case someone missed it, I'd better state that again: I am not currently a member of the BAG. Indeed, I have so far not participated in any way in the bot approvals process (with the possible exception of some offhand comments I may have made on one or two nominations). This is not because I'd consider myself lacking in technical expertise; besides being an administrator on the English Wikipedia, I'm a MediaWiki developer (svn username vyznev) with considerable experience with user scripts, and have (briefly) operated a bot myself in the past. I've simply tended to view the whole bot approvals process as essentially a manifestation of Parkinson's law, bureaucracy growing to accommodate the needs of the growing bureaucracy.

That is not to say I consider the whole thing completely useless — on the contrary, I feel that the review and feedback generated through the approval process can be extremely valuable to prospective bot operators. What I find silly is the notion that the procedure should be a mandatory hoop to jump through, or that the people in charge of it should wield any actual power beyond that possessed by anyone who has demonstrated sufficient technical expertise and trust by the community for their opinion to carry any credibility.

I have tried to articulate my views by drafting in my user space what I would consider a simpler and more sensible alternative to the current, IMHO excessively bureaucratic bot policy. For those who'd rather not read the whole thing, the nutshell box captures the essence pretty well: I believe we could dispense with a lot of drama and bureaucracy if we spent less attention on how people are carrying out their edits and more on what they're actually doing.

To clarify, I do not intend this as a protest candidacy in any way. I do genuinely believe I would have something to contribute to the bot approvals process even in its current form, and I believe myself to possess the combination of technical skills and community trust required of a BAG member. I don't expect I'll be able to devote very much time to handling bot approvals at least in the near future, being currently somewhat busy polishing off my M.Sc. thesis. Nonetheless, I do believe I might be able to be of some use, and maybe inject a modicum of common sense into the process.

I'm aware that the RfBAG process is still being debated. Should this application pass, I'm willing to submit myself to additional confirmation by the old BAG membership process or by any other system that achieves community consensus, and/or to have my acceptance into the BAG deferred until the issues with the approval process have been settled. (Of course, seeing as BAG members have no actual power, other than to advise bureaucrats on the granting of bot flags to accounts, I'm not sure what it is that would actually be deferred.) In the mean time, this should be a fun experiment. :-)


 * Candidate, please indicate acceptance of the nomination here: Why, yes, thank you. —Ilmari Karonen (talk) 03:08, 28 April 2008 (UTC)

Questions for the candidate
Dear candidate, thank you for offering to serve Wikipedia as a member of the Bot Approval Group. You may wish to answer the following optional questions to provide guidance for participants:
 * 1. Are you currently, or have you in the past, operated a bot on a Mediawiki?
 * A. I have, for a few days last year, operated an anti-vandalbot bot as User:Apostrobot (contribs). Written as an emergency response to an ongoing vandalbot attack, it was never flagged or submitted for formal approval; I basically just posted to WP:AN saying I was running the bot and telling people to block it if it did anything wrong.  (As it happens, it did, though I ended up blocking it [ myself].)  The relevant AN thread, for those interested, is archived at Administrators' noticeboard/Archive108.
 * Questions from  Chris 
 * 2. What other languages can you program in apart from perl and php. How experienced are you in these languages
 * A Of the languages most relevant to bot writing, I can program in Perl, PHP, C, Java and JavaScript pretty well. The biggest, most obvious gap in my programming skills in this regard is Python; I've been meaning to learn it for years, but I've never just got around to actually doing it yet.  My Lisp/Scheme skills are also rather shaky — I know just about enough Lisp to edit my emacs config file, but that's it.  Further afield, I have at least some experience with quite a few languages, from BASIC to 680x0 assembly to ColdFusion to Fortran.  I also consider myself pretty handy with Redcode.  —Ilmari Karonen (talk) 15:38, 28 April 2008 (UTC)
 * 3. What is your opinion on BAGers flagging bots?
 * A I think it could be good idea, though I don't really have a strong opinion either way. Certainly it would reduce bureaucrat workload and provide the group with a concrete function beyond its current advisory role.  On the other hand, as I noted in my alternative bot policy essay, I'd really like to see all administrators able to flag and unflag bots, just like we do with rollback.  I think our administrators would, on the whole, possess sufficient community trust and common sense to handle this relatively low-impact job just fine, and are sufficiently numerous to be able to effectively police each other (as we already trust them to do with other admin tools).  Of course, as not all BAG members are administrators, the two possibilities are not mutually exclusive: we could even, if the community so wanted, allow both administrators and non-admin BAG members to grant or revoke the bot flag.  —Ilmari Karonen (talk) 15:38, 28 April 2008 (UTC)

Questions from Dragons flight

 * 4. What is your opinion on the ongoing use of unapproved adminbots?
 * A I feel it's extremely unfortunate that our admins need to run bots in secret in order to keep Wikipedia running. Somehow we — both admins and the community in general — seem to have succumbed to some kind of mass paranoia where any open and rational discussion of the subject becomes well nigh impossible.  The circle feeds itself: Because adminsbots are run in secret, they're not accountable; because they're not accountable, people don't trust them; because people don't trust them, it's impossible to get one approved, and because it's impossible to get one approved, they can only be run in secret.  I don't know where we took the wrong turn.  Maybe we should've arranged a formal approval ceremony for Curps's blockbot, back in the days when it was him versus Willy on Wheels, instead of treating its existence as an open secret.  But it's easy to say that with 20/20 hindsight.


 * As for looking forward, I think the only way out of the current spiral may be to inject some light and air into the system. Having an open list of current and past adminbots at Adminbots would be a good start.  What we need are some admins who have the courage to stand up and openly (not just with a wink and a nudge) admin they're running adminbots, open the source code to their bots and submit them to community review.  By which I don't mean a vote  !vote  gauntlet poll to decide whether or not some threshold percentage of the people who care are willing to sign their name and say they trust the bot, but just plain, simple, non-binding review.  As in discussion.  If there are genuine problems with some particular admin and their bots, the community already has ways to address them (assuming the ArbCom and/or the devs don't render the point moot by stepping in first) — but what the review would accomplish would be to bring the currently obscure adminbots out into the open, hopefully allaying the more unreasonable fears and suspicions and, in time, allowing us to determine the real community consensus, once the current paranoid atmosphere has dissipated, on the automation of different admin tasks.


 * I'd be happy to set an example myself here. All I'd need first, though, would be an adminbot.  —Ilmari Karonen (talk) 03:47, 3 May 2008 (UTC)


 * 5. Ideally, what should the adminbot approval process consist of?
 * (This reply, like the one above, is somewhat long. Sorry about that.  The difficulty is that, as I already feel the normal bot approval process could use some reforming, it's hard to outline my views on adminbot approval without going over a lot of tangential issues as well.)


 * A First of all, the operator should be an admin. This should be enough to demonstrate that they have the community's trust.  If they do something stupid that makes them lose that trust — like, say, run an adminbot that runs amok — their admin status should be revoked, just as we'd generally do to an admin that acted in a harmful or disruptive manner.


 * Second, the task they're carrying out should have consensus. This is kind of obvious, though: most of the things one can do with admin tools, at the kind of rates that would require bot assistance, are either obviously useful or obviously bad.  Finally, the way they're carrying out the task — that is, the bot itself — should also enjoy community approval.  With the current system, that would include passing the current bot approvals process, though there's really more to it than that.  Having community approval (for anything, not just adminbots) doesn't mean just passing a one-time technical review by the BAG, it also means the bot and its operator need to enjoy continuing acceptance by the community at large and a general perception that what they're doing is good for the project.


 * One issue I haven't fully considered is the issuing of secondary sysop accounts for adminbot use. It is clearly something that should be done, if only to let the bots' actions be better distinguished from those made manually by their owners, but I'm not sure what the appropriate process should be.  Given that a malicious person can't really accomplish that much more harm with two admin accounts than with one (especially if it's known they belong to the same user), I'd be inclined to think that merely a general review, as done by the BAG these days, of the bot's usefulness might well be enough to let a secondary adminbot account be given to someone who has already passed RfA and has the admin bit on their main account.  But I'm certainly willing to change my mind on this if someone comes up with a better approach.


 * Finally, though it's somewhat tangential to the subject at hand, I think I ought to mention here one issue that should always be considered with any bot, but with adminbots in particular: whether the functionality would be better implemented in MediaWiki itself. That was most clearly illustrated by the ProtectionBot case, but it's an important general principle to keep in mind.  A bot can fix a problem for us; a MediaWiki feature can fix it for all the Wikimedia projects, and others besides.  —Ilmari Karonen (talk) 03:47, 3 May 2008 (UTC)

General comments


Please keep discussion constructive and civil.

Support

 * 1) Support; wanting "in" on a group in order to help alter its functioning isn't a flaw, it's the very best definition of self-governance. Technical wherewithal is not in question, nor is this editor's judgment.  &mdash; Coren (talk) 04:35, 28 April 2008 (UTC)
 * 2) Support - he is trusted and well-versed in policy, as he has a long tenure as an admin. He's technically competent, as he has svn commit access to MediaWiki. Any extra sparks in the BAG due to attempted reforms are a bonus. I see no reason to oppose. Tito xd (?!? - cool stuff) 04:41, 28 April 2008 (UTC)
 * 3) Support per the above comments. -- S iva1979 Talk to me 17:44, 28 April 2008 (UTC)
 * 4) Support. Yes, please.  &#10154; Hi DrNick ! 19:27, 28 April 2008 (UTC)
 * 5) Support Why not? As long as the bot doesn't start vandalism, I'll go with it.-- B a r k j o n 21:03, 29 April 2008 (UTC)
 * You realize you are not voting for a bot, you are voting for a user to join the bot approvals group, which approves and rejects bots? Ral315 (talk) 04:51, 30 April 2008 (UTC)
 * 1) This addition to the BAG team makes sense. John Vandenberg (chat) 09:53, 30 April 2008 (UTC)
 * 2) -- Naerii  14:59, 30 April 2008 (UTC)
 * 3) - Agathoclea (talk) 15:48, 30 April 2008 (UTC)
 * 4) --Kbdank71 17:06, 30 April 2008 (UTC)
 * 5) dihydrogen monoxide (H2O) 12:19, 1 May 2008 (UTC)
 * 6) I agree with his outlook on the bot process, and I would trust this user to make sensible decisions that account for technical issues, policy, and community sentiment.  r speer  / ɹəəds ɹ  18:39, 1 May 2008 (UTC)
 * 7) Looks fine to me. - Philippe  22:08, 1 May 2008 (UTC)
 * 8) Support per Coren, Titoxd, Rspeer and intelligent, well-considered answers to questions. Franamax (talk) 01:49, 4 May 2008 (UTC)
 * 9) Support. I appreciate his thoughtful answers to the questions.  Dragons flight (talk) 21:46, 4 May 2008 (UTC)
 * 10) Support per above, especially Coren. —Locke Cole • t • c 05:42, 5 May 2008 (UTC)
 * 11) Yes please. &mdash;Cryptic 08:20, 5 May 2008 (UTC)
 * 12) Support - ah, the fear of change. Neıl ☎  23:35, 5 May 2008 (UTC)
 * 13) Support, this is my first awareness of there being BAG membership; Ilmari appears to have the technical expertise and good judgment to evaluate bots, already community endorsed as an admin, and rounds out the committee as someone with an alternative perspective. -- M P er el 03:26, 6 May 2008 (UTC)
 * 14) Support per all of the above. Nick (talk) 12:14, 9 May 2008 (UTC)

Oppose

 * 1) Oppose. This user's technical competence is beyond any question, but I find their opinions on the BAG and BRFA processes impossible to comprehend.  Ilmari seeks a mandate from the community to join a group that he does not believe should have any authority, or even (as far as I can gather) exist at all.  He cites (effectively) disenchantment with the BRFA system as a reason for minimal prior contributions to Bot Approval processes, and demonstrates an almost deliberate ignorance (actually circumvention, which is even worse) of the existing system. However unpalatable he may find the BRFA process, as long as it has consensus, as BRFA did at the time (and arguably still does), ignoring said process constitutes exactly that: ignorance.  I accept Ilmari's claim that this nomination is not intended to be a "protest candidacy", but I cannot avoid the conclusion that he is trying to prove a point; namely, if appointed, he will endeavour to use the accompanying authority to change the BRFA process from within.  I consider this to be a misuse of power: BAG must have authority over the approval and control of bots on Wikipedia, but outside that mandate (even as close to it as the BRFA process itself) its members are just respected members of the community.  BAG's problem all along has been its insularity and lack of community involvement at all levels: its authority must be derived from processes and policies supported by the wider community, not unilaterally declared by BAG.  If Ilmari believes that the BRFA and BAG processes are broken (and I think that the majority, myself included, will agree that there is something wrong with at least part of them), then he should work with BAG and the community to change it for the better.  Seeking the authority of BAG membership purely (as far as I can tell) for the purpose of gaining a louder voice in reformation discussions, is a flagrant misuse of process, particularly ironic in the fact that, were Ilmari to have his way, the position to which he aspires would have no authority whatsoever.  Happy‑melon 14:16, 28 April 2008 (UTC)
 * At the same time, he indicated that he thinks he is useful to the BAG "even in its current form". He has expressed that he has ideas for reform, but at the same time, that if those reforms don't take place, he'll abide by current community consensus. I don't see how that is an abuse of process—it's just a clear explanation of his views on the BAG's current status. Tito xd (?!? - cool stuff) 17:13, 29 April 2008 (UTC)
 * Oppose. --Eatthefood43 (talk) 23:41, 30 April 2008 (UTC) — Eatthefood43 (talk • contribs) has made few or no other edits outside this topic.
 * Blocked indefinitely, vote indented accordingly. Acalamari 23:13, 2 May 2008 (UTC)
 * 1) Very Strong Oppose user has no prior experience with the bot approval system or comments to bot approvals. βcommand 22:14, 1 May 2008 (UTC)
 * 2) Oppose Running to make a point and accidentally breaking the Title Blacklist do not give me confidence.  MBisanz  talk 07:34, 2 May 2008 (UTC)
 * 3) INFINETELY STRONG OPPOSE Obviously in it for personal gain. A no-go candidate. I feel I would be more suitable. Mattbroon (talk) 18:48, 2 May 2008 (UTC)
 * "Obviously in it for personal gain." And what personal gain is there in being a member of BAG? &mdash;Dark talk 03:40, 3 May 2008 (UTC)
 * This opposing user appears to have 21 mainspace edits and 13 edits in 2008. Franamax (talk) 23:09, 4 May 2008 (UTC)
 * 1) Oppose, reluctantly - I have to agree with the above, especially with the part about no prior work with the bot approval system or comments to bot approvals.   -- Cobi(t 09:30, 5 May 2008 (UTC)
 * 2) Oppose. No apparent previous bot approval experience, and says won't be able "to devote very much time to handling bot approvals at least in the near future". Gimmetrow 00:54, 6 May 2008 (UTC)
 * 3) Oppose He says he has no experience. Zginder 2008-05-06T01:20Z (UTC)
 * 4) Neutral for now. While Ilmari is obviously skilled with mediawiki, as he/she is a dev, I'm not seeing any experience whatsoever with bot policy and the BRFA process. Oppose per experience. --uǝʌǝsʎʇɹoɟʇs (st47) 10:28, 28 April 2008 (UTC)
 * 5) Oppose. The lack of any participation over at WP:BRFA poses a real problem for me, as that is, in my opinion, the major job of BAG. This is by no means a comment on technical competency to perform the task or your views on bot policy, simply on your lack of participation as a community member to date. (Note: this comment was left after the scheduled end time for this RFBAG). - AWeenieMan (talk) 01:18, 7 May 2008 (UTC)

Neutral

 * 1) Neutral Your fine with the technical side of things but I would like to see more interaction at bot related pages before I support -- Chris  10:09, 1 May 2008 (UTC)
 * 2) Neutral I have a few disagreements on your opinions that I believe could turn badly if ever implemented. Also, no work for the BAG process is not helping you.  §hep   •   ¡Talk to me!  00:57, 3 May 2008 (UTC)
 * 3) Gurch (talk) 21:41, 4 May 2008 (UTC)
 * 4) Oppose Neutral as Happy Melon. A particular problem of the current/old BAG was the absolute inability to conceive that a bot operator of an approved bot might be problematic. While I am happy to believe that your own behavior is acceptable, being a member of the BAG requires being able to sort the wheat from the chaff. Your policy appears to assume that it is self-sorting, and that approved bot operators are inherently good fellows who need some space to roll up their sleeves. I don't share your optimism. AKAF (talk) 16:19, 30 April 2008 (UTC)
 * Oh dear. That is certainly not the impression I was trying to give.  On the contrary, I fully share your opposition to treating bot approval as a carte blanche, and feel that any such attitude, insofar as it exists, should be strongly discouraged.  (I don't feel myself sufficiently familiar with past BAG practice to  really tell how common such attitudes might be, though I do have to admit to encountering some sentiments in that direction expressed on occasion.)  One change I'd like to see in the bot policy would be a shift of focus away from extensive pre-approval with lots of second-guessing and hoop-jumping and more towards an ongoing review of bot (and operator) reliability based on actual performance in real-life use.  This ties in with my wish, as I've tried to express it in my bot policy essay linked above, to put a greater emphasis on the responsibility and accountability of individual bot operators in general.  —Ilmari Karonen (talk) 18:12, 30 April 2008 (UTC)
 * (Moved to Neutral) I must say that seeing your well-reasoned responses, I wouldn't be against any editor who can bring their civil and intelligent outlook to the BAG. Good luck. AKAF (talk) 11:34, 5 May 2008 (UTC)
 * 1) Neutral, per Chris G. &mdash; E  ↗T C 05:29, 10 May 2008 (UTC)


 * The above adminship discussion is preserved as an archive of the discussion. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the talk page of either this nomination or the nominated user). No further edits should be made to this page.