Wikipedia talk:Bots/Archive 19

Archive 15

←

Archive 17

Archive 18

Archive 19

Archive 20

Archive 21

Archive 22

Question regarding the need of bot approval

I just read through the policy, but i am not sure about some things:

1. Do you need any kind of bot approval for a script that will help you, the person, doing stuff? (i.e. advanced javascripts that will automate some processes while you are still clicking and using the "show preview" function before saving)
2. If you use an external script to change pages, but you are sitting next to it and check all of the script's revisions, do you need it to be approved first? (i.e., a script that will run through an article or a category per manual start)

Thanks. Niderbib (talk) 11:44, 8 April 2008 (UTC)

The things you described don't need bot approval as long as you keep your editing rate low and are authorizing each edit individually. If you want to edit in a sustained way (e.g. 6 edits/min average for twenty minutes or more), then you should consider getting a bot flag. Someone will probably let you know if your editing rate makes you look too botlike. — Carl (CBM · talk) 12:15, 8 April 2008 (UTC)

Any script which is either 1) used to edit "fast", or 2) used to do "a lot" of edits in total, or 3) does not involve a Yes/No for each individual edit, should in my opinion be reviewed for technical and policy compliance. Exactly what "fast" and "a lot" might mean is unclear, but I think if you're planning to use a script to modify 10000 articles, you ought to have consensus and not use a buggy script. Gimmetrow 04:24, 10 April 2008 (UTC)

My guideline is "Anything which decides on the edits to make itself" needs bot approval. That is, a mass-deletion script that takes a list of pages as input does not, but a bot which decides which pages to delete does. — Werdna talk 12:52, 17 April 2008 (UTC)

BAG approval section

That section has no consensus and should be removed and re-written. β^{_{command 2}} 16:55, 9 April 2008 (UTC)

I'm sorry for being blunt, but you should have said so during the 23.5 hours that this replacement policy was drafted, on a page not linked from this talk page. You're too late now. -- Eugène van der Pijll (talk) 17:13, 9 April 2008 (UTC)

I scarcely think it's too late considering the swift move from the draft, and the continuing discussion at the draft talk page, however I think a more effective place to complain would be at User_talk:Coren/Bot_policy, which is very much alive. AKAF (talk) 17:19, 9 April 2008 (UTC)

Let's ignore procedural issues, for the moment. Betacommand, what parts of it are you concerned about? More generally, do you mean the section on BAG approving bots, or the section on people approving BAG? — Carl (CBM · talk) 17:34, 9 April 2008 (UTC)

The fact that it appears that joining BAG now requires a RFA subpage, and will be voted on by strict numbers 2/3. and that current members must re-apply through this joke of a voting system. Joining BAG is not a big deal, be involved in the bot process; commenting on BRFA's, know bot policy, and know how bots operate on at least theoric level. β^{_{command 2}} 17:42, 9 April 2008 (UTC)

Thanks. I thought that was what you meant, but I wasn't sure. I am also curious about how the new BAG confirmation process is meant to work. I find the current method of selecting BAG members (where new members are "inducted" by existing members) makes membership an equally "big deal". One way to make it not a big deal would be to abolish the BAG and go back to the community noticeboard method; I have some sympathy for that proposal. — Carl (CBM · talk) 18:25, 9 April 2008 (UTC)

BAG does NOT "induct" new members. what the Bcrats have done is say, if its crystal clear acceptance go ahead and close it yourself as there is no need for a crat. But if there are any issues then BAG defers to Bcrats. When joining BAG voting is BAD. also your idea about a noticeboard is a VERY VERY bad idea. I was the person who original re-did the BRFA process, before the re-write it could take up to a month from the end of a trial until you got the bot flag/Approval. also the community would probably approve some real bad ideas for bots, see the list of frequently denied bots. β^{_{command 2}} 18:40, 9 April 2008 (UTC)

You're saying that the bots that are now denied would somehow be approved if we had a noticeboard? That seems unlikely. The noticeboard isn't my idea, by the way - other people have been advocating it. If it takes a month to get a bot flag the first time, that isn't a dealbreaker for me. If it takes a week every time a new task will be added to an existing bot, that is a much more significant issue.

On the other issue, tf you already accept that 'crats close the BAG discussions, then I don't see how moving the nominations from the talk page of WT:BAG to a more public location is much of a change. My impression, though, was that nominations currently are essentially decided by the current BAG members. — Carl (CBM · talk) 19:34, 9 April 2008 (UTC)

We already make request public by posting on AN ANI and the VP. RFA is broken and I dont want to see BAG turn into a popularity contest, like RFA is. and yes if there was a noticeboard the lack of formal oversight would be approving bots that should not be operating due to the fact that there will be no central oversight. Also with the old version operators did not bother to get approval for tasks once the bot was approved, it took between a week and a month if you where lucky. there are just very very few people interested in working with bot approvals. it is not flashy and its not a sign of importance like RfA is. BAG and BRFA is just a lot of quite paper pushing with no real rewards. β^_command 21:05, 9 April 2008 (UTC)

I think everyone can agree that we don't want to have a redux of the requests for adminship page. The "old system" I am thinking of (which I never participated in, this is all hearsay) was apparently set up so that if you made a request and nobody objected in a couple days, that was enough for you to proceed until someone did actually raise an objection. So there was no "approval", just the regular consensus process that governs all editing. I find that appealing, in theory, compared to the current system. I will make some more comments later. — Carl (CBM · talk) 21:12, 9 April 2008 (UTC)

And I think that system would be great for the second and subsequent tasks for an approved bot. It would be simple, lightweight, and encourage bot operators to document what's going on. As for BAG joining, although waiting 40 days to join BAG is absurd too, the "proposed" RFA-like system is overkill. Adding it here live without notice doesn't help. Gimmetrow 04:35, 10 April 2008 (UTC)

I think it's fairly clear that "induction" is precisely how BAG membership works at present. Personally, I'd be less than happy about the "grandfather clause" running as long as six months. Admittedly, that may be somewhat moot if there's no general acceptance of a new system, and there continues to be a lack of coherence and clarity about what the BAG's role actually is. It may also be somewhat redundant, if the nomination process doesn't attract much wider interest than has previously been the case. It would indeed be bad to turn BAG membership into a "popularity contest", but it would be worse to leave it as a self-selecting "technical" oligarchy. Alai (talk) 14:07, 12 April 2008 (UTC)

Alai, please before getting into a discussion understand the topic first. BAG is not and never has been self-selecting, nor is BAG a oligarchy, BAG however is run by those who understand bots, and enjoy working with them. I have tried countless times to get the community more involved in the bot approval process but they have not taken any part in it. If bot approval is run like RFA nothing productive will get done. there are things that may not have "consensus" but still need to get done. If you look back with full hindsight you will see that BCBots NFCC image tagging would never have gotten approval, because there are just too many users who dont care about the m:Mission or the m:Vision. and lobby for the liberal use of non-free material. β^{_{command 2}} 20:41, 12 April 2008 (UTC)

I agree fully that converting the bot approval process to run anything like RfA would be foolish in the extreme, for all the reasons you note. But BAG approval is a different matter altogether. Happy‑melon 10:40, 13 April 2008 (UTC)

If one part is ran like RfA the rest follows. because those that are elected like RfA will make a joke out of bots. there might be a strong disagreement about a bot task, but if the task has a strong policy backing it will be accepted over time. when the first ant-vandalism bots were introduced there very strong dis-like for them, but you know what several years later when they go off-line we complain. any RfA like nomination is a vote, bots are not about votes, bots are about discussion and logic, which unfortunately a large part of the community lacks. β^_command 13:43, 13 April 2008 (UTC)

what are these if not miniature RfAs? The only difference is that RfAs are widely watched by a broad cross-section of the Wikipedia community, while WT:BAG has only a narrow group of editors interested. A bureaucrat would not dream of promoting an admin candidate with only half a dozen support votes (even though the tally was 100% in favour); yet in giving someone the authority to instruct 'crats to add or remove user rights, this is considered normal. I think there is merit in the comment that an increased level of popular support in the community would be required to pass an RfBAG at RfA, as opposed to one at WT:BAG, but I do not agree that this is a bad thing. The BAG is, or certainly should be, responsible for (in a manner of speaking, and among other things) 'protecting' Wikipedia and its community from rogue or recalcitrant bots and bot owners; how can a BAG member be the community's knight in shining armour if he or she does not have the support of that community? BAG is currently quite good at approving bots, and I do not wish to see the BRFA process changed significantly at all; but BAG also has a responsibility to control and, if necessary, 'kill' out-of-control bots and their operators, and it is currently not very good at that at all. With clear community approval for BAG members in an RfA-like process, I am confident that BAG will find itself with more authority to amend or revoke bot approval and to enforce compliance with WP:BOT. With careful management, a good balance of bot operators and wider community members can be obtained at RfBAG - I recommend a double transclusion of the subpage, to RfA and also to WP:BON - and thereby ensure that the BAG candidate has the support both of the bot-operating community and the wider Wikipedia editorship. Happy‑melon 16:18, 13 April 2008 (UTC)

Happy-melon, those are not RfAs. RfA is a fucking joke and a straight vote. requesting to join BAG is a discussion, and is not full of absolute dumbasses and vote against user for reasons like "not enough portal talk space edits". when requesting to to join BAG the person is asked to leave notices on AN, ANI, BON, and VP It cannot be helped that the community chooses not to take a more active roll. making BAG a RFA process WILL turn BAG into a joke. As for BAGs authority we have it, just few members choose to exercise it. If someone has problem with a bot I have told then to bring it to WT:BRFA and it would be handled. very few issues have. I personaly take pride in the bots that I approve. Recent events have shown that the comunity wants BAG to have more authority over bots and they will start using the authority. β^{_{command 2}} 16:46, 13 April 2008 (UTC)

You wrote when requesting to to join BAG the person is asked to leave notices on AN, ANI, BON, and VP It cannot be helped that the community chooses not to take a more active roll. I strongly disagree. This is a proposal for "helping" what you say "cannot be helped". It may not be a good one in your opinion (do I sense some dislike of RfA? ;-)), but it is a proposal nevertheless, and certainly one worthy of serious consideration. — Werdna talk 01:36, 14 April 2008 (UTC)

I just noticed this. Glad it was *almost* given a full day of consideration (at least, if the commentor above was correct). Ugh. I can't wait till the first "Not enough Category: space edits" :( SQL^{Query me!} 18:35, 14 April 2008 (UTC)

OK, half-joking aside, a couple things strike me... "In order to allow a smooth transition, currently active members of the BAG at the time this process was put in place (9 April 2008) are grandfathered as members of the BAG for a period of six months. Upon termination of this period, if they have not already been re-approved, they must present themselves before the new system or be removed from the BAG.". I've changed this to current members. I'm fairly sure that was the way it was intended. If not, I'm sure someone will revert me :) Another clarification, it does not really say clearly (just says similar to)... "This process will take place in the same way as Requests for Adminship, as a subpage.". A subpage of what? I'm assuming that WP:RFA is meant, from the comments I saw above, but, is not specified. SQL^{Query me!} 18:56, 14 April 2008 (UTC)

My suggestion is to make the RfBAG (for want of a better acronym) subpage of WP:RFA, but to transclude it to both RfA and somewhere like WP:BON, so it's prominent for both the bot community and the wider wikipedia community. I suggest that the votes be divided into "bot operators support/oppose/neutral" and "non bot-operators support/oppose/neutral", so the crats can see clearly how a candidate's support is divided. I recommend that we very carefully avoid giving any prescription of what 'pass mark' is required in each section: let the crats evaluate the first round of RfBAGs (I expect that the majority will be pretty decisive) and work out for themselves what constitutes an appropriate threshold. The concept of a 'pass mark' is itself anathema to the original intention of RfA, so there's absolutely no reason to prescribe one for RfBAG. Happy‑melon 20:14, 14 April 2008 (UTC)

If we absolutely have to do it this way, then, yeah, I agree with Happy-melon. Just let the crats do what they do, and, interpret consensus. I'm not really clear on what's to be gained by splitting it into bot-ops and non-bot-ops, however. An interesting sidenote, we've covered a lot of ground here recently. From "Anyone can join" to "Join after a short informal-ish pseudo-vote on WT:BAG" to "Voted in at WP:RFA". SQL^{Query me!} 01:18, 17 April 2008 (UTC)

Talk about process creep! A whole RFA-like system for BAG? RFA in general is a bad idea, the only reason we use it for adminship is because no one can come up with a better idea. But for a small group with limited scope, RFA with a bureaucrat deciding the outcome is major process-overkill. There's no need to have a massive vote with subpages and the whole works when a talk page discussion will suit. Mr.Z-man 01:58, 17 April 2008 (UTC)

I really, REALLY hate to be that guy, but, given the wildly varying opinions, and, the lack of much discussion prior to adopting this clause, I am going to tag the relevant section with {{disputedtag}} after this edit. Hopefully, shortly we can work the issues out, and, settle on something that everyone's happy with. SQL^{Query me!} 03:19, 17 April 2008 (UTC)

Yes, the section certainly is under discussion.  :-) FWIW, I agree with removing explicit percentages if they are felt to be unneeded if we put in some wording to the effect that BAG membership shouldn't be a stringent as adminship— putting in actual numbers was an act borne of paranoia, not dire necessity; the intent was simply to indicate that 2:1 support should be quite sufficient given that the BAG does group decisions, not unilateral ones like admins can. — Coren ^(talk) 12:31, 17 April 2008 (UTC)

I don't really see how this new approval system can help the bag. The only problem I can see with the old system was a lack of community participation and this new rfa system isn't going to help fix that. If you want more participation have the old style approval system on the WP:AN, at least people will look at it there --Chris 13:00, 17 April 2008 (UTC)

Section break

Unless someone else comes up with a good idea I'm going to revert that section of the policy back to the old version. This new version has more problems than solutions IMHO --Chris 11:17, 19 April 2008 (UTC)

I think we're actually overstating the extent of the change that's been made here. RfBAGs have always been miniature RfAs - take a look at a some examples (1 2 3 4 5 6 etc - just have a look through the WT:BAG archives). The proposal only actually changes two things: firstly, the RfBAG occurs on WP:RFA rather than WT:BAG, and secondly, bureaucrats determine the final outcome and close the poll, whereas current BAG members do it at the moment. I especially approve of the second part: it is very poor form for nominations to such a group to be administered by current members of that group. We condone 'crats closing RfBs only because there's no one else to do it, and that's what 'crats are for: closing things and determining consensus. Why can't we ask the 'crats to close the occasional RfBAG as well? It's a minute workload (average of what, five or six per year?). The only question is, how do we attract more interest in the process? I think WT:BAG is a stupid place to put nominations, almost as if it were designed to be obscure (of course it wasn't, but it has had that effect). At the very least nominations should occur at WP:BON, which most bot operators watch. In my opinion, we should separate each RfBAG onto a subpage; then it can be transcluded wherever we like (and we don't necessarily have to specifiy where in WP:BOT, although we probably should). It makes absolutley no difference whatsoever what page the nomination is a subpage of (I think Wikipedia:Bots/Approvals group would make sense, although I also want to move that to Wikipedia:Bot Approvals Group to bring some sanity to the naming conventions). Then we can transclude them to WT:BAG, WP:BON, and yes, if we want to, WP:RFA and/or WP:AN as well. In fact, WP:BN is probably more appropriate than WP:AN, as the 'crats are the ones who will have to work with BAG in flagging bots. The point is, the current BAG approval system is only three small steps away from a proper RfBAG. One step ('crat closure) is (IMHO) vitally necessary, and I don't see why we didn't insist on it from the start. One step (splitting to a subpage) is utterly trivial. All we're actually arguing about is step 3: where to post the nomination. You want another idea? How about this wording:

New wording proposal

Any Wikipedia editor may apply to become a member of the Bot Approvals Group. There is no requirement for a BAG candidate to be a bot operator, although a sound knowledge of Wikipedia's policies and guidelines as they apply to automatic- and high-speed-editing, and the capabilities and limitations of automated scripts, is an essential prerequisite. Requests to join the BAG ('RfBAG') are endorsed by the community in a discussion similar in style to the Requests for adminship process.

A RfBAG discussion is held on a subpage of Wikipedia:Bots/Approvals group; this page should be transcluded onto Wikipedia:Bot owners' noticeboard, Wikipedia talk:Bots/Approvals group and Wikipedia:Bureaucrats' noticeboard. The discussion should remain open for seven days, during which time all wikipedia editors are invited to participate. At the end of that time, a Wikipedia Bureaucrat will determine whether there is consensus for the candidate to be appointed to the BAG.

The current membership of the Bots Approval Group can be found here, divided into 'active' and 'inactive' (no BAG-related activity within the past six months). If a member no longer intends to participate in the work of the Bot Approvals Group, he or she should resign the position. Members are removed from the Bot Approvals Group only in extraordinary circumstances, when it is clear that the individual has lost the trust of the community or is incapable of carrying out the associated tasks appropriately. A proposal to remove a member from the BAG should be made on the Bot owners' noticeboard, and publicised on The Administrators' noticeboard, the highest-profile forum on Wikipedia.

Those users who were members of the Bot Approvals Group at the time this process was put in place (9 April 2008) shall retain their membership for a period of six months. Upon termination of this period, if they have not already been re-approved, they should present themselves before the new system or be removed from the BAG.

Comment and criticism on this alternative is welcomed. Happy‑melon 15:49, 19 April 2008 (UTC)

Why does it have to be so complicated? All that does is change the location. It still uses the unnecessary subpage system, it still uses a fixed timespan, its still based on RFA, and it still requires a bureaucrat to close it. This is excess bureaucracy for no good reason. A simple section for a normal discussion on a page like WP:AN should suffice to bring in more community voice to BAG membership and just let the discussion go until it dies down or there is obviously a consensus either way or it is obvious that no consensus will form. Mr.Z-man 17:44, 19 April 2008 (UTC)

There is something to be said for using a process people are already familiar with, in a venue where editors expect this sort of discussion to take place. Personally, I think adding a new, different process is adding more bureaucracy than reusing an existing one, even if the different process is, in some respects, slightly more lightweight. — Coren ^(talk) 17:48, 19 April 2008 (UTC)

A section on WP:AN is basically what we use for community sanctions and is very similar to processes like WP:TFD. Mr.Z-man 17:47, 21 April 2008 (UTC)

My thoughts about bots

I have been thinking quite a bit about bots lately. I find the current BAG system (before the very recent policy rewrite) to be less than ideal for several reasons. The policy rewrite tries to address some of these shortcomings, but I think it still falls short in a few areas.

My assumptions:

1. The benefits of a bot flag are:
   1. To remove bot contributions from the recent changes list, which allows for faster editing
   2. To give the account access to more data per query from the API
   3. A bot flag does not otherwise give the bot more authority to edit than a manual editor would have. Granting a bot flag signifies a level of trust in the bot operator.
2. BAG exists because there is benefit in having technically proficient editors review bot proposals in addition to community review of the need for the task.
3. BAG currently reviews two things: bot tasks, and bot accounts. They typically review bot source code only in the sense that they say "take it for a spin" and then they ask "did it break?".
4. There is no current, easy way for BAG to withdraw approval for a task after it is granted, even if there is a visible lack of community consensus.

Ideas for change:

1. BAG should still advise bureaucrats about which bot accounts should be flagged. They might interview the operator about her background, programming skills, history of interpersonal communication on the wiki, or other relevant issues.
2. The review of bot tasks should be separated from the request for a bot flag. Like all editing, bot tasks are implicitly approved when there is consensus for them to go forward, and lose approval if that consensus changes.
   1. New tasks should be proposed on the bot noticeboard. If nobody objects in a fixed period of time (say 2-3 days), the task can go forward until someone objects. This is like to the current editing system, but with a mandate for caution rather than a mandate to be bold.
   2. If a reasonable objection is raised, the task should pause for discussion. This is no different than any other editing on the wiki.
   3. Everyone is welcome to comment on the bot noticeboard. BAG members would still be able to raise technical objections to tasks, for example if the bot operator proposes something that is technically flawed. BAG would even have "veto power" on tasks, so that they can indefinitely pause a task until technical concerns are met. But they would not approve tasks, except by being part of a broader community consensus.

There is room to tweak these proposals, of course. — Carl (CBM · talk) 21:43, 9 April 2008 (UTC)

It's kind of funny, I had this exact same idea earlier today and was thinking about the details of presenting it. Quite frankly, it is exactly how I feel the system should work. A bot operator who has a number of bots operating should not have to prove that they still know how to write code by requiring a test run for their fifth bot. Any programmer who fails to test new code to see if it destroys things before rolling it out was clearly not deserving of a flag in the first place. Except in the cases where implementation is quite complicated and/or editing rates will be extremely high, what should matter is whether or not the task has consensus to be done at all. It annoys me to no end every time I see a simple find and replace request I could fulfill if only I was in the mood to go through the several day process of proving, yet again, that I know what I'm doing. This sort of noticeboard system would simplify everything, improve communication, and solve the problem of very experienced bot operators ignoring the system altogether. I would only add a couple tweaks to what you have proposed: First, due to the present concern that multiple recurring tasks with high edit rates should be separated into multiple accounts, the obtaining of a second bot flag should be as simple as obtaining consensus for the task necessitating it. Second, I think that these discussions should take place on a noticeboard specifically created for them, as opposed to the existing bot owner's noticeboard (I'm not sure whether or not that's what you intended when you said bot noticeboard, but I just wanted to be specific). The only requirement this system would have is that people who care about these things would actually have to pay attention to the noticeboard, although I would hope that the knowledge that a really wrongheaded task could be quite legitimately be started in good faith if no one was paying attention would motivate people to actually pay attention to what's going on there.--Dycedarg ж 09:18, 10 April 2008 (UTC)

One bit of expansion: If a task has been done in the past by someone else, and is utterly uncontroversial and necessary, then the waiting period should either be shorter or nonexistent. For example, an existing bot operator wanting to tag a few hundred categories that have been mass CFD'd shouldn't have to wait three days to do so; the action is obviously necessary and has consensus.--Dycedarg ж 09:31, 10 April 2008 (UTC)

It sounds good. The only thing I would add is that bot operators who want to add tasks to an existing bot are encouraged (or obliged?) to notify applicable WikiProjects (unless identical tasks have already been approved for other bots). As technical competence is no longer in question, the only questions will be about the contents of the changes, and we want to encourage outsiders who have experience or knowledge of the nature of the changes to comment. -- Eugène van der Pijll (talk) 13:25, 10 April 2008 (UTC)

I suggested something similar at Wikipedia:BON#-BOT_Process. My main concern is encouraging bot operators to document what they're doing and allowing the community (including BAG) to give some feedback on tasks. Gimmetrow 03:42, 12 April 2008 (UTC)

Rewrite

I fully agree with the opinion voiced in various places above that this complete rewrite should have been much more widely publicised and that it now needs such wide publicity to achieve consensus. I would like to say that I support the policy as it currently stands as a vast improvement over the previous version, although there are a few changes that I would like to see. I am not at all surprised to see opposition to the new BAG approval system from a current BAG member (who will need to be reappointed under it); however I have always considered heavy community participation in BAG appointments (by which I mean far more than the token notice on AN and VPP) to be imperative to increase community support for, and the authority of, BAG members. My major concern in this policy is the high level of detail: nowhere, for instance, do we prescribe the level of support required for successful RfAs or RfBs - why should we be prescriptive about the level required for RfBAG? The extensive detail in the "approval" section is similarly excessive, and the "Appeals and reexamination of approvals" section is WP:CREEP to the max. Policy on Wikipedia establishes the general outlines of a process or rule, and precedent and consensus fills in the details. There is no need to be so prescriptive (and in some cases proscriptive) about the finer details. I think that the appeals procedure is unworkable as written and sounds far to much like the ArbCom for bots, but the rest of the policy is sound, and I fully support its adoption. Happy‑melon 19:45, 11 April 2008 (UTC)

"currently stands as a vast improvement over the current version"? Maybe you meant something different. MBisanz ^talk 19:51, 11 April 2008 (UTC)

Hehe thanks. Yes I meant over the pre-rewrite version. Happy‑melon 21:09, 11 April 2008 (UTC)

Part of the reason there may be too much detail rather than, arguably, not enough is that this was repeated often as a cause of problems; that policy was underspecified and therefore too vague to actually apply. As for the percentage in the election process— well, I felt that the recent problems about undefined "sufficient consensus" in the recent 'crat elections should be forestalled. Whether the ratio as it is now is adequate or should be tweaked is, of course, open for discussion. There isn't a Wikipedia Stone Tablet edition planned anytime soon.  :-) — Coren ^(talk) 20:19, 12 April 2008 (UTC)

I agree with Happy-melon's comments. Requiring re-approval from the community is essential, to remedy the fact that the BAG has been self-selecting. The effects of this self-selection are that the goals of the BAG have become detached from the goals of the community, and certain BAG members actually take an antagonistic stance toward the benighted, uninformed community. Since this creates the possibility that some members of the BAG will lose their membership, it is not surprising to see a BAG member oppose the idea. I am skeptical of setting firm approval thresholds, but I understand Coren's response and the need to address all possible issues. rspeer / ɹəədsɹ 16:46, 16 April 2008 (UTC)

Not so much self-selecting. Anyone is/was welcome to participate at WT:BAG. I think, the community has been begged to do such many times before, at AN and the village pump (often times, a cantidate would post there, asking for more input). SQL^{Query me!} 01:21, 17 April 2008 (UTC)

Well, I'm trying to participate in BAG stuff as much as I can right now, but in the past there has been the distinct unwritten impression given off by some of these pages that "if you're not BAG, it's not your business". That was certainly the case back when ST47 protected a page that was getting too much community input; I recognize that BAG is in the process of reforming itself now, but it's hard to deny that previously it's been a pretty insular group.

Incidentally, how do you keep abreast of all the topics of discussion? There are a lot of pages to watch without much consistency in their names. I had no idea there was so much discussion going on on Coren's talk page, for example. rspeer / ɹəədsɹ 07:35, 17 April 2008 (UTC)

That's because that was where the discussion originally took place while the draft policy was next to it; but it quickly became apparent that without doing something bold almost nobody was going to come over to discuss it. An history merge was workable for the policy page per se, but I couldn't figure out how to do this "right" with a threaded talkpage. — Coren ^(talk) 12:25, 17 April 2008 (UTC)

Sorry I missed your post rspeer, the way I keep abreast of discussions and what not, are using my Watchlist. There's a tickbox somewhere in Special:Preferences, that watches any page you edit, has come in a lot of handy for me, so far. I, too, missed the discussion on Coren's talkpage as well, so at the very least, don't feel too bad on that one. SQL^{Query me!} 04:55, 20 April 2008 (UTC)

Quite confused

I'm quite confused about how bot policy was suddenly drastically changed. From what I can tell, a couple of users worked on a new policy in a user subpage. Then, suddenly and with little to no notice, they swapped the "old" bot policy for a entirely different version that includes RfA-style votes for BAG members. Can someone please explain to me what is going on? Why was this policy so suddenly changed, and why was the community input seemingly unheard? --MZMcBride (talk) 03:16, 17 April 2008 (UTC)

Selection via an RFA-style discussion overseen by bureaucrats makes sense to me. I'd even go a step further and require a minimum number of participants (so people can't simply be promoted on the basis of one or two editors input). A quorum, basically. I do think more precautions need to be in place for selecting BAG members though.. —Locke Cole • t • c 04:50, 17 April 2008 (UTC)

The subpage was announced multiple places. Very few people showed up, and as you demonstrate many people missed the announcements. I would imagine Coren thought that the best way to jump start a discussion about the bot policy would be to completely change it. As his modified policy tag freely admits, the changes were sudden and not all sections of the page have consensus. For what it's worth, I agree with his method; even after changing the actual policy page and announcements on WP:VPP, WP:AN and WP:BON, we haven't gotten the level of participation I would have hoped for.--Dycedarg ж 05:44, 17 April 2008 (UTC)

From what I can tell, this was done over a period of 48, maybe 72 hours. If this were a guideline, I'd have no issue. If this were simply a rewrite for clarify or brevity, I'd have no issue. But this is a fundamental change in bot policy. Two to three days really isn't enough for the level and depth of changes made here. And, in all honesty, the criticism that BAG is separate / aloof from the rest of the community isn't really helped when major changes are done like this. : - / --MZMcBride (talk) 06:14, 17 April 2008 (UTC)

(ec)If the intention had been to 'slip in' such a radical change while bypassing the discussion/consensus process, I would quite agree with you. However, that doesn't appear to be what's happened. As I see it, Coren and co started a rewrite process, came up with a draft, and asked for comments on it before it went live. Thoroughly underwhelmed by the level of response, they realised they'd need to do something drastic to provoke any real discussion (provoke being the appropriate word given the depressing level of apathy for this area of Wikipedia policy). Hence the (appropriately qualified) replacement of the live version which, hey presto, has indeed precipitated the necessary discussion. Happy‑melon 09:25, 17 April 2008 (UTC)

Um, Happy meant to say WP:BRD, but forgot the shortcut and had to reproduce it the hard way. MBisanz ^talk 20:15, 17 April 2008 (UTC)

Commenting out my original rant-like post to say instead that this is a fairly dense page of information and perhaps could be improved with better summarization near the top, also summarization of the significant changes to bot policy introduced in the draft version, in the interest of making it accessible to the wider non-technical audience. Franamax (talk) 08:50, 17 April 2008 (UTC)

That, indeed, was the whole point. If it stayed as a subpage, it would have gotten a thourough hashing from, maybe, a half dozen editors. Being bold managed to kickstart some discussion (albeit even then it's getting less that it should). Some very good discussion has already occurred; let's keep at it. — Coren ^(talk) 12:22, 17 April 2008 (UTC)

New policy summary

I note that the general feeling over the policy changes generate quite a bit of ambivalence on some of the detail. In particular, the following points stood out:

• There is a general feeling that this represents, as a whole, a good step in the right direction; but
• There is talk that approval for further tasks to an already approved bot, and for new bots for already established operators, should be streamlined

  What about adding a section in the original request, linking it again on the BRFA page, and reducing the testing requirements under the presumption that the bot operator already demonstrated minimal ability? That looks like a good candidate for "speedy" approvals.

• The RFA-like process for BAG member selection is, in general, supported by "outsiders" and opposed by "insiders" (in fact, the existence of "insiders" is perceived as part of the problem).

I did propose the new approval method; because I felt this was a solution to the perception that the BAG was a clique that self-selects far enough outside of the general community's scrutiny and therefore had no real mandate to watch over bots. Some current BAG members oppose the change because they fear the process will degenerate into a popularity contest and that technical ability will be set aside to the detriment of all.

The best way to settle this, IMO, is to simply go ahead tentatively with the new process. As I've said before, nobody is planning a "stone tablet edition" of Wikipedia anytime soon, so if the change is a failure, nothing prevents us to, collectively, scrap the whole thing. So I'm going to put my money where my mouth is by putting my neck out as the first such candidacy sometime today. Let's hope mixing metaphors is not a valid oppose reason.  :-) — Coren ^(talk) 16:15, 19 April 2008 (UTC)

Wikipedia:Requests for BAG membership/Coren — Coren ^(talk) 16:53, 19 April 2008 (UTC)

As an aside, I've left a note for the 'crats, but I did not publicize the new process or my candidacy to avoid appearance of impropriety or canvassing. It might be a good thing if someone else could poke the VP and such, though. — Coren ^(talk) 17:01, 19 April 2008 (UTC)

The RFA-like process for BAG member selection is, in general, supported by "outsiders" and opposed by "insiders" - I've seen zero evidence that it has significant support or opposition by any group, mainly because this was added to the bot policy without any significant community involvement. Mr.Z-man 17:49, 19 April 2008 (UTC)

Of the subset of editors who commented, of course. I can't speak for those who have yet to express themselves. — Coren ^(talk) 17:51, 19 April 2008 (UTC)

In my opinion, the mere fact that after the number of announcements that have been made and the fact that the policy has been rewritten for a good 10 days now and we still have far less participation than WT:U gets when you change a single sentence speaks volumes about how much people care about the bot policy. What are we supposed to do except proceed as if the people who aren't participating tacitly approve by virtue of their lack of objection?--Dycedarg ж 19:12, 19 April 2008 (UTC)

Perhaps proceed with the non-controversial parts, and, work out the disputed parts here? SQL^{Query me!} 19:17, 19 April 2008 (UTC)

"What are we supposed to do except proceed as if the people who aren't participating tacitly approve" - Uh, no, you invite more people to comment. Perhaps they haven't objected because they have no clue this change even happened. Even with just the comments here right now there is no consensus for some of the changes, unless we count the 5000 users implicitly supporting. Mr.Z-man 19:38, 19 April 2008 (UTC)

We announced twice each on WP:VPP and WP:AN, and once on WP:BON. What precisely would you suggest we do? Post an announcement in the site notice? Make another announcement on WP:VPP and WP:AN in blinking <big> bright pink text with animated gif's of explosions next to it?--Dycedarg ж 20:57, 19 April 2008 (UTC)

Something other than going ahead with it over what is becoming overwhelming consensus not to from the community? SQL^{Query me!} 19:23, 21 April 2008 (UTC)

Reversion of BAG approvals section

I've reverted the BAG approvals section. There have been several users who have objected to it (Beta, Mr.Z-man, myself) and I feel that it is too radical of a change that was not discussed thoroughly enough. There are plenty of people who object to RfA-style votes in general, instituting another doesn't seem to be the optimal solution.

Like SQL above, I hate to be that guy, but a reversion is what is needed, as I see it. And as MBisanz said above, this is really just part of WP:BRD. We've reached the 'd.' : - ) --MZMcBride (talk) 19:32, 19 April 2008 (UTC)

I agree, a reversion is what's needed, and I've reverted you. The nomination was only up for what, about a day, and already it seems to have gotten more input from the community than our current method. Further, I don't see why three editors should be able to derail this "test case" (which so far seemed to be succeeding). Let's see where this goes; I think we'd all prefer wider community involvement, and if this results in that, I believe it's a process we should keep and use. —Locke Cole • t • c 20:47, 19 April 2008 (UTC)

There are many other ways to get community involvement that don't exponentially increase the amount of bureaucracy involved in BAG membership. "I think we'd all prefer wider community involvement, and if this results in that, I believe it's a process we should keep and use. " - Of course if you give a place for people to vote on something they will, that doesn't mean its the best idea and we should not consider any alternatives. Mr.Z-man 20:57, 19 April 2008 (UTC)

see m:Voting is evil bot approval is a discussion not a vote, dont make it one. there was no consensus for the change so reverting is proper. β^_command 21:31, 19 April 2008 (UTC)

Voicing an opinion != voting. Just because one says support doesn't mean they're voting, it means they're discussing, and their contribution to the discussion is that they're supporting the nominee. If you'd like reasons, perhaps you should contact the individual editors directly. But I fail to see the need for lengthy discussion when the question being posed is simply "do you support this editor to be a member of the BAG?". —Locke Cole • t • c 08:55, 20 April 2008 (UTC)

I am very amused at several things. One, the claim that a revert is 'needed', even though nearly no discussion on this process has occurred at all. Two, the opposition to this proposal after the community expressed the need for this very reform, adding power to BAG. Figure out what you want us to do, and we'll find a sane way to implement it, however if you all demand the BAG do something about things like BetacommandBot, and then shoot down things like the proposal which would give us the power and credibility to do that, then you continue to have things not go your way, nothing gets done, and people accuse us of not doing anything. --uǝʌǝsʎʇɹoɟʇs(st47) 22:20, 19 April 2008 (UTC)

ST47, the community has given BAG the authority, but no-one in BAG has had the testicular fortitude to use it, had I not had a conflict of interest with the BCBot issue I would have asserted that authority. any RfA system will turn BAG into a circus just like adminship is. admins have the ability to solve a lot of issues, but very few assert that ability, and even fewer exert that on anything that matters, because everything is becoming a popularity contest and a which POV pushing spamming vandal can I make the happiest? what we need is leadership. BAG kinda does that, but it needs to assert itself more, that can be found by the discussions in recent issues. If someone has an issue with a member of BAG instead of forum shopping and dancing around the issue take it to WT:BRFA and lets solve it. you might not be happy about it, but its something that needs done. If someone wants to join BAG dont vote on it, lets have a reasonable logical discussion about it. with pros, cons and differing opinions. now lets all grow up and start doing what we should be doing improving the encyclopedia. β^_command 22:59, 19 April 2008 (UTC)

Perhaps I'm missing something, but I'm having difficulty understanding just how, exactly, community involvement would help, rather than hinder, selection for the BAG. On RFA, for example, it helps to have community involvement because one can bring up issues of incivility and huge judgment errors that would otherwise be missed, the reason being that issues of civility, judgment, and a cool head are observable characteristics that almost any member of the community can recognize and understand.

  However, being programming savvy isn't like that. It's not something that's inherently comprehended by the majority of socialized people in civilization. It involves complex terminology, abstract concepts, and meta-analysis of the complex interplay between multiple elements of network, server, program, and person. It sucks, but the average editor (as well as the average administrator), doesn't know who to program; therefore, the average editor (as well as the average administrator) should not be heavily involved in decisions regarding issues of programming, and I fear that a process like "requests for bot approvals group membership" would de-emphasize programming ability and practical skill sets in favor of implicit trustworthiness, gentle demeanor, and overall less important qualities that would be better suited for RFA than anything related to bot engineering.

  It understand that this does sound elitist, however, the only people who would presumably be able to detect someone demonstrating adeptness at coding are coders themselves. Asking non-coders to gauge coding sanity is similar to asking a dog to check your homework. We've seen this during bot RFAs where people have compared bots to SkyNet. That said, there are potentially serious long-term problems that could result from having a self-selecting group with no community input; however, should this present a significant concern in the future such that it clearly and demonstratively results in negative consequences for the project, I feel confident that, as always, we will find a way to resiliently bounce back. --slakr^\ talk / 23:12, 19 April 2008 (UTC)

I reject the idea that there must be some minimum level of technical knowledge before someone can voice an opinion on a nominee for the BAG. Further, issues of civility and judgment errors are just as relevant to BAG membership as they are to adminship. Would you really want someone incivil on the BAG? Or someone prone to making "huge judgment errors"? Better to let the community at large in on these discussions. —Locke Cole • t • c 08:55, 20 April 2008 (UTC)

I've been told that the community is welcome at BAG requests even under the old system. Haven't really seen it in practice, but that's what I've been told. So aren't you advocating an even larger amount of self-selection than the status quo? I must object; BAG is self-selecting enough as it is. rspeer / ɹəədsɹ 03:12, 21 April 2008 (UTC)

At no point did I argue for less community involvement than is already in place. I am, however, against an RFA-like requests-for-bot-approvals-group-membership type of thing. I feel that I was very clear about explaining that in my comments. --slakr^\ talk / 22:55, 21 April 2008 (UTC)

OK I'm a bit confused right now, and here would seem the best place to ask. What on Earth is all this nonsense on WP:RfA? Last I checked, RfA was called Requests for Adminship for a reason. I mean, we sometimes push the boat out and stretch to 'cratships which is a fairly natural extension, but for members of the BAG? That page is crowded enough already. Now there may (or may not) be consensus to go for some type of RfA-type voting system for BAG (to be honest I can't find anything more than bare bones discussion surrounding it) but RfA is most certainly not the place for it. Much as you may get more attention sticking it on a completely inappropriate page, that doesn't make it any less bizarre :-) Will (aka Wimt) 00:32, 20 April 2008 (UTC)

Quite clearly it's a first step before we transclude them in the sitenotice. Martinp23 00:34, 20 April 2008 (UTC)

Oddly Bureaucrats are handled on this same page, so I fail to see how this is at all irregular. BAG nominations will likely be a rare thing, just as Bureaucrat nominations are rare. Since they'll all be closed by Bureaucrats, it makes sense to keep them centralized instead of creating multiple separate pages that must be watched. —Locke Cole • t • c 08:55, 20 April 2008 (UTC)

I am now the second person to revert the policy back to the former wording for approving BAG members. I do think we can find a better way to approve them, but pushing changes into the policy over objections isn't it. — Carl (CBM · talk) 13:06, 20 April 2008 (UTC)

Concur. I'm willing to discuss a different system for selecting BAG members, but a major change without prior discussion is not the way to implement a different system. Oh the irony, that to encourage community involvement, a policy page gets a substantial rewrite without community involvement. Anyway, under a different BAG approval system, would BAG receive any buttons (like admins or bureaucrats do), such as access to the bot flag at special:userrights? Gimmetrow 23:07, 20 April 2008 (UTC)

Be bold in updating pages, there was some discussion apparently, and it's such an improvement over the existing system that (IMO) it makes sense to give it a trial run. Anyone disagreeing that it hasn't already improved participation is dreaming: look at the two nominees so far, that's probably eight times the amount of input BAG nominees normally receive. As to your second question: I'd love it if the devs added a user right which allowed for a type of bureaucrat that could only give or take the bot flag/right. If there were such a user right, I would support giving it to BAG members directly (removing the need for a bureaucrat to be contacted to give/take the flag/right). —Locke Cole • t • c 00:19, 21 April 2008 (UTC)

If a user right were to be granted, everyone would likely need reconfirmation. If we discussed this first, we might work out those details and avoid doing things two or three times. Gimmetrow 00:53, 21 April 2008 (UTC)

I would rather have some incremental movement forward than no movement at all. While it's likely very simple to add what I've suggested, I'm not confident it will get done in a timely enough fashion to warrant holding up changing the method of BAG member selection. BAG isn't adminship, if we even decide reconfirmations are needed if/when this new user right is available, it's unlikely to negatively impact anything to go through the motions again (FWIW, I would support giving the flag to anyone coming through this RFA-style confirmation; I think reconfirmation would be unnecessary for something as trivial as being able to give/take the bot flag). —Locke Cole • t • c 03:40, 21 April 2008 (UTC)

This is a trial, and why, exactly, do you oppose this change? It's already pretty obvious it is providing more input and opportunity for discussion than the old method. —Locke Cole • t • c 00:19, 21 April 2008 (UTC)

Wikipedia:Administrators'_noticeboard#BAG_Membership, we'll see which one works best --Chris 08:48, 21 April 2008 (UTC)

How many times do I have to say this? If you give people a place to vote, they will. That doesn't mean its the best idea and we should not consider any alternatives. Mr.Z-man 17:49, 21 April 2008 (UTC)

We're all (I hope) looking to explore all the alternatives. We're not (AFAIK) trying to solve the "problem" of RfBAG, because that would be process for process' sake. The problem is BAG's complete lack of authority except in as much as they control who gets the bot flag from the 'crats. The easiest way to give BAG some teeth seems to be to tinker around with RfBAG, so that's what's happening. Any alternatives to achieve the same end are good alternatives, but the status quo is simply not an option - even the ArbCom implicitly agree that BAG at current is broken, and somehow we need to fix it. If that requires bureaucracy, then we create bureaucracy: we're not an anarchy any more than a bureaucracy. If not a full RfA-style nomination, or even the compromise solution I offered above (which you seem to have shot down in flames), then what do you suggest? Happy‑melon 18:42, 21 April 2008 (UTC)

Well, 1) Why does BAG need more authority? What else do they need to do other than approve bots for flagging. I may be wrong, but I think enwiki is the only Wikimedia project with an group to approve bots, yet we aren't the only one with bots. 2) Your compromise was the same thing as RFA, just not on the RFA page. I cannot think of a justification for jumping from small discussions on an obscure talk page to a full RFA-style vote. 3) I believe I've given my suggestion at least twice now, apparently Chris G is the only one reading my comments anymore. Mr.Z-man 19:00, 21 April 2008 (UTC)

Why does BAG need more authority? because, once an account has been given a bot flag, that's when the potential for damage really begins. It's almost inconceivable that a bot intending to perform a blatantly damaging or controversial task would be flagged, with or without BAG; but once the account has the flag, there is currently almost no authoritative system to stop its owner deciding to perform more controversial tasks with it. What we've hashed over countless times all over the shop, is that BAG needs to be more responsible for the continued oversight of bot accounts and their operators, making sure that their operations stay within the bot policy, not just that they were in line to start with. I won't needlessly bring up specific examples (we all know the events we're reacting to, and it's not just the one bot or one owner), but history has shown that procrastination and apathy is currently more powerful than the BAG; that is untenable. If I had to sum up BAG's most important role, it would be as a first-point-of-contact for bots doing unexpected things: they have the technical knowledge to know if the bot is supposed to be doing whatever it's doing, and the authority to rectify the situation quickly if it's a problem. Because a bot can potentially be breaking dozens of pages a minute if it really is haywire, we have to give BAG more authority to make snap decisions without much discussion, and the authority to command admin action if required (since there is no requirement for BAG members to be admins, and not all of them are). That authority does not equate to power, but it does equate to trust: an admin responding to a BAG request to block a renegade bot must know that the community trusts the BAG member to know what he's talking about. Higher level of community trust means higher level of community involvement in the selection process, and I'm afraid to say that so far it doesn't seem that either the old method or your AN proposal is working in that regard. CrisG's nomination has been on WP:AN for twelve hours now: the response? Zip, nada, zilch. Of course I would be delighted to be proved wrong, for it to be found that the extra bureaucracy of the trial RfBAG process is not necessary, but at the moment it's simply not selling itself to me! As for the old system, well, I was (to be honest) shocked to find Coren's old-system nomination. No 'crat would close an RfA on the back of two supports - it's only one step above threshold for rollback, let alone proxy access to Special:Userrights. In just 48 hours Coren has found fifteen times as much community comment on his suitability for BAG at WP:RFA as he did in almost two weeks at WT:BAG. I am actually glad that the discussion has been two-sided - it is a clear demonstration that the more people are involved, the more thorough the discussion is, and the more reliable the consensus. Happy‑melon 21:04, 21 April 2008 (UTC)

Incidentally, regardless of the result of how I will fare personally over there, the attempt is in my opinion an unmitigated success for the procedure. 26 comments, and a number of relevant and interesting questions. Can anyone explain to me why, besides a blind edict of "voting is bad and bureaucracy is evil", this could be viewed as anything but an absolute success? — Coren ^(talk) 22:26, 21 April 2008 (UTC)

What exactly is it a success for? There isn't agreement that RFA is the right location for the discussion, but I don't think anyone disagrees that RFA will draw more comments than the BAG talk page. I think a middle way is more reasonable. — Carl (CBM · talk) 22:30, 21 April 2008 (UTC)

There isn't disagreement either: the folks up in arms over this respond once or twice here, then proceed to ignore replies acting as if ignoring it will make it go away. That's just not the case. There is support for this, well reasoned support at that, backed by clear evidence that this is a better solution than the old method. You mention a "middle way" but fail to describe what you're suggesting in detail. Please remedy that so we can move forward. —Locke Cole • t • c 22:44, 21 April 2008 (UTC)

There is not consensus for the implementation of this idea. for an idea to become policy it must have consensus, something this idea does not have. instead of attempting to force your POV why not try and find consensus. β^_command 22:52, 21 April 2008 (UTC)

See also this which shows three users for this idea, 23 against and no other comments. Please dont try forcing policy that has zero consensus. β^_command 22:59, 21 April 2008 (UTC)

That page also shows opinions clearly stating that Administrators are all incompetent fools, that the ArbCom is about to bring the end of times. Clearly very valuable. — Coren ^(talk) 23:18, 21 April 2008 (UTC)

There isn't consensus for the current method either. Since you seem unable, or unwilling, to discuss this, I believe it's time to shutdown BAG nominations until we can reach a consensus for something. —Locke Cole • t • c 23:17, 21 April 2008 (UTC)

(←) I don't see how you can say there isn't disagreement, given that at least five users have disagreed in this very thread.

My thought is to discuss the BAG nominations on a high-activity page like the technical village pump, which is frequented by the sort of people who have the technical background to judge BAG membership. I don't see any reason to think that most people who routinely participate at RFA have that background. — Carl (CBM · talk) 22:52, 21 April 2008 (UTC)

Given that the (currently) 28 users opposing this on the adminship poll and the people opposing here don't actually count for anything anymore, we're just an endemic problem, I'm going to stop wasting my time arguing this. I really don't care enough. If people want bureaucracy for the sake of bureaucracy, fine. Mr.Z-man 15:46, 22 April 2008 (UTC)

Bot list

Where is, or how do I generate, a list of active bots. Ideally I'd like to be able to easily import it into VandalFighter. Thanks in advance. - RoyBoy 03:52, 21 April 2008 (UTC)

Wikipedia:Bots/Status. I'm not sure if this absolutely accurate but to my knowledge it's the best we've got.--Dycedarg ж 04:09, 21 April 2008 (UTC)

this list is autogenerated and will be bang up to date - of course this will not include bots which are not flagged, and will include bots which are still flagged but are inactive. This is the equivalent API query, which you can easily import into a script. Happy‑melon 08:34, 21 April 2008 (UTC)

I've got an approximate list of approved bots (taken from the list above, the API query above, and some internal lists of approved bots) in a database on my server. If you like, I could throw together a page to generate a list of all approved bots. — Werdna talk 03:41, 22 April 2008 (UTC)

Interesting, I haven't looked at it closely yet; but I would need to work out a way to strip the bot list and add the formatting to paste the list directly into the .dat file of VF. Although VF does import the Admin list automatically, so there must be a way to import the Bot user list... though the obvious solution is to ask the maintainer to do it. But I've had limited success with VF versions after 3.3. - RoyBoy 04:31, 24 April 2008 (UTC)

Wikipedia:Bots/Requests for approval/Approved Wikipedia:Bots/Requests for approval/Approved/Archive 1 --uǝʌǝsʎʇɹoɟʇs(st47) 13:09, 25 April 2008 (UTC)

Edit warring on a policy page

Ideally, please stop edit warring on the page. If you feel that the wording should revert to its prior version (i.e., "revert back to the way it's always been until there's consensus to change it"), simply state so below and we can continue discussing here about proper changes instead of edit warring. If you feel that the new process should stay until discussion has concluded (i.e., "install the new process until there's consensus against it"), then please also state that here.

• Revert — I believe there is no reason to introduce more bureaucracy to BAG selection at this point in time per my comments above. --slakr^\ talk / 23:01, 21 April 2008 (UTC)
• Revert: Yeah, I've been watching this situation and from what I can tell, there was never a consensus to change it in the first place. - Rjd0060 (talk) 23:04, 21 April 2008 (UTC)
• I already expressed my opinion by reverting to the old wording. I'm not planning to revert again, at least not in the forseeable future. — Carl (CBM · talk) 23:10, 21 April 2008 (UTC)
• Revert. --MZMcBride (talk) 23:17, 21 April 2008 (UTC)
• Leave old style β^_command 23:23, 21 April 2008 (UTC)
• Revert to the old style, until a clear consensus to change to the RfA-method can be demonstrated. SQL^{Query me!} 03:47, 22 April 2008 (UTC)

This is a joke

Yes, let's "discuss" this. Oh wait. That's right, we tried that (^^^), except when points were made people seemed to disappear (or suddenly have an aversion to posting on this particular talk page, but have no problems joining discussions elsewhere). Oh, and don't forget, voting is evil. Except for people who think voting is evil, they can vote whenever the mood strikes... —Locke Cole • t • c 23:21, 21 April 2008 (UTC)

Locke Cole, Im sorry but do not edit war to force something that does not have consensus. β^_command 23:23, 21 April 2008 (UTC)

You realize the same applies to you? —Locke Cole • t • c 23:26, 21 April 2008 (UTC)

I have consensus on my side, its been long standing policy. if you want to change policy please create consensus. Also please revert the edits that were made after protection. β^_command 23:31, 21 April 2008 (UTC)

You don't have any more consensus than I do. It's clear there's disagreement about the current system, so maybe shutting down BAG nominations will prod people into discussing this rather than ignoring it. —Locke Cole • t • c 23:37, 21 April 2008 (UTC)

Unfortunately, when a page resorts to edit warring, enumerating assenting and dissenting opinions in the most concise way is one of the only ways to wade through the muck and establish consensus. There is little ambiguity over what change is proposed, so a simple "yes, change it to that version" or "no, don't change it to that version" is appropriate to settle the edit war.

  Additionally, consensus seems to generally default against change when there is no consensus to change. For example: on RFA, if a candidate does not achieve clear consensus, the change to promote him does not occur. On AFD, if there is not consensus to delete an article, the change to delete the article will not occur. In policy changes, the same seems to hold true. If there isn't consensus to change a policy to a certain version, then we default to not having that version on the policy page until there is consensus to do so. Since there was confusion as to whether or not a change was being made and whether or not it should be made, I felt it necessary to consequently rectify the situation. --slakr^\ talk / 23:32, 21 April 2008 (UTC)

I think this situation is unique in that there are a number of editors who think the current system is broken, and there seems to be a similar number of editors who dislike the new proposed method. It seems neither solution has a consensus in support of it, so, IMHO, we should scrap both and try to work on something that can gain consensus. —Locke Cole • t • c 23:37, 21 April 2008 (UTC)

Its not unique. You just described rfa. Every one agrees it broken but the can't agree on how to fix it --Chris 23:48, 21 April 2008 (UTC)

FWIW, I'd support shutting down RFA as well if what you say is true (I know it was true long ago, but I haven't participated much in RFA discussion recently). If nothing has consensus, then nothing should be available until some compromise is reached. Maintaining the status quo only helps to maintain inertia, which only makes it harder to change things the longer things go unchallenged. —Locke Cole • t • c 23:53, 21 April 2008 (UTC)

As an aside, am I the only person who reads that customized {{policy}} tag at the top of the page which says, quite clearly, "Parts of this policy have been recently reworked and rewritten. Those parts may still be in development, under discussion, or in the process of gathering consensus for adoption."? The appropriate thing isn't to revert, it's to edit/improve or discuss as the policy is still being developed/reworked. —Locke Cole • t • c 23:37, 21 April 2008 (UTC)

Locke Cole: I'd support shutting down RFA as well if what you say is true — I feel that that might be a straw man of what I was saying. To clarify my point as I said before, the trend, as far as I have seen, is that we tend to value establishing consensus first; then we reflect it in our policies/actions— not the other way around. Again, most of the time on RFA, we don't make a user an admin first and then discuss the action of making him an admin. And again, most of the time on AFD, we don't delete an article and then decide if deleting it was the best course of action. Yes, there are both ARB and DRV to reverse decisions, however, the overwhelming majority of the time, we discuss controversial actions before implementing them.

  The entire process of changing consensus is, indeed, sparked by a bold action. However, as part of building consensus, that action doesn't necessarily stick, and like it or not, discussion is critical to actually determining just what edit, exactly, needs to happen in order to accurately reflect it. You feel that maintaining the status quo may lead to stagnation; however, I would argue that implementing change and forcing a consensus against implementation presents an equally problematic scenario of avalanche policymaking. The real life analogy would be somewhat like a dictatorship, whereby a decree is passed and the only way to override it would be through an act of congress. Given enough changes to policy in a short enough period of time, it would be impossible to demonstrate adequate opposition to all motions for change, thus creating the illusion of consensus through silence when there actually is none (because nobody has the literal time to state their objections repeatedly).

  Therefore, I hope you will sympathize with our current approach: when a large change is proposed, we tend to default back to the status quo until consensus is actually determined to support the change. Thus, "no consensus," at least, from what I've seen, defaults to opposing a change in the status quo unless the change is non-controversial. --slakr^\ talk / 01:21, 22 April 2008 (UTC)

Discussion is critical, but over 24 hours later (going on 48 in some cases), with many of the opponents of the new method editing elsewhere (but ignoring this discussion), I opted to try again given the lack of continued reasonable objection. As to the idea that we must revert to the status quo, I disagree. Is it appropriate in AFD? Yes, AFD asks a very simple question: "should this article be deleted?". Is it appropriate when deciding policy? No, policy matters are asking if anything has support/consensus. There's no consensus for either the new method or the old method. In this case we have a clear and simple solution: neither process should be used, and discussion should continue until something agreeable can be found. This has the bonus of forcing the issue to be dealt with, rather than ignoring it and letting continued inertia carry it forward. BTW, what I said was not a straw man: you're the one that brought RFA in to this discussion, I was just being clear that I'm not being selective with my opinion; if RFA is indeed in state similar to what we have here, I believe it should be shut down until something which has consensus is available. —Locke Cole • t • c 01:47, 22 April 2008 (UTC)

If you would like, then, start up a !vote to establish the consensus that the bot policy and/or BAG is broken and must be stopped, at all costs, until a fix is determined. It's the quintessential {{sofixit}}. It's not like it requires a grand policy change— just someone who is willing to start a new header and say what you want to be done. If you feel that your way is the right way to go and that nothing should be done in either direction until we decide a new plan, then suggest it. If people agree with you, then it'll happen. *shrug* It's as simple as that. It's totally up to you, because that's the beauty of having a community that runs itself: if you're ticked about something, and people agree with you, then the people can demand action. It's your choice to take the first step. --slakr^\ talk / 04:18, 22 April 2008 (UTC)

Protection

I've locked this page until people can agree on it. Thanks, Majorly (talk) 23:25, 21 April 2008 (UTC)

{{editprotected}} To complete the edit I had started before Majorly protected; to include both version to clarify things as they stand in the dispute— specifically:

---

Members of the group are experienced in writing and running bots, have programming experience, understand the role of the BAG in the BRFA process, and understand Wikipedia's bot policy. Those interested in joining the group should make a post to the talk page explaining why they would be a good member of the team and outlining past experience, and then should make posts to relevant community noticeboards to invite community participation. After ten days, an uninvolved bureaucrat or an uninvolved current BAG member will close the discussion.

A currently proposed methods is that members of the BAG are to be elected in a consensus evaluated by the Wikipedia bureaucrats in a process similar to the Request for Adminship.

This process would take place in the same way as Requests for Adminship, as a subpage. The community may ask questions of the candidate and express their views in the voting sections. A 2/3 majority is expected after a one week period to promote the user. A 2/3 majority is also required to, through the same process, remove a member. Both types of request should be closed by a bureaucrat, who has discretion to weigh arguments and determine consensus in close votes or in extreme circumstances.

At least two current members of the BAG have voluntarily sought approval with this process, as discussion progresses, to validate the concept.

---

— Coren ^(talk) 23:36, 21 April 2008 (UTC)

Anybody have objections to Coren's edit? --Chris 01:10, 22 April 2008 (UTC)

My only objection would be to removing the note that no current method has consensus for continued use. I believe that accurately reflects the situation as it stands right now. —Locke Cole • t • c 01:49, 22 April 2008 (UTC)

Why has someone added a description of a proposal to a policy page? The old system isn't even mentioned right now. Gimmetrow 04:01, 22 April 2008 (UTC)

From the customized {{policy}} tag at the top of the page: Parts of this policy have been recently reworked and rewritten. Those parts may still be in development, under discussion, or in the process of gathering consensus for adoption. —Locke Cole • t • c 05:06, 22 April 2008 (UTC)

(from WP:VPP)

The games surrounding this have got to stop. It was moved in place, without consensus, and now that that's not working, apparently edit-warring, and claiming neither version has consensus is being tried. ENOHGH. This war of attrition has to stop. SQL^{Query me!} 04:06, 22 April 2008 (UTC)

At this point, I think the approach taken trying to push this through makes it less likely to gain acceptance, even if a lot of the issues were worked out. Did we learn nothing from ATT? Gimmetrow 04:43, 22 April 2008 (UTC)

It's a bit silly to compare this to ATT; ATT was taking three policies and combining them into a single policy page (which was, as I recall, heavily rewritten). We're talking about one somewhat major change to a process that only forms a small part of the overall policy. Not exactly Earth-shattering stuff, really... —Locke Cole • t • c 05:06, 22 April 2008 (UTC)

Although the BAG change drew the most attention, there are other changes on the page, some I agree with and some I don't. And because of the way the new draft was introduced, we're mired down discussing the procedure to change policy rather than the actual changes, even the good ones. And that's comparable to ATT. BOTS is much more narrow than ATT and could have been hashed out in a few weeks, had it been started as a draft on a subpage *here*. But that's water under the bridge now. Gimmetrow 05:26, 22 April 2008 (UTC)

I'm trying hard, really hard, to assume good faith throughout this dispute. But in my experience, at least so far as this group of people goes, unless you make a bold change people will either respond once and ignore your proposal or they'll ignore it outright. For whatever reason there seems to be a major aversion to change here at WP:BOTS. So somehow I doubt doing as you suggested would have accomplished much beyond having a proposed policy revision die a painfully slow death before being archived. —Locke Cole • t • c 05:37, 22 April 2008 (UTC)

Well, you have a point there. My own policy proposals have not been addressed by this rewrite, though, so I'm not particularly happy. Gimmetrow 05:41, 22 April 2008 (UTC)

Pros and Cons

Ok in order to make a proper judgement on what process we should use for electing bag members I think we should weigh up the pros and cons of each idea.

Old Version:

Pros:

• Not much bureaucracy
• Not political
• Less likely to bring silly opposition
• People who do comment are likely to have technical experience

Cons:

• Very little input
  • Almost no involvement outside the bot-operating community
  • Input largely from existing BAG members
• Can be mildly intimidating
• Little-known system in an out of the way venue reduces scrutiny

RBAG

Pros:

• Lots of input
  • Involvement from outside the bot-owning community
  • Closure by bureaucrats makes process more obviously transparent
    • Crats were initially supposed to close noms in the original version as well, but could not be convinced to do so
• Reuses a known system in a visible venue

Cons:

• Overly Bureaucratic
• RfA regulars may not be familiar with the technical experience of the candidate or the requirements for BAG membership
• Overly political
• Can be extremely intimidating
• Subject to the same weaknesses as RfA in general

Discussion

I've just gone and listed the dead obvious, add more as you think of them. Hopefully by having a clear list of the problems with each idea we can make a decision on which one to use. If you come up with another idea feel free to list it --Chris 04:29, 22 April 2008 (UTC)

I added a few, the only one I feel I need to explain is 'not political'm under pros for the present system. I say so, because I got in, just after rather harshly criticizing BAG, and the decision to move back to voting members in, instead of allowing them to add themselves. SQL^{Query me!} 05:10, 22 April 2008 (UTC)

I also added a few, and rewrote one - I hope it's still neutral. Happy‑melon 08:56, 22 April 2008 (UTC)

I've added a few as well. — Coren ^(talk) 12:23, 22 April 2008 (UTC)

Added one bullet. My suggestion has always been to move it from the BAG talk page somewhere else, but not RFA. I suggested the technical village pump as one possible location. What do people think about that? — Carl (CBM · talk) 13:52, 22 April 2008 (UTC)

That could possibly work very well. I'd like to see the format drastically changed as well (away from the present formats, both of which (the newer one moreso) strongly encourage voting over rational discussion. SQL^{Query me!} 20:48, 22 April 2008 (UTC)

I am always keen to move away from voting where practical, but I'm not sure how much discussion is ever going to occur. Either a !voter thinks that a candidate is suitable, or they think they are unsuitable - since each is entitled to their opinion, there's not much productive discussion that can be undertaken. In controversial cases, yes, past actions or traits will require, and receive, a more loosely organised discussion, but offers of support rarely extend to more than a single sentence. All we'd do by forcing people to "discuss" their !votes of support rather than just listing them would be to make it more difficult for a 'crat to evaluate the consensus. Plus it would discourage those who have unconditional support (and hence nothing substantial to say) from commenting at all, IMO. Happy‑melon 20:58, 22 April 2008 (UTC)

Well putting in on an doesn't seem to be working. The Village Pump might work better. However I do agree with Happy-melon that it will be hard to actually get discussion started --Chris 04:33, 23 April 2008 (UTC)

I suspect it would be even less successful on the VP. My biggest problem with BAG is the seeming self selection done for membership: the RFA method solves this by moving the decision to a bureaucrat, and by involving more of the community (not just current BAG members or the occasional passerby). FWIW, Happy-melons comments sum up my feelings overall as well. —Locke Cole • t • c 04:47, 23 April 2008 (UTC)

I definitely and very strongly believe that, wherever we hold RfBAGs and whatever the format, asking a bureaucrat to evaluate the consensus and close the discussion is an easy way to improve the integrity and transparency of the process. The extra workload for the 'crats is inconsequential, and it's what crats do best (and largely what we created them for). That, combined with an increased input from outside existing BAG members (and those few bot operators who watchlist WT:BAG), will solve the legitimate criticism of self-selection. Happy‑melon 10:26, 23 April 2008 (UTC)

'Crats would close any BAG nomination that wasn't clear in the old talk page system, too. There should be some examples in the archives. Most BAG nominations are unanimous, though, and don't require a complex evaluation of consensus. Gimmetrow 04:55, 24 April 2008 (UTC)

I'll take your word on it that crats have closed BAG noms in the past, but the troubling part is when a non-crat closes them (which I believe is more the norm). And we're still leaving off the lack of input (usually a handful of people, often less than four, and some or all usually either BAG members already or bot operators or a mix of both). RfBAG simply gives us a lot more input (and due to the exposure, is more likely to turn up any issues we should consider before promoting them). —Locke Cole • t • c 05:06, 24 April 2008 (UTC)

Generally, they're closed by *anyone* when unanimous. I closed 2 myself, when I could not get a crat to do so. How would you suggest 'unanimous' be closed instead? SQL^{Query me!} 05:32, 24 April 2008 (UTC)

By a 'crat, just as an RFA wouldn't be closed by someone unable to actually flip the bit (I know there's presently no bit setting going on, though that's something else I'd like to see in the future (BAG members being able to give/take bot flags). What's your thoughts on the added input at RfBAG? —Locke Cole • t • c 06:00, 24 April 2008 (UTC)

As Martinp23 jokingly suggested above, if it were in the site notice it would get even more attention. I'm skeptical of the need for that level of attention. If a userright were involved, even something you call "as trivial as being able to give/take the bot flag", I would be less skeptical. The bot flag was originally handled by stewards. Gimmetrow 06:24, 24 April 2008 (UTC)

I would have absolutely no objection to the creation of a 'botapproval' usergroup, with /− 'bot' at Special:Userrights, but I wouldn't dream of supporting it if promotions to that usergroup were to be on the basis of two supports at WT:BAG. Are we approaching a possible compromise here? The "unnecessary" level of attention of RfBAG balanced by a technical ability that genuinely requires such scrutiny? Happy‑melon 09:27, 24 April 2008 (UTC)

(undent) I should point out that, as far as the "get more community input" metric is concerned, the two currently running RfBAGs are unqualified successes; in particular Cobi's, which is unburdened with the "protest over the process" comments, has gotten 31 editors to contribute— more than the entire current BAG have gotten with the old system combined! I see neither the flood of ridiculous opposition nor the popularity blind support doomsayers have been dreading. Intelligent questions have been asked of the candidates, and the comments seem to indicate that the editors who chimed in understand what this is about.

I know I've been bold to the point of being pushy in bringing those changes; but Wikipedia has grown to acquire the inertia of a black hole, and changing anything in policy is now an exercise in futility unless you hammer it in— discussions over minute points of policy manage to swamp out anything productive until everyone leaves in disgust, or apathy wins the day. Stop whining about how I've worked to bring those improvements to come to fruition and start evaluating the results on the merits.

The old system was broken. The Arbitration Committee even took the effort to point out that it needed fixing. The proposed system successfully manages to be more transparent, and to bring in (a lot!) more community involvement, which was unarguably the primary concern expressed with the old system.

Yes, it's "more bureaucracy". Perhaps we need to collectively grow up and stop rejecting everything just because "bureaucracy is teh evilz"; minimizing bureaucracy is a laudable principle when it doesn't hinder, not an objective in itself. The RfBAG is, quite opposed to the apparent feeling, very little bureaucracy. It reuses a system which is well know, well understood, and in a venue where such is expected. It is less added bureaucracy than inventing yet another selection system, in a venue nobody looks in, and with new different rules nobody is familiar is. — Coren ^(talk) 15:18, 23 April 2008 (UTC)

Damn right Coren, consensus is for losers. If you can't get people to agree to the changes you want, just force it. SQL^{Query me!} 20:24, 23 April 2008 (UTC) Agree, not helpful, sorry. SQL^{Query me!} 21:10, 23 April 2008 (UTC)

That is not what I said, and I would appreciate not misrepresenting me; but "consensus" also doesn't mean "unanimity" or "nobody objects". — Coren ^(talk) 20:33, 23 April 2008 (UTC)

(ec)I don't think it's fair to say that Coren has been forcing something against consensus; if there is one thing that is manifestly not present in this discussion, it is consensus, either for the new system or the old. The merits of both systems need to be weighed up once we are in posession of all the facts (I want to see how the crats close the open RfBAGs, for one thing), but I feel that right now we have a useful equilibrium: we are in a very real sense at a tipping point, and we can either return to the old system, or proceed to the new, with very little difficulty, when consensus is found for one or the other. Coren (or any other user) would be out of line to attempt to push this change any further without consensus, but I think it's unfair to deride the valuable task that Coren has performed to get us to where we now stand against the enormous inertia of the policy system. Happy‑melon 20:34, 23 April 2008 (UTC)

Let's not lump these together, there's enough hubbub about this already. We are discussing the new system, and, as you agreed, there is presently no consensus for that (as discussed here, and at WP:BN). While we should continue the discussion (carl, and a couple others seemed to have some good ideas and proposals that may be worth discussing), I think, the exercise with RFBAG should be closed at this point, and revisited or reopened when it has the support of the community (or, if it's binding, perhaps we should allow the open ones to close, I'm honestly not sure it's ever come up before). A note, just because people participated, in a new process (which often garners a lot of attention initially), does not mean they support doing it that way, and should not be interpreted as such. It's unknown if any of them came here, and read the discussion (except for those that commented here), or, read / even care about the policy. I'd like to continue on, but I've got to get dinner ready. SQL^{Query me!} 21:00, 23 April 2008 (UTC)

I think we should let the four open RfBAG nominations run their full course, and see how the 'crats deal with them; but I don't think we should encourage any more to open at this point. I think it's arguable that, amongst the participants, those who were opposed to the process as opposed to the candidate made their views very well known, while those who support the process seem to have restricted themselves to comments on the candidates; as you say, we shouldn't try to read too far into the comments. The level of interest as compared to concurrent RfAs and previous RfBAGs, however, would be useful to consider. Once the discussions are closed I'll try and work out how the participants are divided in terms of bot-operators/non-bot-operators, those who also posted here vs those who didn't, etc. Should throw up some interesting data. Anyway, we seem to agree that now (or certainly once the trial RfBAGs have closed) is the time to step back and rationally discuss which is the better method for selecting BAG members. Happy‑melon 21:23, 23 April 2008 (UTC)

Hear, hear! I, for one, put forth my candidacy to gather that information without which any decision would be ill-advised; speaking for or against a system without any information on how it performs in practice just irks my rational fibre. I had most certainly not expected that others would jump ahead and do so as well, but in hindsight this may have been for the best: their candidacies are considerably more representative than mine could hope to be given that it doubled as a focus on the process itself.

And, unlike what some people may seem to believe, my position isn't "RfBAG or bust", although I (obviously) think this is a workable system that correctly addresses the legitimate concerns about transparency and wider community participation that have been raised repeatedly here and elsewhere. I do beleive that unless I had been bold and simply went ahead, the discussion would have mired down in the hypothetical merits of the proposed system and never gone anywhere. — Coren ^(talk) 22:52, 23 April 2008 (UTC)

Moratorium on RBAG nominations

To help calm things down, and since half a dozen nominations are already up, I've moved to place a moratorium on adding new nominations to RBAG while community deliberations of the process continue. Hopefully as the current batch of nominations take place the community can get a feel for how they would play out and come to a solid consensus on this issue. krimpet✽ 03:22, 28 April 2008 (UTC)

Ugh... things are calm, and the process is working fine. Why introduce a moratorium when it's pretty obvious people are interested in trying this out and seeing if/how it works? I'll refrain from reverting, but I strongly object to this idea of a moratorium. —Locke Cole • t • c 03:27, 28 April 2008 (UTC)

If people are interested in trying it out, they can gladly participate in the half dozen nominations that are ongoing; too many nominations and it may become chaotic. It's best to take this one step at a time. krimpet✽ 03:31, 28 April 2008 (UTC)

You speak of "chaos", but I'm not seeing anything remotely close to chaos with how things have been going. It's been quite orderly and proven itself well. A moratorium seems unnecessary at this point. —Locke Cole • t • c 03:43, 28 April 2008 (UTC)

Actually, Locke, I think the current batch will be enough to allow everyone here to evaluate how things have gone; certainly there is no need to rush things any further, and a pause is quite reasonable. — Coren ^(talk) 04:28, 28 April 2008 (UTC)

Especially given that when the "community" was polled on this issue, 4 people supported having BAG membership on RfA, while 31 opposed (see here). That's 89% opposed. Surprisingly, one of the people currently under an RfBAG said "We need community involvement. THis is not the best way to solicit it." Which has left me entirely confused. --MZMcBride (talk) 04:50, 28 April 2008 (UTC)

With all due respect, MZM, it's because that "poll" is quite worthless; as you are well aware. That area is mostly populated by malcontents, and is little but a crying wall. As for someone changing their mind, well, that happens. — Coren ^(talk) 12:33, 28 April 2008 (UTC)

No need to call fellow editors names just because they disagree with you. SQL^{Query me!} 13:02, 28 April 2008 (UTC)

Which would be why I didn't. — Coren ^(talk) 17:09, 28 April 2008 (UTC)

Bot flags

Some mention should made in the policy of why some bots do not have the bot flag set. Bots whose sole purpose is to revert vandalism do not use the bot flag. From what I can see, this is just an unwritten rule/practice. -- Mufka ^{(u) (t) (c)} 02:21, 25 April 2008 (UTC)

There was some verbiage to that effect, but the consensus that emerged is that while some older bots may not have the flag for historical reasons, all new bots should. AV bots would simply not flag their edits as bot. — Coren ^(talk) 23:33, 25 April 2008 (UTC)

Perhaps this needs to be brought up again because it doesn't seem like this is the case. For example, Pseudobot was flagged as a bot and its edits were tagged as bot edits. I talked to ST47 and he removed the bot flag. I'm only asking about this because I want to know how it works and it helps if it is consistent. -- Mufka ^{(u) (t) (c)} 23:52, 25 April 2008 (UTC)

What do you mean by "AV bots would simply not flag their edits as bot"? Since when is there a way for a bot-flagged account to refrain from having their edits marked as bot edits? Isn't that the whole point of the bot flag? (Aside from the non-obvious reduction in technical limitations amyway.)--Dycedarg ж 00:47, 26 April 2008 (UTC)

A patch that I made to MediaWiki allows bots to use &bot=0 when posting an edit. If that is done, the edit will not be marked as bot, even if the bot has the bot flag. The reason for this patch is bots get higher limits than other users from the API and from rollback (and maybe some other things), so just marking an edit as bot is not the only thing the bot flag is useful for. Bot flags are also sometimes used to simply denote that a bot has gone through BRFA and been approved by the BAG, when the speed of the edits would not flood recent changes (like bots that make one edit per day that have been approved still have the bot flag). -- Cobi^(t|c|b) 01:44, 26 April 2008 (UTC)

By the way, ClueBot is flagged, but its edits are not marked as bot edits. -- Cobi^(t|c|b) 01:45, 26 April 2008 (UTC)

Interesting. I knew about the higher rates for API and such, but always assumed that most bots that don't flag their edits operate slow enough for that not to be a problem anyway. I never noticed that ClueBot had a flag. That thing about bots being able to make unflagged edits with &bot=0 should probably be noted somewhere in the policy; I don't think it's particularly common knowledge.--Dycedarg ж 03:12, 26 April 2008 (UTC)

I don't think this should be in the policy, because it's generally of little interest to anyone but bot writers, but it should certainly be explained somewhere conspicuous. Someone did tweak the policy to explain the reason why some bots did not have the flag but now should, however. — Coren ^(talk) 13:03, 26 April 2008 (UTC)

That was me, actually, and that edit was pretty much what I meant by noting it in the policy. Just that it was possible, not necessarily how to do it. I would agree with you that it should be mentioned somewhere. Maybe on the Creating a bot page?--Dycedarg ж 17:37, 26 April 2008 (UTC)

Sounds logical to me; giving &bot=0 is an implementation detail, after all. — Coren ^(talk) 17:39, 26 April 2008 (UTC)

WP:Bots

I think it would be handy if WP:Bots was an overview page about bots with links to relevant pages, rather than a redirect to here. This page isn't especially helpful for the non-technical who just want to know a bit about wiki bots. --kingboyk (talk) 15:13, 25 April 2008 (UTC)

I'll work on that. — Carl (CBM · talk) 12:36, 28 April 2008 (UTC)