Wikipedia:Administrators' noticeboard

Welcome — post issues of interest to administrators.

For urgent incidents and chronic, intractable behavioral problems, use Wikipedia:Administrators' noticeboard/Incidents. If you are new, try the Wikipedia:Teahouse instead. Do not report breaches of personal information on this highly visible page – instead use Wikipedia:Requests for oversight. For administrative backlogs add {{Admin backlog}} to the backlogged page; post here only if urgent. Do not post requests for page protection, deletion requests, or block requests here. Just want an admin? Contact a recently active admin directly. If opening a close appeal, use {{RfC closure review}} When you start a discussion about an editor, you must leave a notice on their talk page. Pinging is not enough. You may use {{subst:AN-notice}} ~~~~ to do so. Sections inactive for over seven days are archived by Lowercase sigmabot III.(archives, search) Start a new discussion

Current issues

Adding useless revisions to pages to make them undeletable

Just FYI: [1]. And I'm sure Tim's not the only one. --bainer (talk) 03:14, 4 February 2008 (UTC)

Fully support Tim's comment. The history of the Main Page is a joke. - auburnpilot talk 03:16, 4 February 2008 (UTC)

While I'm sure this was done with good intentions... boy, this strikes me as spectacularly bad judgement. It also highlights a function that could be used in ways less benevolent. This ignores the fact that the Main page appears to have been deleted once *(Per CSD G6!) in the process, albiet briefly. I'm a new admin, so maybe there's conversation on this topic I've missed... But, I have to agree with Tim's comments. UltraExactZZ ^Claims ~ Evidence 03:24, 4 February 2008 (UTC)

Tim has already blocked BetacommandBot for one week, due to "abuse of system resources". [2] - auburnpilot talk 03:31, 4 February 2008 (UTC)

Ok, so how did BCB, a non-admin bot, edit a protected page. I'm seeing some odd things going on in the page history with restoring deleted edits, etc. And shouldn't something like this get approved somewhere before its done? Isn't this the purpose of WP:BAG? I know there is a technical switch that would make the main page undeletable by anyone including the devs, which isn't flicked since we don't want to do anything that can never be undone. So this is basically doing that (aking it undeletable) the way I see it, which is somewhat against consensus IMHO MBisanz ^talk 03:37, 4 February 2008 (UTC)

The edits were made in userspace and the page was moved by an administrator into the mainspace. Nakon 03:37, 4 February 2008 (UTC)

Was this discussed anywhere beforehand? --Rory096 03:50, 4 February 2008 (UTC)

Nope, no approval and no consensus that this was need - a well deserved week long block. The main page history is destroyed now. Ryan Postlethwaite 03:54, 4 February 2008 (UTC)

Isn't there some way to selectivly delete prior revisions like oversight but not as extreme). If it cant be automated, maybe hand-deleting (oy!) will be required. Worst case, there might be a consensus to oversight the interjected edits (yea I know its against policy, but I'm not seeing the harm). MBisanz ^talk 04:03, 4 February 2008 (UTC)

Oversight works on a single revision at a time; good luck finding an oversighter that's willing to go through 5,000 revisions by hand. ;-) Kirill 04:09, 4 February 2008 (UTC)

According to BCBs history, it made about 1100 of the edits to that page, still it would be an unfair burden on an uninvolved oversighter to have to do that. MBisanz ^talk 04:21, 4 February 2008 (UTC)

I believe the whole point is that nobody other than the devs can delete pages with over 5000 edits. They'd have to be the ones to remove all of the bot edits, if I'm remembering correctly. - auburnpilot talk 04:05, 4 February 2008 (UTC)

I don't see how Betacommandbot's block is any more "well deserved" than a block of the administrators who collaborated on this venture would be. All the bot did was make a bunch of null edits to a user subpage. Mike R (talk) 04:06, 4 February 2008 (UTC)

Frankly, I'd have blocked the users involved as well; the administrators, in particular, are expected to consider the consequences before they do something of this sort. Kirill 04:16, 4 February 2008 (UTC)

The admins involved obviously screwed up, but remember blocks are preventive, not punitive. --Rory096 05:09, 4 February 2008 (UTC)

It was obviously discussed somewhere beforehand. My guess is IRC. Mike R (talk) 04:09, 4 February 2008 (UTC)

I think the reason the bot was blocked and the admins weren't is that 1. Tim was afraid of the bot doing this to many other pages and had to act quickly. 2. Bots go through a special process to get the BOT flag and that process allows harsher action when they mess up, Admins, generally have the grace of an RfC/AN/Arbcom discussion. MBisanz ^talk 04:15, 4 February 2008 (UTC)

I do think blocking Betacommandbot is justified, since as far as I know, it isn't approved for this (though maybe I'm wrong, given the massive list. At any rate, it seems like it should definitely have been discussed first (on wiki), and was probably a bad idea. Apparently, East718 was the one to do the move, and it was done in his userspace. It might also make sense to block him and Betacommand. Superm401 - Talk 04:20, 4 February 2008 (UTC)

{unindent) Tim Starling has proposed here [3] that East 718 be de-sysoped for his behavior in this matter. What forum should this request be discussed in? MBisanz ^talk 04:47, 4 February 2008 (UTC)

Not sure what the proper venue would be, but I'd assume ArbCom. Thankfully one of the devs has removed the bot's junk edits from the Main Page history.[4] - auburnpilot talk 04:55, 4 February 2008 (UTC)

This is kind of silly anyway. Why do we restrict administrator tools like delete and protect to administrators? Because we trust them to know what they're doing. We restricted deletions to pages with less than 5000 revisions because it's not obvious that it's going to break the site. This is something completely different — an administrator clicked a button that said "Delete the main page". An administrator should not click such a button unless they actually wish to delete the main page. Testing is for test-wikis — this is not the sort of thing you mess about with on the main page of the 8th most used website in the world! Applying restrictions like this is not needed if we have an appropriate mechanism for distinguishing prospective administrators who know what they are doing with those who don't — or a culture of caution with regards to administrative functions. As for the use of the bot, we have a bot approvals group for a reason. I am aware that Betacommand is on it, and should know better than this. — Werdna talk 05:00, 4 February 2008 (UTC)

What need has there ever been to delete the main page? Or move it? Why are the tabs even there? The delete tab is currently hidden, but why not just remove the options altogether? Lara❤Love 05:31, 4 February 2008 (UTC)

bugzilla:9625 seems to indicate it won't ever happen. MBisanz ^talk 05:38, 4 February 2008 (UTC)

I think the main concern here is not with the end result (making the main page undeletable) but rather the means to get there (the apparent unilateral and unsupported move by two editors to dump 1200 junk edits into the main page) which represented an unintended exploit of a recently hacked-in safety feature and is rather an object lesson in how to use said safety feature to be disruptive. The two editors in question were NOT trying to be disruptive per se, but now it is plainly clear that one could use their means to be disruptive. I am not sure that the recommended blockings and/or desysoppings are justified or not, but this does seem like the wrong way to go about doing things. --Jayron32.talk.contribs 05:40, 4 February 2008 (UTC)

I think it is actually good that there is something highly visible for compromised admin accounts to delete or vandalize. A bit of good old delete-the-main-page-for-lulz will send people searching for stewards pretty fast. The latest case was desysopped in three minutes... – Sadalmelik (talk) 06:22, 4 February 2008 (UTC)

Bugzilla doesn't appear to have reached any sort of consensus. In fact, the page reads like more people agree that there's no need to have the option to delete the main page. Certain pages just don't need the option. Considering the delete and move tabs serve no useful purpose on these pages, and having them leaves the risk of abuse and server lockdown, why not just remove the option? Having the main page to the (currently) 9th most viewed website in the world be down for a few minutes is bad times when it's pointless and avoidable. Lara❤Love 06:49, 4 February 2008 (UTC)

Note that Bugzilla is not consensus-based. Brion Vibber, the lead developer and Wikimedia CTO, resolved the bug as "won't fix" and then marked it closed. But that bug was filed as a request to have the feature exist and be enabled by default in the core software. A Wikimedia-specific hack (as Tim has now implemented) is a different question. I still don't think it's a good idea, for the reason I gave on the bug. —Simetrical (talk • contribs) 14:55, 4 February 2008 (UTC)

Vital tasks

I seem to remember that BCB does some tasks like RFC bot's job, CfD (I don't know what that is), Spamreports, Image moving, and image renaming. What is our contingency plan for it being unable to do those tasks for a week? Yes, I know it does non-free image and orphan image work, but I'm not considering that vital given the existing huge backlogs at those areas. MBisanz ^talk 04:13, 4 February 2008 (UTC)

I don't know that we have a contingency plan for such things. The bot system is like the wild west. Everyone runs their own code and there is very little redundancy. I have supported for a long time the division of Betacommandbot's tasks into separate usernames instead of a single username - BetacommandBot 1, BetacommandBot 2, etc. — Carl (CBM · talk) 04:23, 4 February 2008 (UTC)

I have no idea how complex these tasks are, but the RFC one in particular seems pressing as part of the WP:DR process. How hard would it be for an uninvolved bot operator to code up a quick and dirty substitute? Or is there a by-hand process that explains how to replace the bot with actual editors. MBisanz ^talk 04:27, 4 February 2008 (UTC)

The RFC bot has somewhat complex behavior. I estimate that someone with a lot of experience with bot programming should be able to code a replacement in a day of dedicated work with no interruptions. But it would be better if the code was publicly available (I don't know whether it is). Even then, it might take a few hours for a new operator to get the code running on their machine. — Carl (CBM · talk) 04:34, 4 February 2008 (UTC)

Nope, BCB is (as of a discussion in 2007 [5]) proprietary code. MBisanz ^talk 04:36, 4 February 2008 (UTC)

Actually, RFC bot is/was operated by Messedrocker. BCbot seems to handle this task now, though. – Sadalmelik (talk) 04:52, 4 February 2008 (UTC)

Hell, just unblock the bot. That script isn't running anymore and Betacommand presumably knows not to run it again, so there's no reason to keep it blocked. --Rory096 05:12, 4 February 2008 (UTC)

Well Tim is a developer and I think is an employee of the foundation under Wikipedia:OFFICE#Who_does_office_actions so that might be a consideration in unblocking. MBisanz ^talk 05:17, 4 February 2008 (UTC)

Firstly, this is not an "office action". Do as you see fit. Rory096 says "Betacommand presumably knows not to run it again, so there's no reason to keep it blocked". I don't believe Betacommand has learnt any lessons for this, he's a stubborn kind of guy. He certainly didn't make any apologies when I was talking to him about it on IRC. If he does it again, I'll block him again. I've written the script to clean up his mess now, so it won't be so much trouble for me the second time. -- Tim Starling (talk) 05:42, 4 February 2008 (UTC)

In that case, we should make sure we get a promise out of him not to do it again (as he seems to have given below), and then consider actions to make sure anything like this doesn't happen again- not only exploiting some fix for some problem in an way that wasn't intended by those who implemented the fix, but also any unapproved, undiscussed bot functions, and to consider splitting BetacommandBot's functions so any one of the functions can be blocked without causing all the others to stop working. --Rory096 18:25, 4 February 2008 (UTC)

Rory096,Ive been working on seperating the accounts, its just not that simple. And Ive been pushed for other requests, see commons:User:BetacommandBot, for the recent image re-naming. As Ive said before this was a single request, and with no future plans on repeating. β^_command 18:34, 4 February 2008 (UTC)

Here's some back story behind RFC bot. I came up with the idea; Betacommand wrote it for me. It exists under the MIT license at this location. It is supposed to run on my toolserver account, but it takes up too much memory as the result of being somewhat broken. Therefore, it goes down once it reaches the toolserver's threshold of 1GB of RAM, and it is killed off by memory management software. While Betacommand is fixing it, he is using his account (that or the bot's) to continue operations; this is not unheard of, as Betacommand's account and his bot are used as testing vectors for new versions of the software.

The issue behind the bot is that objects in memory are forming too quickly without being given enough time to die off. I am going to slow the refresh rate from 5 minutes to 60 minutes to see if that will allow it to be ran on the toolserver without complication. MessedRocker (talk) (write these articles) 05:41, 4 February 2008 (UTC)

While I know BCB doesn't like to spread his code around, but honestly, I feel like more bots should take up some of his tasks either in case he doesn't wish to run it or a situation like this happens again. User:Zscout370 ^{(Return Fire)} 06:55, 4 February 2008 (UTC)

Note from developer

Tim Starling has sent an email to the wikitech-l list in which he strongly criticizes the practice of adding meaningless revisions to a page. At the very least, everyone should take away the message "don't do that". — Carl (CBM · talk) 05:15, 4 February 2008 (UTC)

I thought it was a good idea at the time ... BCBot hadn't quite added enough revisions to make the page undeletable. I have therefore given myself a suitable punishment. I now understand the slippery slope issue this could cause. Graham87 08:26, 4 February 2008 (UTC)

Ah... what? That's not the point I can't understand that matters. The point I can't understand that matters is why there is so much inertia. Could this not have been discussed publicly, on-wiki? Some lessons need to be drawn from this. And Tim noting that Betacommand seems unwilling or unable to do this, well, that does not inspire confidence. El_C 08:55, 4 February 2008 (UTC)

The Main Page cannot be deleted or moved on any Wikimedia wiki now. Graham87 10:06, 4 February 2008 (UTC)

I think it can still be moved. WODUP 10:16, 4 February 2008 (UTC)

It can't. I just tried on testwiki:Main Page. When you submit the form, you get a big red "You are not allowed to execute the action you have requested." WODUP 11:07, 4 February 2008 (UTC)

Why not let admins delete the main page? It is a canary in a coal mine, if an account is compromised or has gone berserk then they often delete the main page and lose their admin bit within 10 minutes. This is a good thing, it keeps the damage to a minimum. _(1 == 2)^Until 17:42, 4 February 2008 (UTC)

There are other pages they can delete inappropriately. MessedRocker (talk) (write these articles) 23:47, 4 February 2008 (UTC)

Messiness of bot records and COI

Just a note on an issue that's relevant to this case: For the past several days, I've been working on the records over at Wikipedia:Bots/Status. A brief look at through records yields the fact that they're in horrible condition, both inaccurate and out-of-date. There's also the outdated page, Wikipedia:Registered bots. And both of these records don't likely match up with the list of users with a bot flag.

This makes it practically impossible for admins to keep track of bot abuse.

Overall, there appears to be a COI with WP:BOT:

• The bot approvals group are required to be bot owners or programmers themselves, so naturally, they tend to be liberal about handing out bot privileges and may turn a blind eye to this sort of thing, or defend it (that's a speculation -- not an assumption -- of bad faith). Though they take into account the community's commentary, they still have a leading role.
• The bot approvals' group primarily maintains the bot records. Well, again, why should they care about maintaining good records? If they don't, then it's a lot easier to get away with this kind of thing.

I suspect that cleaning up the records and a thorough review of all users flagged or listed as bots would yield the fact that there's more bot abuse going on than people are aware of. ☯  Zenwhat (talk) 11:21, 4 February 2008 (UTC)

I don't want to sound like a beancounter, but maybe we could audit the active bots on the project. There should be a list of everyone with a bot flag somewhere, as well as the list of registered bots. We sent a talk page comment to every operator and bot on both lists: "Hey, we're updating records, please update the status of your bot here. If we don't hear from you in a week (or whatever), your bot will be listed as inactive." Whether inactive bots are de-flagged, blocked, or otherwise noted is up to consensus - but, I'd recommend that the operator checks in and updates status before resuming his/her bot's operations. This might also be a good chance to audit approved functions that current bots no longer do, so we can identify functions that other bots might pick up. Just tossing out an idea or two, UltraExactZZ ^Claims ~ Evidence 13:37, 4 February 2008 (UTC)

It seems to me that most complaints about bots are not in line with policy, and those that do are dealt with promptly. _(1 == 2)^Until 18:01, 4 February 2008 (UTC)

I agree that they are often dealt with quickly but, in some instances, it seems that the previous intervention does not preclude future issues from the same bot operator. Which is frustrating, at best. --Iamunknown 21:10, 4 February 2008 (UTC)

You can get a list of flagged bots from special pages. I'm sure if you offer to compare with the Status and Registered Bots pages, and the user contributions of those bots, and invite updates where needed, the bot community would welcome it. Rich Farmbrough, 14:37 6 February 2008 (GMT).

Just a note- the opposite of assuming good faith is not assuming bad faith, it's not assuming good faith. Speculation of bad faith is, therefore, violating AGF just as much as an assumption of bad faith would be. That said, you could have made the same points you just did while at the same time assuming that everyone's acting in good faith. --Rory096 18:12, 4 February 2008 (UTC)

Ah, yes, you're right, Rory. Sorry. Yes, I am probably just being paranoid. On the bright side, it did make for a neat addition to WP:List of cabals.

Still, at the very least, I'm right about the bad record-keeping, just wrong on the lack of WP:AGF. ☯  Zenwhat (talk) 04:17, 5 February 2008 (UTC)

The record keeping on approvals is pretty good last time I looked. But the status page relies on bot operators to update it, which if they go-away, die, just plain forget, whatever, is not always going to happen. Rich Farmbrough, 14:37 6 February 2008 (GMT).

Summary

Trying to summarise the above:

• (1) The Main Page got deleted accidentally after a "joke" discussion on IRC led someone to test the assertion that the Main Page couldn't be deleted. See Wikipedia:ANI#I deleted the Main Page.
• (2) Some other off-wiki discussions led to two users (an admin and a bot operator) to implement a workaround to prevent this in future, based on the recent "5000+ revisions = can't delete" fix. They did this by having a bot do lots of null edits to a subpage (User:East718/empty), and then the admin deleted the Main Page and merged the two histories (see also here), thus bring the edit count up above 5000 (well, in fact it seems they possibly fell short, and two other admins (User:Nakon and User:Graham87) made some manual null edits to bring the total above 5000; one later blocked himself for doing this, see here). The bot in question was BetacommandBot, but the edits are no longer visible because they got removed at some deep level by developers (because the page was over 5000 edits, only developers could do this - see point 5).
• (3) Tim Starling found out about this (see here and here) and blocked BetacommandBot for 1 week for "abuse of system resources", and has called for East718 to be desysopped.
• (4) A technical fix now means that the main pages on all WMF wikis can't be deleted or moved (regardless of how many edits they have). However, this is not a true fix. See here for a process that could be used for emergency deletion.
• (5) The 'junk' revisions have been removed from the Main Page history.
• (6) Several other discussions are ongoing, on mailing lists and on-wiki and (presumably) elsewhere. The wiki-tech mailing list disucssion has been linked, and there is this WP:AN thread. Two others are: Main Page talk page discussion, wiki-en mailing list, and a bugzilla discussion.

I think I have that about right. What needs to be sorted out moving forwards?

• (A) Consequences of BetacommandBot's block and how to handle the work it does.
• (B) Whether an arbitration case should be opened to handle the desysopping points.
• (C) Whether there has been abuse of a bot flag.

My views are that the block on BetacommandBot should remain, if only to force the community to face up to the consequences of a permanent block or future departure (for whatever reasons), and hence become less reliant on this (and other bots) in case things like this happen again. It is not acceptable to have bots be "unblockable" because they do "vital work". Splitting of bot functions seems well overdue. I'm not too fussed about East718's sysop flag - I don't think he will do anything like this again any time soon. Carcharoth (talk) 11:53, 4 February 2008 (UTC)

This isn't an RfC, but I concur with Carcharoth's summary of the issues. I would also add as a footnote that, whatever happens, I believe East718's actions were undertaken in a good faith attempt to improve the encyclopedia (and the security of the main page). The elements of concern in my mind are the lack of discussion beforehand, not necessarily on the intent itself. I also believe that that should be taken into account during any further proceedings, if and when. UltraExactZZ ^Claims ~ Evidence 13:45, 4 February 2008 (UTC)

Of benefit to the discussion of Task A, in re: Betacommandbot's block, the prefix list for all Requests for Bot Approval regarding Betacommandbot. I'm sure there's a more recent summary somewhere, but this might be a good place to start. The critical task I am aware of is tagging Disputed Fair Use Rationale images, but - given the active backlog on that category, adding more images to the backlog seems to be a low priority at this time. UltraExactZZ ^Claims ~ Evidence 13:59, 4 February 2008 (UTC)

Tim Starling made it clear above that he will accept unblocking Betacommandbot. Doing so would be the simplest way to move forward. Rewriting all the bot code would be nice, but ultimately it's probably not worth it. — Carl (CBM · talk) 14:27, 4 February 2008 (UTC)

Probably is the simplest solution, I agree. But whoever does it needs to make clear in the log summary that bots should not be merely unblocked to carry out other tasks. Becoming indispensable through one set of bot tasks is not a free license to carry out other (unapproved) bot tasks. This is a serious concern that has been brought up in the past and never satisfactorily addressed. It is effectively the same thing with editors (eg. Giano is effectively being asked to split his functions as article writer and Wikipedia namespace contributor). Humans can't always be asked to split between role accounts, but bots can and should be. Carcharoth (talk) 14:32, 4 February 2008 (UTC)

This is one reason I haven't unblocked the bot myself. — Carl (CBM · talk) 14:55, 4 February 2008 (UTC)

Just a note, my actions were a single one time request, that I did not think would cause as much trouble as it did. I did and do not have any plans on repeating the incident. Im am sorry that my good faithed attempt to protect the encyclopedia, caused as much drama as it did. I dont like drama. As for the source for my bots, I am willing to share it with people that I can trust. I wrote RfC bot and gladly handed that code out to a user that I know is responcible. I have also written code for other users and they have abused it, since then I only give it to people I can trust. within the next month or so I also plan on releasing the code for my image renaming script. (I need to finish testing and clean up the code). β^_command 15:28, 4 February 2008 (UTC)

Betacommand, do you want to tell Tim this and ask if he will unblock the bot? Or would you prefer the community to review the block of the bot? Carcharoth (talk) 15:40, 4 February 2008 (UTC)

I told tim prior him blocking that that was a single event that I was not planning on repeating. But he was fairly mad at the time. if someone wants to try and talk to tim for me Id be thankful, or if the community wants to review it. β^_command 15:56, 4 February 2008 (UTC)

Of course, all this would be an excellent argument for requiring that all bots on Wikipedia be entirely open-source, and that this be periodically verified by someone attempting to run the bot on a test wiki and making sure it actually works as advertised. Why Wikipedia has not yet agreed on this I'm not sure, except to the extent that it seems never to be able to agree on anything. (Yes, yes, anti-vandal bots' source code will be open, I'm sure that will be a great aid to the huge number of vandals who are also programmers and malicious enough to spend hours analyzing twisty heuristic-based source code. The idea of security through openness is that they'll be outnumbered by the group that's identical but willing to help out by sharing any exploits they find.) Without open-source bots, it seems to me Wikipedia is asking to have major bot contributors get annoyed with the project and leave, or just disappear for any reason, seriously inconveniencing everyone. Actually, this has happened in the past, if I'm not wrong. How is it that The Free Encyclopedia is relying so heavily on non-free software? If not for the bots and scripts that are permitted to be closed, you could come close to saying that the only proprietary software used in creating and serving the encyclopedia is routing software.

But I doubt this is the first time that argument has been made. —Simetrical (talk • contribs) 15:06, 4 February 2008 (UTC)

I can't agree more with Simetrical. Moreover, publishing the code, particularly during a BRFA, allows code review by other bot owners, and several improvements : DumZiBoT got significantly improved by Dispenser during its BRFA.

There is some pride in maintaining a bot, and I understand why some bot owners are reluctant to the idea of publishing their sources... However, keeping the source "secret" is not helping at all non-English wikis. From what I know, I have several examples of bots that could have been put in a great use at fr:, but that aren't, because the source is not available, and because English bot owners don't have time to spare for other projects... NicDumZ ~ 15:28, 4 February 2008 (UTC)

NicDumZ, have you asked these bot operators about this? β^_command 15:36, 4 February 2008 (UTC)

Positive. 'Want a concrete example ? I was thinking about SineBot : Slakr wants to develop a new tool, completely rewriting a PHP framework, and doesn't want to release SineBot's code until this is done. That's honorable, and I'm far from blaming him for this : Having such a new tool would certainly open a lot of new possibilities. However, this is taking a significant time : SineBot is running here since september... During that time, the only possibility that we, on fr, have to run such a bot, is to rewrite its code, and I personally consider this as a waste of time.

I mean. I am a bot owner, and I code everyday as a living : I don't want to blame any coder for their efforts, I also by myself have a lot of troubles when someone comes, looks at my hard work, and tells me : "This part of your code is useless, delete it", or when some random guy comes and add dubious functionalities to my script. But a strong fact is that several developers working together usually develop better tools than one developer alone, and eventually, I always consider these interventions as useful and helping. Just consider how efficient is the pywikipedia community ! At some extent, that's the way wikipedia works : others sticking their noses in your articles, in your work; but eventually, "your" articles are most of time far better with the help of others... Sharing the code is an immediate way to improve it. NicDumZ ~ 16:06, 4 February 2008 (UTC)

NicDumZ, Ill see what I can do about sinebot. I think your approach to the operator could use some work. β^_command 16:16, 4 February 2008 (UTC)

Thanks for your help, even if that wasn't why I wrote this. I'm just saying that every coders have reasons not to publish their code (code cleaning, refactoring, new project, new functionalities, not enough time to maintain it/document it, and so on... ). Sometimes I just think that making a little effort to clean our botcodes to release it every week or so on one of our subpages could help *a lot* the community. NicDumZ ~ 16:52, 4 February 2008 (UTC)

Simetrical, your comment about vandals getting a big help from the source to anti-vandal bots is not really true. ClueBot has had its source open since its inception, and I haven't seen any vandals who have been trying to get around ClueBot.  :) -- Cobi^(t|c|b) 00:50, 5 February 2008 (UTC)

Precisely as I said. I don't believe an open-source anti-vandal bot is any problem at all. You can sometimes make an argument for security through obscurity when heuristics are used, because a lot of correlations can be avoided with some care if the subject knows they're being looked at, but this isn't such a case. The overwhelming majority of vandals aren't going to trawl through source code or even know it exists, and the tiny number who might are too clever and careful to be tricked by a heuristic-based bot anyway. They'd probably spend their time getting through a rogue admin account or something, if they were interested in their vandalism not being immediately reverted.

So I don't see any reason for continuing the practice of permitting closed-source bots to operate. If it were up to me, an iron-clad criterion for bot operation would be publication on the web of the live source files actually being executed by the bot as it runs, and this would be enforced retroactively after a suitable grace period. I don't, unfortunately, have the time or inclination to immerse myself in Wikipedia policy-mongering hell sufficiently to actually get anything resembling this agreed to, but if anyone else does, you certainly have my support. —Simetrical (talk • contribs) 02:33, 5 February 2008 (UTC)

Bot Tasks

Somewhere floating around in the back of my mind, I have a recollection that the Bot Approval Group has to approve new tasks on Betacommand Bot. Or maybe I'm mixing that up with something else... can someone confirm/deny that? I think it has bearing on this: if such a requirement exists, and it was ignored, then we have a problem. If not, and I'm confused (which is more and more likely every day) then someone should feel free to strike through this section. ;-) The link I'm concerned with is Wikipedia:Bots/Requests for approval/BetacommandBot, but I may be misinterpreting it. - Philippe | Talk 16:42, 4 February 2008 (UTC)

that was related to the mess back early last year. approval for those tasks were rejected and I had to re-file. bots are usualy exempt from needing approval for user subpage editing. β^_command 16:48, 4 February 2008 (UTC)

OK, I didn't know they were exempt from needing approval for user subpage editing. Thanks for explaining that. Cheers. - Philippe | Talk 16:56, 4 February 2008 (UTC)

I am not aware of any such exception. Can you point me to the location of this guideline? Happy‑melon 21:24, 4 February 2008 (UTC)

It's not part of the guideline, as far as I'm aware, but neither is every single task completed or to be completed in Wikipedia. It has been allowed in the past at reasonable rates, mostly for statistics and record-keeping, but Betacommand's actions were neither at reasonable rates nor, uh, sane (although in good faith). Gracenotes^T § 01:40, 5 February 2008 (UTC)

Betacommand, were you aware that the edits, even though made to a user subpage, were going to be moved by East718 to the page history of Main Page? If so, then I'm afraid your excuse is the flimsiest and most disingenuous one I've ever seen. It is absolutely clear that any bot would need approval for a task to make edits to the main page. I see that WP:BAG are having elections at the moment, but this is something they will need to discuss as a matter of urgency. Carcharoth (talk) 23:44, 4 February 2008 (UTC)

I've thought about this more, and I'm disappointed in Betacommand's reasoning here. Given that the task was run in userspace with the sole purpose of then affecting the main space, I think this reasoning is a cop-out. I believe the task should have been cleared through BAG and anything else is unacceptable. Betacommand knew that this was to be used on mainspace, and found a way to work around BAG restrictions. - Philippe | Talk 17:11, 5 February 2008 (UTC)

And from a BAG member at that. Soxred93 | talk count bot 17:12, 5 February 2008 (UTC)

Eh? I'm not a BAG member - and neither is Betacommand. Or do I misunderstand? - Philippe | Talk 04:55, 6 February 2008 (UTC)

Betacommand has been reinstated as a BAG member since December Wikipedia:Bots/Approvals group MBisanz ^talk 05:01, 6 February 2008 (UTC)

Good lord, I gotta re-watchlist that page. See what I miss? - Philippe | Talk 05:13, 6 February 2008 (UTC)

The next step

I believe Carcharoth has laid out the problem here perfectly, but I'm concerned about where we're going now. East and Betacommand have yet to comment on the issue, and until then I believe an ArbCom case with what we have is overkill. I'd prefer to see an RFC on user conduct initiated (preferably using Carcharoth's analysis as an introduction), so both users can share their views and rationales. There is unanimous consensus among the community that their actions were wrong, let's hear their responses and let the community decide what to do next. - Mtmelendez ^(Talk) 14:46, 4 February 2008 (UTC)

• Lets take a step back for a minute and consider some points. East was quite clearly attempting to help the encyclopedia with what he did, as did Betacommand. I'm sure they didn't expect the drama that arose because of it. Whilst Tim Starling is a key developer, and don't think we need to take his advice without consideration of the facts - let's not start the desysopping talk - there was no malicious intent here, and by stopping the main page getting deleted, they were doing an honourable service for the project (although they went about it the wrong way). All seems sorted now, the revisions are gone - let's just move on, we really don't need this escalating further than it already has. I fail to see how an RfC would accomplish anything. Ryan Postlethwaite 14:48, 4 February 2008 (UTC)
  • While I assume (and don't doubt) that Betacommand and East were acting in good faith, that doesn't mean there certainly shouldn't be any action taken. Desysopping could be necessary if it's likely something like this can happen again in the future, even if this happened in good faith. It should be discussed whether the involved users have learned from their mistakes and will be sure to have a consensus before they do controversial things like this in the future, and we should seriously discuss what to do with BetacommandBot, so Wikipedia doesn't rely so much on a single bot to keep things running smoothly. There are certainly things to be done, or at least discussed, even though the users were acting in good faith. --Rory096 18:34, 4 February 2008 (UTC)
• I agree with Mtmelendez. I don't have time to initiate an RfC, but if anyone does, please feel free to copy paste my summary above, and add anything else that is needed. Possibly wait until the end of the day for more responses here, and pray that no-one files a presumptuous request for arbitration. Let's see if we can get things done productively, and address issues while minimising drama (Ryan, just letting it all go doesn't address the valid issues and concerns - this is precicely the sort of things that user conduct RfCs are for - pointing out lapses of judgment and getting community opinion on said lapses of judgment - doing this sort of thing without any on-wiki discussion, as far as I can tell, was unacceptable). Carcharoth (talk) 14:52, 4 February 2008 (UTC)
  • I'm not suggesting that we let this go per se, many people have weighed in now, and I expect East realises his mistake - I just think it would be wrong to start an RfC over someone who made a good faith (yet extremely misguided) effort to help the encyclopedia. My advice would be for someone to go to Easts talk page, say that he messed up, don't let it happen again but that it is the end of it now. Ryan Postlethwaite 14:57, 4 February 2008 (UTC)
    • Waiting until East and Betacommand reply here is probably a good idea (but see the note below about East and the notice he has left). For the record, here is what Tim Starling said in that mailing list post:

      "East718 and Betacommand got together, and decided between themselves, apparently without review or approval by any other party, that they would add 1200 junk revisions to the main page. Betacommand edited User:East718/empty the requisite number of times, and then East718 deleted the main page, moved his subpage to Main Page, and then undeleted it to merge the histories."

      Not the best judgment call there by either of those users. As far as I can tell, neither Betacommand or East718 have responded to the messages left for them so far. East was notified as early as 01:32, 4 February 2008. That comment by MZMcBride specifically mentions IRC, and also mentions Nakon (presumably User:Nakon, recently renamed). Was there an IRC discussion and decision to do all this? Also, East718 went silent for two hours after carrying out the deletions, and then made this edit to leave the message: "Something's come up and I won't be editing for a while. Feel free to leave messages." Is it too presumptuous of me to ask what "came up" to prompt that notice? Carcharoth (talk) 15:15, 4 February 2008 (UTC)
  • I don't see the need for an RFC just to cover a single action; just pointing out a mistake is enough feedback for an isolated incident. — Carl (CBM · talk) 14:58, 4 February 2008 (UTC)
  • Is there a pattern of bad judgement, or is this an occasional mistake? I don't see evidence of a pattern. Jehochman ^Talk 15:01, 4 February 2008 (UTC)
    • The lack of response worries me. Sometimes there is an attitude that hunkering down for a while is the best way to avoid drama. Sure, things do happen in real life to prompt people to stop editing for a while, but the impression is that sometimes the timing of such interludes is convenient. Until a fuller explanation emerges (and at that point I will be the first to apologise), the impression given is not good. In East's case, probably no pattern. Betacommand's case is more problematic. These were good-faith attempts to protect the encyclopedia, but there needs to be an open admission that they quite simply got this wrong and realise they shouldn't do this in future. Carcharoth (talk) 15:21, 4 February 2008 (UTC)

Let's review the facts. User and Admin collude to do something that they Probably Shouldn't Do. UserBot does garbage edits to a subpage, and Admin moves it to the Main Page, thereby misusing the admin tools. The admin tools were granted on an implicit agreement to not misuse them. Solution: Sternly warn said misuser of tools, and take away UserBot's access since this is the 500th time it's done something that makes the entire community mad at it. Luigi30 (Taλk) 15:03, 4 February 2008 (UTC)

Anyone who deals with fair use images, even if it is a human, will have people knocking on their door on a constant basis, sometimes in an angry state of mind. This is the first time I noticed the bot do anything very wierd such as this. User:Zscout370 ^{(Return Fire)} 15:44, 4 February 2008 (UTC)

• I've asked User:Graham87 and User:Nakon to comment here, as they made the null edits to the Main Page (still visible in the history). What I want to know is how they were aware of all this and what communications were made to take these actions? I've also asked User:MZMcBride, as he seemed to become aware very early on (he posted to East's talk page at 01:32, and the wiki-tech mailing list post by Gurch was at 01:39). I'll ask Gurch as well how he became aware of all this. Carcharoth (talk) 15:50, 4 February 2008 (UTC)

  As one of the admins who made a few null edits to the main page, I've been asked to comment here. I have Talk:Main page on my watchlist (and thus Main page) and noticed this edit. I assumed that since the page was deleted earlier, someone was trying to ensure that it wouldn't happen again. I found that the number of edits was lower than the bigdelete threshold and I made a few more to bump up the number. I apologize for any problems I may have caused. Nakon 16:34, 4 February 2008 (UTC)

  The Main Page was deleted. It's the sort of thing that is fairly easy to become aware of – Gurch 17:14, 4 February 2008 (UTC)

  As far as I'm aware, deletions don't show up in watchlists. Let me be frank. Was there an extended discussion off-wiki (be it mailing lists or IRC) before it was raised on-wiki? If so, why? Carcharoth (talk) 17:18, 4 February 2008 (UTC)

  The developers aren't in the habit of hanging around Talk:Main Page or the administrators' noticeboard. However, they do read the mailing list. Since the mass editing was an attempt to use the deletion size limit in a way the developers had presumably not foreseen when they added the limit, I thought they should know about it – Gurch 17:30, 4 February 2008 (UTC)

  One would think that recent Sandbox deletion would hint at the neccessity to discuss all these "bright" unusual ideas first ∴ AlexSm 17:26, 4 February 2008 (UTC)

  Gurch has explained in more detail on my talk page. I'm satisfied that the dummy edits are what attracted some people to this - but I'm still unsure how those making the null edits became aware of the deletion and page history merge and the "let's get it above 5000 revisions" plan. I just want to be crystal clear who contributed to these off-wiki discussions before during and after this incident. Ultimately, as Gurch says, it comes back to the judgment of the users in question, but I'm still (like others) worried that people just don't seem to be getting the message. Off-wiki discussions, or those with a limited number of people participating, are inherently risky due to lack of review and lack of transparency. Please, please, please can those using IRC, e-mails or whatever, think next time something like this comes up? Carcharoth (talk) 17:49, 4 February 2008 (UTC)

  You're still trying to put blame on the mode of discussion. Had the handful of users involved discussed this on each other's talk pages before doing it, it still wouldn't have been any better. Conversely, if they had asked for an opinion in the #wikipedia-en channel before doing it, they would have been told not to, by me. And needless to say, if they had asked on the Wikitech-l mailing list, they would have been met with the same threat of desysopping that they have now been met with anyway. Please stop labelling all "off-wiki" discussion as bad and "on-wiki" discussion as good, because that simply isn't the case – Gurch 18:19, 4 February 2008 (UTC)

  The problem is that those who give advice in closed areas don't get shown up for the purveyors of bad advice that they are. In the Hoffman arbitration case, Moreschi commented in an ANI thread that Hoffman was "obviously a sockpuppet" - that bad advice was in the open for all to see. If that had taken place on IRC, Cuerden would have not had the ANI thread to point to and say "look, I got support for my block", but equally, the giver of bad advice would have remained in the dark. Similarly, here, we just don't know who else gave bad advice. Carcharoth (talk) 18:28, 4 February 2008 (UTC)

• Just a small note: please do not call these null edits, this is very confusing to those familiar with already established terms, see m:Help:Dummy edit: null edits are not even recorded in the page history ∴ AlexSm 17:26, 4 February 2008 (UTC)

Wow. Simply incredible that so many people can make so many poor choices in the span of about 12 hours. Tim laid it all out pretty well. I wasn't able to be on IRC while this was going on, thus my post to East's talk page. Beta did a task for East; Beta's really not all too much to blame -- almost any bot user could've done the exact same thing. Alex is correct in the post above this one -- null edits do not add revisions to the database. What Beta made are called trivial or "dummy" edits. Going forward: (1) All bot discussion should go to WP:BON or WP:BOTREQ. This page is inappropriate for bot discussion; (2) I'm of the opinion that ArbCom and RfC are both unnecessary here; I propose that East (and perhaps Maxim and Ryan) stand for a new RfA in a week (when everyone's called down a bit). Meanwhile, I think that he / they should be barred from using any admin tools. Kudos go to Krimpet and Animum who attempted to reduce the possibility of a Main Page deletion without fucking everything up and causing all this drama. Kudos go to all of the admins who didn't unblock BetacommandBot; if there's one thing this community doesn't need, it's another sysadmin who's particularly pissed with us. --MZMcBride (talk) 17:54, 4 February 2008 (UTC)

Addendum: I mentioned User:Nakon and User:Betacommand in my post on East's talk page because they were the ones I could see in the Main Page history and they were the ones that I knew for sure are regularly in the admins channel. --MZMcBride (talk) 17:57, 4 February 2008 (UTC)

I think MZMcBride is referring to this by Krimpet. Animum undeleted the Main Page three minutes after Maxim deleted it. Presumably Maxim had noticed his mistake and was trying to undelete it, but Animum got there first. Carcharoth (talk) 18:12, 4 February 2008 (UTC)

When referencing Krimpet and Animum, I was referring to their edits to the global CSS files. On a side note, I completely agree with Gurch that it makes no difference whether this conversation between East and Beta occurred on-wiki or off-wiki. Had I been able to get on IRC last night, I would have immediately told them what a stupid and ill-thought-through idea it was to add null edits to the Main Page. --MZMcBride (talk) 19:10, 4 February 2008 (UTC)

They didn't talk about it on IRC either. Well, they might have done in a private message, but not in any of the channels – Gurch 20:14, 4 February 2008 (UTC)

I for one just checked out the history after the Main Page deletion to see if anything happened in the hour or so since its undeletion. I found BCBot's edits and decided to try the final push to 5,000 edits. Graham87 23:06, 4 February 2008 (UTC)

And that is one of the most ironic things about the whole incident. It seems that even after doing all this, they miscounted or something, and the bot hadn't done enough edits! I mean, really, getting the number of edits right isn't that difficult is it? Carcharoth (talk) 23:21, 4 February 2008 (UTC)

I may have skimmed through all of the comments above, but I don't see where this has been addressed. Where was Betacommandbot authorized to do these edits? The bot should have gone through bot review to do any other tasks than what it has already been approved for. If Betacommand is adding new features to his bot without approval, he should stop, now. Corvus cornix_talk 23:17, 4 February 2008 (UTC)

In the subsection titled "Bot tasks" there is something about how editing user subpages is an exception. Of course, if Betacommand knew that these edits were going to be moved to the page history of the Main Page (as it seems he did), then this whole excuse disintegrates like a mass of soggy tissue paper. Carcharoth (talk) 23:24, 4 February 2008 (UTC)

I have made an extended comment on this situation in a user subpage. — Werdna talk 09:20, 5 February 2008 (UTC)

What next?

It has been almost 24 hours since the deletion and page history merges by East718 took place. The following is an attempt to refocus the discussion again and see if anything remains to be done before moving on.

• (1) BetacommandBot remains blocked (it is a week-long block that started at some point on 4 February 2008). Betacommand has asked above if anyone will speak to Tim Starling for him, and failing that, would be happy for the community to review the block. At the moment, it seems no-one is willing to unblock, though maybe someone will put forward a good reason for unblocking. One question is whether the block is causing problems - are Betacommandbot's other functions being carried out? See also the questions below that I've posted to the WP:BAG talk page.
• (2) East718's actions - it seems that nothing much more can be done until East718 returns and responds to the concerns raised above and elsewhere. What is the normal procedure in situations like this? Leave a note on their talk page summarising what has been discussed and pointing them to somewhere (back here?) where they can respond to the concerns? How long should should someone be given to respond?
• (3) Bot actions and issues - as was pointed out above, this needs to be discussed at the various bot pages, such as WP:BAG. Betacommand is a member of BAG (the bot approvals group). I've posted the following questions over there: (a) Did this bot action need approval? (b) What are the views of WP:BAG on the block of User:BetacommandBot? (c) How can the issue of too many functions being tied up in one bot be addressed?
• (4) The Bugzilla discussion about Main Page deletion hacks, patches and whatnot. Anyone want to summarise that? Not sure what more can be done with that other than to note the outcome somewhere if it is relevant. Most of the technical stuff seems to have been sorted or is being discussed elsewhere.
• (5) Other issues - auditing of bots (cleaning up existing records) and dealing with BetacommandBot's tasks, if needed. Again, all bot stuff. Can these issues be safely turfed over to WP:BAG and similar venues to deal with?

In summary, I think the remaining admin issues are (and we should really concentrate on these issues and not be distracted by the other ones):

• (I) The block of BetacommandBot (let it run out, or lift it early based on what Betacommand has said?).
• (II) The actions of East718 (were the following acceptable: the decision to do this, the discussion of the actions, the actions themselves, and the response afterwards). We can review the first three. The last one (the response afterwards) is not fully known yet, and will need East718 to make an on-wiki response.

Apologies if anyone feels I'm prolonging this thread. I think stuff is slowly being directed off to the correct venues to be discussed, and hopefully the above will help focus on stuff relevant to this noticeboard. Carcharoth (talk) 00:09, 5 February 2008 (UTC)

Response

Hey, everyone. Leave for a day and look what gets stirred up. :O First off, I'd like to apologize for any disruption caused; I hope you all recognize that this was one of those things meant with the best intentions that went completely off the rails. It was a good-faith effort to lock down the main page which was messed with twice just yesterday, once by a compromised account and another time as a joke gone wrong. It was very bad form of me to just unilaterally do this, but I've always been of a mind to just get things done. In the end, it's been rather harmless as the history is now cleaned up and a permanent solution is in place. I'm not going to put up a fight if anybody wishes me to be sanctioned in some manner, but it won't be useful as I'm disappearing for a long while due to unrelated reasons.

BCBot should probably be unblocked, since Betacommand wasn't aware of what I was doing - I made an open-ended request for a bot account in #-admins and he just happened to be the first person to respond. It was my mistake for passing him a bot that made a bunch of garbage edits at reckless speed without informing him of its nature, although I suspect he's learned a lesson to not run unvetted code on his account.

One last bit of housekeeping: it's a very poorly kept secret that I run a bunch of unapproved adminbots that perform repetitive tasks so other admins can worry about more pressing matters. The most important of these is a bot that hunts down and reverts/deletes XRumer (aka /w or index.php) spam in real time, and blocks spam-only accounts and IPs if the spambot is running from a zombie proxy. Around 30 or so spambots are stopped daily with 10% or so being open proxies or compromised computers. Obviously I've taken it down already but can pass it on to any other admin willing to run it, just email me. east.718 at 03:15, February 5, 2008

Hi East. Thanks for the apology and for explaining what happened. Hopefully you will be around for long enough to reply to this, but if not, then I guess it will have to wait until you get back.

• "It was very bad form of me to just unilaterally do this" - can we have assurances that you won't act unilaterally like this again?
• "I've always been of a mind to just get things done" - in future, will you discuss things like this before doing them? There is boldness and then there is recklessness. No harm done this time, but what about next time?
• "but it won't be useful as I'm disappearing for a long while" - the question is whether you will repeat the misjudgments made this time round, so this thread will be useful in determining that - sure, it can wait until you get back, but the attitude that going away for a long time means that the possibility of sanctions (even if it is only a thread like this with lots of criticism of your judgment) should be discounted, is, well, rather strange.

Then there are the three points I raised above:

• (1) discussion of the actions - did you in fact discuss this with anyone? I thought you had discussed it with Betacommand, but it seems now that you didn't.
• (2) the actions themselves - I think it is clear now what you did - can we have assurances that you won't add dummy edits like this in future, or do page merges like this in future, without discussing it first? Especially given that Tim Starling has said he will block anyone who does this?
• (3) the response afterwards - if you do do something unilaterally in future (boldness is sometimes good), can we have reassurances that you will make every attempt to be around in the aftermath? The notice you put up two hours later saying that something had come up is fair enough - but can you tell us what happened in those two hours? Did you see the talk page messages people had left you? Did you get lots of people asking you what had happened, and did you respond to them? Off-wiki response are all very well, but the on-wiki records just shows silence, a notice after two hours, and then this response. If you are not going to be around to deal with the follow-up to something, discuss (on-wiki) with others and maybe let someone deal with it - there was no urgency here.

I'll let others respond to the other points, and I'll respond to the Betacommand bit in the section Ral315 started. I appreciate the image work you do, East, so I hope things do work out. If I'm happy with the above points, I won't be taking things any further, and I would hope no-one else would either. Have a nice wikibreak! Carcharoth (talk) 08:27, 5 February 2008 (UTC)

I have a lot of tolerance for good faith screwups, for the simple reason that, if they're truly good faith, they're unlikely to be repeated (unless the user in question is demonstrably incorrigible, which East isn't). I say we leave things were they now sit. Sarcasticidealist (talk) 08:32, 5 February 2008 (UTC)

How difficult is it for East718 (when he gets back) to take the time to answer the above? Ideally, East718 won't be the only one to learn how not to do things. Others watching this will learn what not to do in more general terms - ie. discuss first, really think about the consequences of the actions, and stay around to deal with them. Carcharoth (talk) 08:36, 5 February 2008 (UTC)

Sorry, what I meant was that I was satisfied with the response, not that further questions to East were inappropriate. Sarcasticidealist (talk) 08:37, 5 February 2008 (UTC)

Unblocking BetacommandBot?

Given that Tim's willing to let the community handle whether Betacommand should unblock, I'm wondering whether BetacommandBot should be unblocked. Clearly this was a lapse of judgment. And this isn't the first lapse in judgment that Betacommand's made. But I think that, with no permanent harm done, and Betacommand promising not to do this again, there's no reason to make him sit the block out.

This unblock would be with the understanding, of course, that he not run bots like these without prior approval from the BAG. I'm personally willing to do this, so long as it isn't controversial; any thoughts? Ral315 (talk) 04:43, 5 February 2008 (UTC)

Bot owners are responsible for the actions of their bots. The edits were a waste of system resources and should not have been made. I hold Tim in very high esteem, and I imagine many other admins do as well. It is pretty rare that a sysadmin steps in to a situation like this and takes action; it's even more rare that revisions are then deleted from the database. Things like that generally indicate quite an error on someone's part. A week is not a very long time, and I hope this block gives people time to appreciate the work that BetacommandBot does for the community and perhaps other bot owners can write similar bots in the event that Beta someday decides to no longer be as active as he is. BetacommandBot can be out of commission for a week -- we'll live. --MZMcBride (talk) 06:58, 5 February 2008 (UTC)

I agree with MZMcBride. Let's see how things pan out for a week. I understand the argument that East718 was also responsible, but before any unblocking, Betacommand needs to state clearly and unequivocally what lessons he has learned. For one thing, I wouldn't be happy to see BetacommandBot unblocked until Betacommand explains the following from East718 (see above): "passing him a bot that made a bunch of garbage edits at reckless speed without informing him of its nature, although I suspect he's learned a lesson to not run unvetted code on his account" Beta, did you really run unvetted code on that account? How much did you know of what East718 was trying to do? Carcharoth (talk) 08:33, 5 February 2008 (UTC)

Just above, you state (and, I agree), that in this instance, no real harm was done. I find it REALLY hard to believe that Betacommand didn't know at least what the bot did, and, I can reasonably understand why it might have been thought to be a good idea (a good enough idea, to protect the main page from deletion, that Tim tweaked the main page to be not be deletable himself!). They were just trying to protect the project. I'm not sure, what the point of punishing keeping BCBot blocked for a week at this time, exactly is. SQL^{Query me!} 13:25, 5 February 2008 (UTC)

I'll repeat myself: "before any unblocking, Betacommand needs to state clearly and unequivocally what lessons he has learned" I normally disagree with this "make them show they know what they have done wrong" approach, but then that approach is too often applied to new editors who can't be expected to know all the site standards. Betacommand should know by now what is and isn't acceptable - what will and won't create drama. Plus the question that East718's response has thrown up (the "unvetted code" bit) - either Betacommand knew what the bot request was for (and thus shares responsibility for thinking it should happen without discussion) or he ran the bot request without really thinking about what it was for - not suitable behaviour for someone who is now back on the bot approvals group and is trusted with helping to approve or deny other bot requests. I would prefer that Betacommand actually steps up to the plate and addresses these issues (and gets his bot unblocked early), rather than him just staying quiet and waiting for the block to expire. I agree that this block is not the right point to talk about splitting up BCBot - but that discussion shouldn't be deferred much longer (BC needs to lay out a clear timetable by when he intends to get this done - and there needs to be checks to prevent over-reliance on single bots, or any bots). Carcharoth (talk) 13:59, 5 February 2008 (UTC)

Until BCB is split up into multiple bots (a completely reasonable request that has been made multiple times in the past) I see no reason it should be unblocked. Opening the source would be good too. Haukur (talk) 10:24, 5 February 2008 (UTC)

While that's a fair point, and one perhaps Betacommand should do, I don't think it's fair to hold a block over Betacommand's head over something as trivial as that. Nevertheless, I see the points made above as well. Ral315 (talk) 10:46, 5 February 2008 (UTC)

Agreed, I strongly oppose holding BCBot's block over his head, to get what you want. If you believe that every bot should be forced to have separate accounts (and, be open source), per task, then, please get consensus and modify the bot policy. SQL^{Query me!} 13:08, 5 February 2008 (UTC)

Agreed. BetacommandBot might emphasize some necessary changes to the Bot policy, but BetacommandBot's case must be considered as an independant thing, no matter what. NicDumZ ~ 16:03, 5 February 2008 (UTC)

I haven't suggested any broad policy changes - I just want Betacommand to split his bot. With any reasonable code design this would be an easy task. Haukur (talk) 15:27, 5 February 2008 (UTC)

I was saying, that, if you want bot owners to be required to run separate tasks, under separate accounts, this isn't the time, nor the place. Until then, there really isn't any requirement, for BCBot to do so. SQL^{Query me!} 16:06, 5 February 2008 (UTC)

I am not in favour of an early unblocking; Betacommand has not been blocked, his bot has. If he cannot run his bot within the agreed limits, it should stay blocked. An early unblocking merely confirms to Betacommand that he can pretty much do what he likes with his bot. Neıl ☎ 11:02, 5 February 2008 (UTC)

BetacommandBot has a long history of problems, especially bugs and unauthorised actions. Do not unblock it. I suggest we ban it. —Preceding unsigned comment added by 59.189.64.44 (talk) 13:46, 5 February 2008 (UTC)

Absolutely do not unblock BetacommandBot. You don't keep giving the car keys to the kid who keeps driving off of Deadman's Curve. This whole latest hullabaloo happened precisely because of a failure to stop and think. Had this issue been calmly deliberated on-wiki, it would have quickly become obvious that it was a hideously stupid idea from the get-go. Instead, it was dashed off as a "hey, let's try this!" idea on IRC, and quickly implemented, to the project's detriment. I think that making BetacommandBot cool its heels for a week (or more) is clearly necessary in order to prevent this sort of slapdash irresponsibility in the future. Nandesuka (talk) 15:06, 5 February 2008 (UTC)

Nandesuka puts it better then I did. Apropos of very little, this is the 36th block to be issued to BetacommandBot, for various reasons including misuse, continuing to work outside trial periods, general bugginess, and so on. Most of them get undone early because Betacommand always promises he's learnt his lesson and won't screw up again. How many more chances will this bot get? Surely we could find a user or a group of users more suitable, civil, and with the capacity to learn from their mistakes to run bots that cover the tasks BetacommandBot does, split by task and with open code to allow for collaborative improvement. Neıl ☎ 15:16, 5 February 2008 (UTC)

Neıl, you fail to note that 90% of those blocks are groundless and quickly reverted. I also take offense to your attacks against me. Ive personly done over 60,000 contributions to this project, and BCBot has ~800,000 edits. yes there will be mistakes, errors and bugs with that number of edits. In this case I did one act to protect the encyclopedia people completely fail to assume good faith with me. I now know why so many good users are leaving the project in droves, on average we loose an admin ever two days. we seem to want to hang every good user for the slightest mistake. it seems that now users like creating drama, and banning users. it almost seems that the current consensus is to have be banned and sent to hell. β^_command 15:45, 5 February 2008 (UTC)

As Haukurth says, "90% were groundless" was patently untrue. The fact many were quickly reverted is because you have already been extended a lot of good faith and many admins have been willing to unblock the bot because you promised it would work from now on. But it's clearly not working, as there's at least 16 blocks that are valid (even if they were undone before expiration). "People completely fail to assume good faith with you" - does this mean you don't even understand why what you did was wrong? Neıl ☎ 16:16, 5 February 2008 (UTC)

lucasbfr's count below has it that 50% of the blocks are valid. That's still a lot of blocks. A bigger problem is that you react with this same indignant attitude every time you've made a mistake and been criticized for it. Haukur (talk) 16:00, 5 February 2008 (UTC)

Haukur, Ive admited that I fucked up, I was acting in what I though was the betterment of wikipedia. Im sorry if I get a little irratated when people miss-quote facts and attack me for making good faith efforts to help. it seems that people ignore all the good that I do and they just attack me and call for my head on platter. or that I get banned β^_command 16:19, 5 February 2008 (UTC)

Indeed, from a COOP perspective, I'm a little worried on how reliant elements of wikipedia seems to on the BCbot performing rountine tasks and how little redundancy we have. As a matter of urgency, we need to work towards developing bots that are under the control of the community (by that, I mean the community has access to the code via some mechanism - so that if the owner leaves or falls under a bus we have a fallback position). I agree entirely with Neil's other comments. Is there a working group or other "body" who could take a lead here? --Fredrick day (talk) 15:22, 5 February 2008 (UTC)

Theoretically, WP:BAG. But that nearly got shut down by the community at one point. One idea would be to get developers more involved. See User:Werdna/Comments on main page deletion:

"It is my suggestion that the development and systems administration team should play an integral role in the administration of a successful bot system. While the community of the English Wikipedia should be tasked with determining whether the purpose of a bot is sound, it is the general Wikimedia technical community which must evaluate a proposed bot, feature, or other technically-sensitive change for its impact upon performance, and on whether it is better achieved with, for instance, extensions and modifications to MediaWiki." - User:Werdna

This would also help build bridges between the community and developers. I really think this is a great idea, and should be followed up as far as possible. Carcharoth (talk) 15:34, 5 February 2008 (UTC)

Carcharoth, it's a good idea in theory, but it's bad in practice, simply because the developer squad is already spread thin and overworked. If they were interested, yes, that would be great, but I really doubt that is the case. Titoxd^{(?!? - cool stuff)} 05:00, 6 February 2008 (UTC)

Here's a suggestion. Why can't Betacommand spend the week-long block tidying up his code, running it on test wikis, and preparing for this splitting of tasks and opening up of the code that he seems to be on the verge of doing? We can't force him to do this, but we can politely suggest that he do this. Carcharoth (talk) 15:34, 5 February 2008 (UTC)

It would be lovely if BC were prepared to do that. I'm pretty sure he has been asked politely to consider this ona number of occasions and refused to do so. Neıl ☎ 16:12, 5 February 2008 (UTC)

Neil, last time I counted (out of 27 blocks) I counted 11 valid blocks and 10 out of process. I guess that 50% of blocks on the bot are still rubbish. -- lucasbfr ^talk 15:45, 5 February 2008 (UTC)

16 valid blocks (plus at least 4 valid ones on User:Betacommand for malfunctioning automated scripts, running the bot on the main account, bot-spamming RFCU, automated deletion etc) makes 20. Even with a number of blocks being "out of process", 20 valid blocks suggetss there's something to be resolved. Splitting the bot tasks would be an excellent start. Neıl ☎ 16:12, 5 February 2008 (UTC)

Keep blocked until functions are atomized. Folks keep saying that we shouldn't keep it blocked because no permanent harm was done. That's an insane standard that we would never apply to an editor. Cool Hand Luke 16:04, 5 February 2008 (UTC)

I hate this, but I believe the bot should stay blocked. I've tried to be open minded, but I've seen, over the past year, too many mistakes coming from Betacommand and I simply don't trust his judgment with the Bot, unfortunately. I tried - really tried - to wipe the slate clean after the fiasco of a year ago, and I had almost rebuild my trust in Betacommand and then he went and did this. I'm sorry, in my opinion, the bot should remain blocked for the full duration. I also strongly suggest that the BAG get involved in this and codify whether or not bots should be allowed to run unapproved tasks in userspace. By the way, given that the task was run in userspace with the sole purpose of then having it impact on main-space, I am horribly disappointed in Betacommand's reasoning that "since it's in userspace" it was okay. - Philippe | Talk 17:09, 5 February 2008 (UTC)

I believe that if BCB's edits were not so crucial as they are, his block should remain. I believe that until we can get his tasks split up between trustworthy users, he should remain unblocked, but merely to do work that is extremely difficult and/or tedious by hand. When this is finished, he should be blocked. I barely even edit Wikipedia much anymore, but I still know of the infamous qualities of BCB. I don't mean to slam on its master, Betacommand, because his edits have been very helpful. Unfortunately, I must agree that these crucial tasks must be handed over to someone who will not make such errors in judgement. - ђαίгснгм ^таιќ 04:52, 6 February 2008 (UTC)

Unblocked BCB, left note for Tim. Rich Farmbrough, 11:07 6 February 2008 (GMT).

Rich, how do propose the above concerns should be addressed? A warning to Betacommand not to run new bot requests or code without getting them approved first? You do realise that is what he did, right? And that Tim's block was after the bot had ceased its "system resources abuse", and was probably to prevent future abuses of system resources? How can this be prevented in the future? Has Betacommand said clearly what he has learnt from this? If you are going to unblock after a long discussion like the one above, I think you should say a bit more than just:

• Tim's talk page: "BetacommandBot. Hi Tim, just to let you know I'm unblocking this account as the system resource issue is no longer. Rgds" (10:52)
• Unblock log: "System resource issue no longer an isue - bot not doing that" (10:53)
• Betacommand's talk page: "Bot unblocked. Effective now. Verb sap. etc." (10:56)
• In this thread: "Unblocked BCB, left note for Tim." (11:07)

Otherwise it seems that you are ignoring the above concerns. Carcharoth (talk) 11:49, 6 February 2008 (UTC)

As the unblocking means that no further admin action is required, it seems that this discussion will have to move to a requests for comments, which I was hoping to avoid. If Betacommand is more responsive to the questions above, then maybe this can be avoided. Carcharoth (talk) 11:49, 6 February 2008 (UTC)

Well, the block served no further purpose. BCB isn't going to do "something silly" in the next six days, so we need to avoid cutting off our nose to spite our face, here. What BCB did in this case was well intentioned, and could have been done without breaking any "rules" in a dozen different ways. The question of user behaviour can be taken up with Betacommand on his talk page, or RFC if appropriate, of changes to bot processes or standards on the appropriate discussion pages. This particular incident is, it seems to me, only important as part of a pattern of behaviour, which if seen as a problem, should not be addressed by temporary block. Rich Farmbrough, 12:24 6 February 2008 (GMT).

I have raised the bot issues at Wikipedia talk:Bots/Approvals group#Bot issues at WP:AN thread, but there has been no response so far. If you feel this was meant as a temporary block to prevent abuse of system resources, an abuse that is indeed no longer occurring, why not ask Tim that first? My reading of it is that you are wrong, and that the block was aimed to be left in place until Tim or the community (not just you) were satisfied that Betacommand had responded adequately to the concerns raised. Neither East718 nor Betacommand have indictated what they discussed, who they discussed it with, and why they failed to discuss it with anyone. Just saying "I fucked up" is not answering those questions. I will copy my comments above over to Betacommand's talk page. The level of responsiveness is a concern, and is a constant refrain. It shouldn't have to take all this to get Betacommand to respond to concerns. Carcharoth (talk) 13:40, 6 February 2008 (UTC)

Whoa, what, Rich? You unblocked? There are clearly concerns, that is not a consensus action. I'm not sure that's the route I'd have taken. - Philippe | Talk 13:53, 6 February 2008 (UTC)

I'm less concerned about the unblock than the poor response by Betacommand so far. He has said he is sorry for causing drama and doesn't intend to repeat this action. No apology for the failure to discuss. TO make it easier for him to respond, I've laid out my concerns at his talk page. See here. Carcharoth (talk) 14:20, 6 February 2008 (UTC)

And Betacommand has responded there. As I said here, I'm satisfied now those answers have been given, and am happy for this AN thread to be archived (though others may want to continue). Some issues remain for WP:BAG to discuss, and for East to respond to when he returns. Carcharoth (talk) 14:53, 6 February 2008 (UTC)

source code?

An important issue that is central to this is that of the source code. My understanding is that it is currently closed? I see a number of requests on here for BC to open it to the community (or at the least "approved" members of the community?) Can we get a yea or nay on that request? If it's a nay - that's fine and I have no issue with that but the community needs to know so that we can start developing bots to perform those important but routine tasks and are under the control of the community rather than single editors. This is no slight on the good work that the bot had performed to date but rather a pragmatic way forward and an attempt to insure that the best needs of the community and project are served. --Fredrick day (talk) 16:25, 5 February 2008 (UTC)

As I was about to post above, I do release my code to a very very limited number of trusted users. one I dont like clones of BCBot floating about that I cannot control. the code is very powerful (Ive clocked it at 700 edits a minute). I dont make my code idiot proof. I build it so that I can use it without a lot of hassel. Also I dont have the time to review code changes made by other to my code. As for splitting the bot into several accounts Ive been working on that. I have also been doing other requests (Wikipedia:Image renaming) changing several thousand lines of code spread accross multiple files just takes time. with RW commitments I need 30 hours a day, and as we all know there are only 26 hours in a day. β^_command 16:54, 5 February 2008 (UTC)

"the code is very powerful (Ive clocked it at 700 edits a minute)". Come on. Editing a page is pretty simple, code-wise, and there are vast quantities of free code available. I have very strong doubts that any code can be any more "powerful" than other code at making edits. And even if your code has some magic that makes it ten times faster than everybody else's, accounts that are not autoconfirmed can make only 8 edits per minute, and other accounts tend to have internet connection latency and speed as a much greater limiting factor than how "powerful" the code is. — Werdna talk 06:02, 6 February 2008 (UTC)

I started a conversation to this end at Wikipedia talk:Bots/Requests for approval#Code publishing.3F yesterday and the response has been lukewarm at best. My feeling is that bot operators don't want to release code

1) To keep control of it tightly, and have it be "theirs". I.e., no clones.

2) Arguments that the bot codes are "sensitive".

3) Operators not wanting to have to clean up their code.

4) Bot operators not wanting to be responsible for others using their code.

The only way to force release would probably be if the community (not the BAG) ordered that all code be published to get authorized to run new bots. Lawrence § t/e 16:35, 5 February 2008 (UTC)

I can understand all of those feels but the community really needs to achieve a better balance between considering the feelings of the people creating the bots and ensuring that we have sufficient control to ensure Continuity of operations/development if a developer goes under a bus. I think retroactively asking for all code to be released can be a problem but certainly there must be scoop for improving the situation with future approvals? --Fredrick day (talk) 16:51, 5 February 2008 (UTC)

As I understand it, the crats actually oversee the BOT process, but have delegated the review part to the WP:BAG, so if the crats said as a group "no new bots/bot tasks may run unless the code is released via X means" that would solve this debate IMHO. MBisanz ^talk 18:14, 5 February 2008 (UTC)

Bureaucrats only give out the bot flag. They don't say whether automated scripts are permitted to run to begin with. —Simetrical (talk • contribs) 02:43, 6 February 2008 (UTC)

All of those reasons are standard and patently wrong-headed objections to open source.

1. Possessiveness is not an acceptable excuse for anything on Wikipedia. You can choose to participate or not, but if you choose to, in certain respects you need to forfeit all possessiveness for the good of the project. That already applies to edits' copyright and there's no sound policy reason it should not be applied to bots as well.
2. No bot code is sensitive. Any idiot can already hack up a ten-line Perl script to spam as many edits as he pleases, once he creates an account and figures out how to calculate a valid edit token. And anyone willing to look at the source code of an anti-vandal bot to figure out how to more effectively vandalize would evade the bot somehow in any case, it's not very hard.
3. Nobody says you have to clean up your code, you just have to release it, and provide just enough documentation that others can actually get the thing running. If it's an ungodly mess, that's not great, but people can still run it if you disappear ― at the very least, until someone can write up a replacement.
4. Every bot will still have to be registered and approved, and it's the operator who's responsible for what the bot does, not the author. If someone wants to run a bot whose source code they don't remotely understand, they should probably be denied the right to do so without fairly good reason.

I think that, in retrospect, it was always a bad idea to have the BAG consist of bot operators. It would be better for it to consist of programmers who can understand the issues involved but don't actually run any bots. There's too much conflict of interest right now, and as a result, too much self-indulgence, IMO. Bot authors and operators provide invaluable services to the community, but that doesn't mean they should be given any right to operate against the project's principles or its interest. Better people get by with manual methods all along than rely on a closed-source bot that may suddenly disappear one day. —Simetrical (talk • contribs) 02:43, 6 February 2008 (UTC)

Here is a bold suggestion: Treat all (bot) source code in a similar way that we treat articles. Let everybody be able to edit the code: Create new tasks, improve any code, debug, etc. Vandalism is dealt with similar to main space vandalism. Unwanted tasks (not compatible with policy) will be deleted. Good-faith-errors will be corrected. Disputes solved by trying to reach consensus. (Utopia or straighforward?) If we require all bot runs to be revertible (similar to ordinary edits – when the developers come up with a robust rollback function), then running a bot is not such a big deal. Oceanh (talk) 02:36, 6 February 2008 (UTC).

That would be a HUGE security risk, in my opinion. SQL^{Query me!} 04:41, 6 February 2008 (UTC)

Very bad idea. It would only take a change to one line of OrphanBot's code and it would start replacing unsourced images with penis pictures. --Carnildo (talk) 05:53, 6 February 2008 (UTC)

Indeed very bad. A few lines changed in ClueBot could either completely disable it or make it revert *everything*. Furthermore, one character in ClueBot's source could make it negate the heuristics and thus revert only good edits but not bad edits. Also, the huge security risk to the servers that run these bots ... Most of these bots are written in very powerful languages. Languages that are capable of lots of stuff outside of Wikipedia. Languages that could run DDoS code ... Languages that could forkbomb the server running them. You know the limit of 100 ifexist calls per page? A forkbomb could (and indeed would, if coded properly) use billions of more resources than an ifexist call. As the bots' password has to be stored on the server somewhere, readable by the bot, what keeps someone from writing a piece of code to read the password and post it to a Wikipedia page? What about the ability to run any command on the server the bot is running on under the user running the bot (often the same user account that the bot owner uses). We know that bots can go at very high speeds, now what if someone were to compromise 5 or 10 bots? Someone could very easily vandalize a lot of Wikipedia before someone could block the bots. Suppose 10 bots at 10 edits per second (this is possible with forking and such), and it took 1 minute to block all 10 bots, in that short time, the bots would have vandalized 6,000 pages. Now, let's assume that it took a bit longer to block them and it took 5 minutes, the bots would have vandalized 30,000 articles. -- Cobi^(t|c|b) 06:33, 6 February 2008 (UTC)

Good grief. I don't think anyone was suggesting that the live code be made publically editable without review. — Werdna talk 07:26, 6 February 2008 (UTC)

What about the idea of a trusted repository like OTRS? They maintain copyright permissions and secure user communications. As part of the bot approvals process, each bot task could be issued an OTRS ticket number verifying that a copy of the code has been given. Then, if something like BCB's block or Gurch's travel happened, an OTRS rep could issue the code to a new, vetted bot operator (probably an existing operator) who would at least have that far a head start on coding a replacement (if not being able to implement the bot automatically).

As a followup, some bot ops have released their code and there are somewhat useful (to a programmer) documentations on how to code in specific languages used on wikis. What is to prevent someone from coding their own bot, setting up several sleeper-socks, and doing the same thing you suggest? Using rotating names and proxies, a concentrated attacked, like what happened to GRC.com several years ago is probably just as possible with or without bot code published.

Certainly, somebody could do something like that if they had the time, motivation and resources. But somebody that motivated is not going to be stopped by a lack of available bot code. I mean, there are countless wiki editing modules available on CPAN, and it's not exactly difficult to write one. I think it took me an hour or two. — Werdna talk 07:30, 6 February 2008 (UTC)

In fact ClueBot has had its source open since its inception, and I haven't seen any vandals who have been trying to get around ClueBot.  :) -- Cobi(t|c|b) 00:50, 5 February 2008 (UTC)"

so isn't there already a risk of someone switching the heuristics or forkbombing (assuming that part of the code is released) and re-running it on their own system to attack your server or wikipedia? MBisanz ^talk 07:13, 6 February 2008 (UTC)

No more risk than there is of somebody picking up pywikipedia and sending through some mass-changes of replacing images with something offensive. It is my opinion that the place of anti-vandalism bots is to revert unsophisticated vandalism by bored schoolkids — bored schoolkids who are, on the whole, unwilling and/or unable to look up the source code of particular antivandalism bots and to figure out how to get around them. — Werdna talk 07:21, 6 February 2008 (UTC)

No, you misunderstand, forkbombing is an attack against the server executing the source code, not against a remote server like Wikipedia. And, no, ClueBot doesn't have any fork bomb code in it. Anyone technically savvy could easily create a bot (or use a heavily modified version of any open source bot) and run it on their own computer. That isn't the reason it is a bad idea. The reason it is a bad idea is that most people aren't going to take the time to do that, but if they could insert a single '!' into User:ClueBot/Source to make it negate its heuristics, that is a different matter entirely. Furthermore, the major bots on Wikipedia are trusted when you see them make an edit. How many times have you checked MiszaBot III's edit to see if it actually just archived stuff without changing anything else? -- Cobi^(t|c|b) 07:25, 6 February 2008 (UTC)

I've checked Misza's a grand total of once, when I set up an archiver for another user and screwed up the date coding and had to go and undo Misza's move to correct my error. And I'm not saying code should be editable, but right now someone with the proper knowledge could take User:ClueBot/Source, copy it to a non-protected page, insert the !, and do damage across many sleeper accounts. On the other hand, requiring all code to go to OTRS for storage, would solve the problem of disappearing bot ops, compromised bots (we've had compromised admins), and maybe even eliminate the need for bot ops to feel some duty to publish their code. MBisanz ^talk 07:33, 6 February 2008 (UTC)

I don't think you really understand what you're saying there. Putting the code in a non-protected page and editing it doesn't magically make it run (try it!). You need a server, you need several accounts, you need decent bandwidth, you need the appropriate software installed on your server. Developing and operating a vandalbot is certainly not a particularly easy task. The easiest part is figuring out how to edit Wikipedia from a script. — Werdna talk 07:44, 6 February 2008 (UTC)

(ec) No, they couldn't. They would have to have access to a Linux machine, a PHP CLI interpreter, a working knowledge of PHP, cURL, MySQL, and a ton of other prerequisites. That page is solely for reference, that page has no special status. As a matter of fact, you can edit it now. It won't change a thing, believe me, many have tried. You would have to copy the source into the correct files, create a proper directory structure, MySQL schema, and create a config file for it from reading the source and figuring out which variables need setting. They would then have to create User:TheirBotName/Run and set it equal to true, otherwise the bot wouldn't run. Then they would have to actually start the bot and after the bot did some sanity checks and read some wiki pages, then it would need to connect both to Freenode and to the IRC feed. After it did that, it would attempt to login as the user defined in the config file, then and only then would it start doing what it is supposed to do (with any changes the malicious user decided to make). -- Cobi^(t|c|b) 07:54, 6 February 2008 (UTC)

(ec2)No, I get that I can't run a bot by magically entering a script. What I mean is that if Cobi's fear is that a user will edit 1 character of his code, so he puts it on a protected page, {or doesn't I see now), along comes a vandal programer, and now that code is in an area that it can be copied to an editable area (I said page, but compiler, text file, etc would also do), then a vandal could copy it, change the code, compile it, run it on their server, and do all sorts of damage. MBisanz ^talk 07:58, 6 February 2008 (UTC)

The difference is the access to the trusted bot's account. User:ClueBot is recognized and trusted not to cause havoc. This saves a lot of time: people don't have to double-check all its edits. In the case of many bots, they even have bot flags, so that not only do people not have to double-check their edits, they can't (at least on RC) unless they go out of their way to show bot edits. Currently nobody does, precisely because the bot flag indicates trustworthiness.

What you are suggesting is, to all intents and purposes, that anyone (who, granted, knows a little programming) be allowed to make any edit they please using the account of any bot. The problem with this should really be very obvious. The entire idea of accounts is to allow recognizability, the ability to associate different edits with each other and use that to draw conclusions that allow you to treat different edits differently. If that weren't important, we would just do away with all account names and IP addresses, and have no way of telling one edit's author from another. It's a bad idea, and no better for being suggested only for bots. Why don't you just have all bots run under one account? It would amount to pretty much the same thing.

This is on top of the fact that arbitrary code execution by untrusted third parties is completely unacceptable in any context. Arbitrary code execution immediately implies trivial DoS capability, intentional or unintentional. This is why you cannot edit the source code of MediaWiki. This is why you cannot edit site JavaScript unless you're an administrator. This is why you cannot run arbitrary SQL queries. Even if you were totally well-meaning, inefficient code or a simple programming error could kill whatever server is running the code. This is why load on the toolserver, where anyone can sign up for an account to run programs, is ridiculously high. Nobody there is malicious (well, except for one guy who apparently ran a Counterstrike server, but that didn't last long). They just aren't being as careful as they should be about optimizing code. And they still bog the server down to a crawl.

So no, really, this isn't a good idea. I promise.  :) If you want to advocate anything more extreme than just requiring the bot to be open-source, you could suggest that it be put in a central repository on a Wikimedia server that a large pool of trusted users can access, and a much smaller pool of trusted users can put live after review. This would be basically like how MediaWiki development is handled, or how most open-source projects are. I'm not advocating that, at least for the present. —Simetrical (talk • contribs) 01:11, 7 February 2008 (UTC)

Wikipedia is one of the strictest sites about using only freely licensed images and other media. If we are so strict on images that appear on the main page, why, no, how can we not look at the code that a bot runs to make edits to thousands of articles? How can we allow this code to be proprietary, when the rest of the site is free? If featured articles require sources, multiple editors, etc. How can we allow bots to just be approved by a committee and ran? Does the community as a whole get a say in this? I firmly believe that bots on Wikipedia must be open source, that all editors on Wikipedia can view and comment on. Though you may realize that there are a million reasons that this "would never work", everyone here realizes that this has to happen. Something like this fiasco cannot happen again. - ђαίгснгм ^таιќ 05:08, 6 February 2008 (UTC)

This current episode happened because an admin and a bot operator did not communicate with developers and larger community – licensing of the code has little relevance. Bear in mind also that the key role of BAG is not really vetting the code, it's vetting the operators. There are examples of people picking up some code, and wanting to run it without understanding how it works (just browse the rejected bot requests). I don't believe the community on the whole is geeky enough to be trusted with this role, as one must know something about the subject to be able to act as a judge. Turning bot requests over the community would simply turn it into a popularity contest, a bit like RfA. Most of the people who know enough of programming and are interested in bots have already gravitated towards BAG. – Sadalmelik (talk) 06:27, 6 February 2008 (UTC)

But that's not really the issue from an organisational point of view, the issue is that we are becoming highly reliant on closed sourced bots and their owners to perform many routine or important tasks. If that editor falls under a bus or goes rogue, then we don't currently have any redundancy in place. Yes - those things can be done manually, yes someone could write a replacement bot - but it's still a waste of time and resources when we can develop processes that minimise disruption from the start by bring bots more tightly under the control of the community. Hell we don't even have to do anything with the sourcecode, just log it somewhere (which does not have to be publically available). If you want to register a bot, you turn over the source code and you agree that in the effect of your incapacity that someone else can manage/develop it. I have a bit of a background in organisational resilience and this stuff seems like a non-brainer to me. --Fredrick day (talk) 11:32, 6 February 2008 (UTC)

This issue has been discussed extensively over on the Bot pages. Two more useful things (than open source per se) would be if each significant task has a number of operators that can perform it, and if rather than being extrinsic, bot functions could be made intrinsic to MediaWiki. Nonetheless this would not stop, nor should it, people using automation in general, without redundancy. Rich Farmbrough, 12:46 6 February 2008 (GMT).

Agreed. Rich, is there a well-organised list of bot functions, the history of who performed them in the past, and who performs them now? If not, shouldn't some effort be made towards carrying out that sort of organisation? Carcharoth (talk) 13:46, 6 February 2008 (UTC)

No, and what is needed is carefully constructed list of important bot functions (i.e. that would be missed). However that's the easy part, the hard part is ensuring that they are all "covered off" by another bot operator - and maintaining the list as operators come and go. Whether BAG would take on this task I don't know, most BAGers seem inordinately busy. Rich Farmbrough, 14:47 6 February 2008 (GMT).

If bots were open-source and acceptably documented, each significant task would automatically have a number of operators who could perform it, i.e., anyone who wants to download the source and get it running. That would be the major point (other than moral freeness issues) in requiring open-source bots. As for adding bot functions to MediaWiki, the main problem is that there's necessarily a much lower barrier to running a bot, than to submitting code to a widely-used application that is relied upon for performance and functionality by not only Wikipedia but thousands of third parties.

There are definitely bot tasks that should be in the software, though. One thing that would obsolete quite a few bots would be a reasonable discussion system (no need for archiving, signing, ...). Another thing that would help a ton is a sufficiently flexible task-management system, which could assist in things like AFD, RFA, etc. by automatically creating pages, maintaining lists of open tasks in a category, tracking time limits, listing ended discussions in an appropriate place, and so on. Improved handling of templates, categories, and interwiki links would kill a whole bunch more bots. If anyone wanted to write new features to help out with that kind of thing, it would be great. Anti-vandalism is probably best to keep in bots, though, because it doesn't really benefit from tighter connection to the software at all and is very fluffy and heuristic-based, so hard to get unambiguously right. —Simetrical (talk • contribs) 01:11, 7 February 2008 (UTC)

No, those things should not be in the software, because they're not set in stone. This is one of the strengths of wikis - that we can get processes working without the software enforcing them. These are social policies set on each wiki, and should never move beyond that. – Mike.lifeguard | ^@en.wb 15:01, 7 February 2008 (UTC)

A built-in discussion system and improved category/template/interwiki handling should definitely be in the software. A task-management extension would also be very nice to have. I think all of these could be made flexible enough to accommodate most needs currently served by bots that handle those tasks, and to the extent they can't, the bots can be retained. Compare to the introduction of categories as an improvement over lists: lists have become much less important, but are still used due to limitations in the category system (e.g., no ability to associate different names to articles, no ability to add comments next to entries, no ability to add custom headings, . . .). That doesn't mean categories aren't an extremely valuable feature, it just means they aren't perfect.

Integrated features have a variety of critical advantages over bots, including speed; reliability; ease of use (using a specialized interface vs. editing wiki pages, the latter possibly requiring arcane templates, etc.); and working out of the box, which is extremely important not only for third parties, but for smaller wikis that don't have large numbers of people willing or able to run bots. (I doubt there are a lot of things done by bots on any but the ten largest wikis.) As I say, in a few cases, like anti-vandal bots, none of these qualities are as important as unlimited adaptability, which bots have and the software usually doesn't. But in many cases, the more limited adaptability of a software feature is easily sufficient.

The task-management extension is perhaps the most arguable of my suggestions. I do think it would be very valuable, and could enormously simplify participation in processes like AFD. There's been some discussion at mw:User:Robchurch/Work queue, but no work has been done on it, at least not by MediaWiki developers. —Simetrical (talk • contribs) 17:04, 7 February 2008 (UTC)

Possible vandal bot

I've read a lot of mistakes in the discussion above, I'd like to try to correct some.

1. A DoS attack using one of the multiple bot frameworks that are available, or one modified bot, is very unlikely to happen. Fork bombs are absolutely off-topic here : That would require several servers running very speedy bots to work, and from my point of vue, there are others types of attacks, way more efficient that could be used to take down the fundation's servers. Modifiying a bot in order to attempt a DoS attack is a waste of time, and that's for sure.
2. The other type of attack that you should worry about would be a simple mass automated edit wave. Not some attempt to take down any server, but simply some characters switching, some page blankings, etc, to alter the content and not the service. And such an attack is VERY EASY to set up actually, without having to modify any particular bot scripts:
   • Pywikipedia only needs : python, an internet connection, and a fresh wikipedia account. Easy.
   • With one of the basic script included in pywikipedia, you can perform automated replacements and customize your edit summary
   • a 10 (Actually, the limit for unconfirmed user is 8) edits per second edit rate can be very, very easily reached with a very slow connection and/or hardware. (And I insist on that : Our bot scripts are using timers around every corner to slow down their processes)
   • Setting up 10 computers to start such an attack at the same time is very easy for an individual.
   • Also some IT students have access to powerful servers that are way more speedy than individual computers: I'm currently sitting at a computer school, which has 4 *huge* servers for its student needs, with a 15Gbps connection : I will not try, but setting up 50 threads to make automated edits from 200 different accounts, using automatically changing edit summaries to make them hard to detect (hum... generated from random google searches ? or, better, from edit summaries from the last XML dump ? ) is EASY STUFF, really.
   • I don't think that there would be any way to easily counter such an attack, and yet I'm not even this experienced on computer attacks. Really, I do think that wikipedia is already vulnerable to such an attack, so please, please, please, stop worrying about a possible attack following the release of the code of an antivandal bot...

Wannabe hackers have more chances to succeed monitoring the latest vulnerabilities, or using a very simple replacement bot.

How many of you, bot owners, 'crats, and admins, are logging in using HTTPS to secure your password ? Really, I think that there are some security issues requiring way more attention that this so-called problem about bot code releases...

NicDumZ ~ 13:02, 6 February 2008 (UTC)

Most of your points are true, however, you have lost the context. All of this was in response to someone suggesting that we let the bots' live code be changed by anyone at any time. This opens all sorts of attack venues against the server that the bot's code is running on, including a forkbomb against the server running the bot (while (1) pcntl_fork(); for PHP) or using the bot to download a DoS program to the server running the bot in order for the attacker to add another server to their DDoS botnet. The problem is not releasing the code ... ClueBot's code already is. The problem is letting anyone change the live, running copy of it and having the server running the bot execute the new code. -- Cobi^(t|c|b) 18:30, 6 February 2008 (UTC)

Oh, did someone really proposed this ? :S

I assumed that live code was meaning that a running bot should have its actual source published somewhere, and allow others to change it, but... using a SVN, or this sort of system ?!

NicDumZ ~ 13:27, 7 February 2008 (UTC)

Yes, above:

“ Here is a bold suggestion: Treat all (bot) source code in a similar way that we treat articles. Let everybody be able to edit the code: Create new tasks, improve any code, debug, etc. Vandalism is dealt with similar to main space vandalism. Unwanted tasks (not compatible with policy) will be deleted. Good-faith-errors will be corrected. Disputes solved by trying to reach consensus. (Utopia or straighforward?) If we require all bot runs to be revertible (similar to ordinary edits – when the developers come up with a robust rollback function), then running a bot is not such a big deal. Oceanh (talk) 02:36, 6 February 2008 (UTC). ” — User:Oceanh, Wikipedia:Administrators' noticeboard

-- Cobi^(t|c|b) 05:50, 8 February 2008 (UTC)

RFC/RFAr against BetacommandBot

Betacommand is stubborn and does not want to fix his bot. And you all ignore the problems and let BetacommandBot cause more damage. Everytime BetacommandBot gets blocked, it gets unblocked very fast, whether the block is correct or not. <redacted sentence> Please deal with the problems. File an RFC/RFAr against BetacommandBot. Also, shut it down until it is fixed. --Kaypoh (talk) 01:43, 7 February 2008 (UTC)

If you've got a problem with BCB, you take it to RfC or ArbCom. Don't come here demanding others do it. ➔ REDVEЯS has changed his plea to guilty 14:28, 7 February 2008 (UTC)

And, if you want people to take you seriously, start your userpage; That bright red link just looks stupid. (and inviting to vandals). BETA 00:26, 8 February 2008 (UTC)

Having a userpage isn't mandatory, and isn't a prerequisite to being taken seriously. There are admins (such as User:JzG) who do not have userpages, but still expect (quite rightly) to be taken seriously. Dreaded Walrus ^{t c} 09:54, 12 February 2008 (UTC)

• giggle*

Possible Solution

Attack the source of the problem:

1. Change the revision limit protocol to only take the size of the article and history into account, not the amount of edits. (if that makes a difference)
2. write in a permanent redirect (or add a normal redirect and protect the article) to a 404-like "article has been deleted, click to go back" page.
3. Progressively delete the original article.

Hope this helps,

BETA 00:11, 8 February 2008 (UTC)

Regarding (1), you misunderstand the cause of the problem. A 1 GB article (not that those are permitted on Wikipedia) is no slower or faster to delete than a zero-byte article. Deletion moves a row from the revision table to the archive table; it doesn't touch the external-storage servers, where the text of the relevant revisions is stored. The text of the article remains exactly where it is, it's just now associated with a deleted revision's ID instead of a non-deleted revision's ID. This is why the limit is on revisions per page, and not the size of the revisions.

I don't understand (2) or (3). What are you proposing to redirect to what? What do you mean by "progressively delete"? —Simetrical (talk • contribs) 21:05, 8 February 2008 (UTC)

The reason for the revision limit, as I understood it, was to stop the server from grinding to a halt when a large article and its history are deleted. That's the reason for this suggestion. Progressive deletion was just an idea to alleviate this congestion. delete it piece by piece so it's easier on the system. I must have gotten it wrong somewhere. Oops BETA 15:17, 9 February 2008 (UTC)

Well, if by progressive deletion you mean deleting a few hundred revisions, waiting a while, deleting a few hundred more, waiting, etc., then yes, that would certainly work, and would be a feasible solution. It and similar solutions have been discussed, for instance, here. The problem is that nobody's written the code to do it. The current code is just a patch job of a few lines: it's not meant to be a definitive or permanent solution, it's just meant to stop people from breaking the site until someone writes a proper fix. —Simetrical (talk • contribs) 00:38, 10 February 2008 (UTC)

The solution is to block the bot and leave it blocked until Betacommand fixes it. The only reason he hasn't is because he doesn't want to. Jtrainor (talk) 07:15, 11 February 2008 (UTC)

Request for more admin eyes on an issue (RE:RfCU result)

Moved to Wikipedia:Administrators' noticeboard/Mantanmoreland.

Request from Abbott Corp.

I blocked the username User:Abbott Corp. Public Affairs per Wikipedia:Username policy. I got the following response by email:

I am a member of the Abbott Corporate Public Affairs team. We have a great respect for the importance of wikipedia and the need to have the most accurate information contained within its pages. In an effort to ensure that wikipedia entries are factually accurate and up-to-date on Abbott-related pages, we attempted to set up a username that would provide complete transparency for any edits that we propose. However, before we had the opportunity to suggest even a single edit, our account was blocked indefinitely.

It is extremely important to us -- and I would hope all users -- that the information contained on the site be accurate. We have no desire to make editorial changes and we believe that the credibility of our edits can only be enhanced by full disclosure as to who we are. We do not want to use this site as a promotional tool. However, we know that the site is an important reference and, therefore, we believe that factual edits, made with complete accountability can only add value to the content.

We would appreciate if you would unblock our editing capabilities and allow us the full disclosure we seek. Alternatively, please advise us as to the best method to make factual corrections for Abbott-related sections. Thanks.

Tom

Should I do anything other than cite the username policy back to them? NawlinWiki (talk) 16:54, 8 February 2008 (UTC)

I would probably cite WP:COI to them and ask him how his boss will react when Abbott appears in a news article as a Wiki-spammer. Ronnotel (talk) 16:59, 8 February 2008 (UTC)

(ec) I'd write a polite note asking them to mindful of WP:COI, reliable sources and verifiability, and thank them for acknowledging that they do not want to use the site as a promotional tool; and of course citing the username policy, as you suggest, is the right thing to do, IMO. Cheers, Antandrus (talk) 17:01, 8 February 2008 (UTC)

(ec):I'd say that's OK, particularly "Accounts that represent an entire group or company are not permitted". They can always email the Foundation if they seriously dispute any content. --Rodhullandemu (Talk) 17:02, 8 February 2008 (UTC)

A particularly helpful site to point him to is Wikipedia Scanner. Ronnotel (talk) 17:03, 8 February 2008 (UTC)

(ec) It looks like he's being pretty open about who he is and who he works for, which is refreshing, but the username block was the right thing to do here. I'd offer an explanation of the username policy, point out COI issues as above, and maybe give him a pointer towards OTRS as an alternate route to expressing any concerns with articles. Tony Fox (arf!) 17:04, 8 February 2008 (UTC)

• The name would be OK if it's confirmed as a single editor, but of course role accounts are forbidden. Let's make sure, though, that we're not preventing them fixing a pressing problem of neutrality (I don't know, I haven' seen the article). Guy (Help!) 17:10, 8 February 2008 (UTC)

Have they made bad edits? No? Then let them be. If a company is openly delcaring a potential COI for all to see and observe (as we tell them to by our own policy and we so commonly encourage), then by all means let them edit. Does it really matter in the grand scheme of things if 1 person or 4 are editing from the same username? If the company shows issues and doesn't comply with our rules, /then/ block, don't block because you think they /might/ create problems. ^demon^{[omg plz]} 01:51, 9 February 2008 (UTC)

Unless I've got it badly wrong, the GFDL licence is a personal licence, not a corporate one, and if this is the case, lumping contributions together is probably the reason we don't permit role, or group accounts. It's an issue of accountability, as I see it. --Rodhullandemu (Talk) 01:58, 9 February 2008 (UTC)

If that's the issue, then I more than willingly will accept that as a reason to block role accounts. Historically, I thought it was for other reasons. ^demon^{[omg plz]} 02:01, 9 February 2008 (UTC)

Well, (puts on lawyer hat) if the Foundation is sued for libel, it would be great if it could join the perpetrator of the allegedly libellous edit as a co-defendant and sit back as an "innocent publisher", as long as that editor can be identified. The problem with group/role accounts is that one would then get into difficult issues about vicarious liability, so it's better just not to allow them. Makes perfect pragmatic legal sense. --Rodhullandemu (Talk) 02:20, 9 February 2008 (UTC)

It doesn't matter. Copyright is copyright; they can own it, they can license it, and it says nothing about excluding them. Corporations can and do contribute under the Gnu licenses (IBM under the GPL, for example). We can say with near certainly that some corporations have already paid people to edit Wikipedia. If the foundation gets sued, we say "sec. 230, not content creator" and turn over the editing records. It doesn't really matter to us whether a corporation or individual might be liable.

I applaud their disclosure of COI. Let's see what they have to offer. Cool Hand Luke 07:51, 9 February 2008 (UTC)

So once again, we're encouraging people to edit under innocuous usernames where no one may perceive their COI. Cool Hand Luke 18:02, 9 February 2008 (UTC)

Actually, we encourage people to edit under innocuous names so that there is no illusion of "fake credentials", i.e. if you are named after a corporation your edits could be taken to mean you officially represent that corporation. As we have no way to prove that, the use of a corporate name may be by an unrelated person, for example someone wishing to slander or libel the company, but who takes their corporate name to "fake us out". --Jayron32.talk.contribs 04:15, 10 February 2008 (UTC)

For what it's worth, a corporation is a juristic person within the meaning of law, or so my business law prof tells me :-) As I understand it, a corporation can libel and be libeled. However, I've only ever been accused of WikiLawyering, not actual lawyering, so take this with a grain of salt. --SSBohio 04:10, 10 February 2008 (UTC)

For what its worth, its' not really our job to worry about that. The foundation pays someone very well to handle those issues. --Jayron32.talk.contribs 04:15, 10 February 2008 (UTC)

I'm familiar with the man. He's a law unto himself, as it were. I was offering that information to bolster what Cool Hand Luke was saying above about corporate responsibility. --SSBohio 05:48, 10 February 2008 (UTC)

Let's be nice! The block was appropriate but so too is a polite outreach to explain that there's a problem with role accounts but that if someone wants to create their own account with a proper name they may. The person is probably familiar with WP:COI but a polite reminder there is in order. If they want to be transparent they can mention their affiliation on the talk page, and add a mention on the talk page anytime they do more than a spelling or format correction. If they do that, and they're modest with their edits, nobody is going to blame them for being a spammer. Wikidemo (talk) 23:07, 11 February 2008 (UTC)

A simple answer might be to create usernames user:Abbott Corporate P.R. (John) or the like?? FT2 ^{(Talk | email)} 19:35, 12 February 2008 (UTC)

Sporadic editors, sleeper socks, and single-purpose accounts

Probably not quite the right place for this, but as a spin-off from an ANI thread, I thought here might be as good as anywhere. I wanted to put forward a couple of quotes about sleeper socks, sporadically editing accounts (no link for this) and single-purpose accounts. See the rather long Wikipedia:Administrators' noticeboard/Incidents#Is it just me... thread for the full background, but first try looking at the following and see what you think.

Consider the following accounts and their contributions:

• Amelia9mm (talk · contribs · deleted contribs · page moves · block user · block log)
• Jrichardstevens (talk · contribs · deleted contribs · page moves · block user · block log)
• Drstones (talk · contribs · deleted contribs · page moves · block user · block log)

Now consider the following quotes:

• Trusilver on Amelia9mm (after declining an unblock request) - "Amelia has very little of an edit history and dropped in to participate in a hot button issue after a long time between edits. That smells extremely strongly of a sleeper sock to me. At the very least it is incredibly suggestive." - Trusilver 03:16, 10 February 2008 (UTC)
• Amelia9mm on herself and her block - "I don't know why this is blocked, nor do I know what I could possibly have done that is disruptive. Why would I want to bother to come to improve info on wikipedia if some total stranger can block my account? How is it that someone I have never heard of can block my account INDEFINITELY? His linked page says "I am here for some very limited purposes.." Hello? Some explanation could be useful here!" - Amelia9mm 16:29, 9 February 2008
• Relata refaro on Drstones - "I certainly think that an account that has had three or four edits over months and suddenly explodes into life on one issue contentious on RW is, if not necessarily a sleeper sock, hardly anything other than an SPA. DrStones made five edits to three articles before spending a lot of time on this one." - Relata refaro 07:28, 10 February 2008 (UTC)
• Carcharoth comparing himself to Drstones - "My editing history shows that up until the end of July 2005 I had made most of my edits to 7 July 2005 London bombings and related articles and debates (about 100 out of 110 edits). For all intents and purposes, I was a single-purpose account focused on that article. In today's climate, if I had made a mis-step, said something unsourced about one of the bombers or bomb victims, or encountered a bad block for a multitude of other reasons, I might have been so affronted at my treatment that I might never have returned." - Carcharoth 07:57, 10 February 2008 (UTC)
• Carcharoth on Jrichardstevens - "Eight innocuous edits to unrelated articles over a period of around two years. Not the greatest of contributions, but the potential is there for someone who was interested enough to register an account to (one day) start contributing more. Indeed, the foray into Wikipedia namespace showed someone who might well have started contributing more." - Carcharoth 23:25, 9 February 2008 (UTC)
• Orderinchaos on Jrichardstevens - "Eight edits two years ago is hardly 'history'." Orderinchaos 01:49, 10 February 2008 (UTC)
• Jrichardstevens on himself - "As someone who reads much, edits little and talks even less, I would suggest that none of you can judge the intentions of an editor based on sporadic participation. Silence or only occasional editing might mean that someone is inclined to read more than write, something I think you would favor in at least some of your membership." - Jrichardstevens 14:58, 10 February 2008 (UTC)"

I obviously disagree in the strongest possible terms with Trusilver's judgment of what can be reliably said to be a 'sleeper sock'. I'm struggling to find a definition for this on Wikipedia. I think part of the problem arises because people who administrate Wikipedia are, naturally, those who edit it a lot. It is difficult sometimes for those who edit a lot to realise how little those who just read Wikipedia may edit under their account (remember that registering an account helps with various settings for reading preferences). Equally, such readers may do lots of editing "anonymously" from an IP address. Thus it is not uncommon to find accounts with few or no edits, even over several years. And it is not uncommon for new editors to take several years to do more than a few edits here and there. We should not be looking suspiciously at accounts that only edit sporadically. Any suspicion of sockpuppeting needs more evidence than this. Does anyone agree? Is it worth doing an essay on sleeper socks (what they are and how to identify them and what to be careful about) and sporadically editing accounts? Carcharoth (talk) 16:05, 10 February 2008 (UTC)

I think an essay describing the differences between a POV-pushing new account and a new account; Accounts here to edit a single topical area (SPA) and accounts here to edit a single article (SPA); and link to edit-countitis (addressing infrequent editting, would be a good addition to our project literature. But please do make a good shortcut for it, between WP:SNOW, WP:CLUE, and WP:SPA there are pages I've discovered that are wildly quoted out of context based on their shortcut. MBisanz ^talk 16:20, 10 February 2008 (UTC)

(ec) If you look at my early contribution history you will see sporadic editing, some of it not very good.[6] People who start on the wrong foot can often be coached to participate in a better way. There is no urgency to block suspected sock puppets. There is an urgency to welcome new users and explain to them how they can be successful Wikipedians. Those who are intent on malfeasance will provide actionable evidence soon enough. There are occasionally very obvious examples of socks, which nobody would question such as User:Earthboat, User:Earthenboat and User:Earthenwareboat, but even then it is useful to employ WP:RFCU in order to empty the sock drawer completely and block the underlying IP address. Shoot-from-the-hip sock hunting needs to stop before we lose any more of our experienced editors and admins (See Example 1 and Example 2). The algorithm is quite simple: 1/ Use WP:RFCU or WP:SSP to present your evidence, 2/ await independent confirmation, 3/ if a case is so sensitive it needs to be handled confidentially, then email your evidence to a checkuser or ArbCom and let them handle it. Jehochman ^Talk 16:24, 10 February 2008 (UTC)

I wonder if Wikipedia:Sporadic editing, with the shortcuts WP:SPORADIC and WP:SPORADICALS will catch on? :-) Carcharoth (talk) 16:29, 10 February 2008 (UTC)

I don't know, is the answer. We do see a lot of sleeper socks (for good examples of what they look like, see the many sockpuppets of User:Jon Awbrey), but yes, there are editors who only log in very rarely. Guy (Help!) 16:41, 10 February 2008 (UTC)

I have argued at length on related issues at the WP:SOCK page; I indicated that the climate of suspicion that surrounds new accounts that know what they are doing is counterproductive; and that people frequently change accounts the first few times they start contributing. (ArbCom subsequently cut the ground out from under me by effectively making policy in the Privatemusings RfArb, and I quit discussing that in dismay, but that's another matter.) What is certainly the case is that a lot of new accounts are returning or otherwise experienced users. When we see an account that is new, there are several alternatives: a former IP editor with "mature", wide-ranging interests; a relatively new editor who had a few initial blocks on his account and wants to start afresh having learnt the ropes; a permissible sockpuppet of an experienced current editor; a disruptive "bad hand" sockpuppet of a current editor; a block-evading editor; a banned user; a new editor with a focus on one area; a new editor with a focus on one article or set of articles. (The last two could be, in addition, of three types: currently non-disruptive; currently disruptive but with potential; currently disruptive but not worth the effort.)

Half our disagreements come about because of confusion of categories. I would strongly recommend setting up a page or a project that shares people's experiences with how these different accounts operate, and - most importantly - figuring out approximately what their relative numbers are. Relata refero (talk) 17:55, 10 February 2008 (UTC)

I'm not an admin, but I have some experience dealing with at least SPA sock-like behavior: the Killian Documents related articles, once hot topics, are somewhat stagnant these days, with "Citation Missing" tags left unaddressed for a long while, but as soon as I even propose making any changes, suddenly things get very, VERY busy, including the appearance of IP's and editors whose edit histories seem a bit curious in relation to the Killian articles. Apparently germane to what's being discussed here, check the edit history of IP 67.168.86.129 and then look at this curious comment made to UBeR. Someone who's been around since 2001 showing up as an IP with a very short edit history to work on just one article? What can you say about something like that? -BC aka Callmebc (talk) 19:06, 10 February 2008 (UTC)

I take it at face value and advise you to leave that article alone, given the trouble your involvement there has caused in the past. Guy (Help!) 20:11, 10 February 2008 (UTC)

For what it's worth, I probably fit the pattern of a "sleeper sock" - I first signed up in 2004, made one contribution, then promptly vanished until February 2007, when I returned as an SPA because it was brought to my attention that a number of articles on a cappella music, an interest of mine, were up for deletion. Heck, you could likely consider me a meatpuppet, since I'm pretty sure I heard about those AFDs on a mailing list. Since then, I hope I've become a reasonably productive contributor, but I think it's worth remembering that many of us probably did start out as SPAs and/or lurkers/sleepers. JavaTenor (talk) 06:08, 11 February 2008 (UTC)

The sporadic editor viewpoint

This is the reply made on her talk page by User:Amelia9mm when I asked for her insights into why some accounts only edit sporadically. It seems obvious, but I think many involved Wikipedians forget this:

Carcharoth, I suspect you'll find that most folks are "sporatic" because they haven't quite got the hang of the detail of editing (syntax-wise, etc.) and they have very busy lives outside of the internet.

The incident in which I was blocked along with a large group of others, is almost entirely comprised of academics (virtually all PhDs in a variety of fields teaching at major U.S. universities), most of whom are busy publishing their own research. These aren't youngsters with nothing else to do with their time, nor are they hard-core wiki-supporters who spend time here for the cause, they are just highly educated folks who have an interest in a particular topic or several and they come here just for that topic or those topics.

I don't think that is all that unusual for wiki users, is it? If they are made welcome, they might stay around and begin to contribute more substantially.

When I realized I was blocked, I was trying to update some information on the page about my hometown. No one had told me I was blocked nor had any (understandable) explanation been given. So, since I had to fumble around to figure out how to request an unblock and it was at first categorically denied, you can imagine why I might not have bothered to return. This sort of thing is probably true for many others. If you want busy intelligent folks to partake of the wiki and to contribute, they need to be made welcome, not blocked or banned in droves.
I'm still not even sure if I should be replying to this here or on your talk page or on the page you linked above [see, more confusion about how the wiki world works], or if you'll see this at all. I guess I'll wait a bit and see. :o)

For most new users, wikipedia and its "administration" and lingo [What is a sleepersock?] are like being in a maze with no idea what the maze looks like and no idea if it's worthwhile to bother finding out. Thus, many are just sporadic users who come looking for information on a topic and see they can contribute a bit, and do that and not much else.

— Amelia9mm 20:21, 10 February 2008 (UTC)

None of this is actually new, or non-obvious, but it seems pretty easy to forget in the hurly burly of Wikipedia administration. Do we need to try and remember this more often, or not? If so, how? Carcharoth (talk) 00:03, 11 February 2008 (UTC)

I do think this is important to remember, and it's great that you've brought it up, Carcharoth. A huge portion of the material in the 'pedia was contributed by sporadic editors, and we should value them. Unfortunately, we don't hear much from them outside of the mainspace, so it's easy to forget that there are viewpoints other than those common to us wikifanatics :-P Though we interact with the hardcore contributors most often, we should remember that people who are willing to devote hours a week to writing an encyclopedia are a tiny minority. If people learn that you have to be willing to devote a ton of time to the project to have your contributions valued, most of them likely won't bother to edit at all. Even though sleeper socks are undoubtedly a problem, treating large numbers of people that fit that "MO" as though they are sleepers is casting too wide of a net and will likely be more damaging to the project than the socks themselves could be. And that's to say nothing of the more general harm caused by blocking based on shoddy evidence. I think it would be great if we could come to a formal agreement that we should not be making blocks based on such weak evidence as a sporadic edit history. (not commenting on any specific blocks here, the particular examples may have been valid) delldot on a public computer talk 10:43, 11 February 2008 (UTC)

A case in point

Right, let's take a real-world example.

• MonaVie (edit | talk | history | protect | delete | links | watch | logs | views) (see also Wikipedia:Articles for deletion/MonaVie)
• JerryGuyB (talk · contribs · deleted contribs · page moves · block user · block log)
• Janos (talk · contribs · deleted contribs · page moves · block user · block log)
• Ulcer boy (talk · contribs · deleted contribs · page moves · block user · block log)
• Travis.M.Allen (talk · contribs · deleted contribs · page moves · block user · block log)
• Holannakata (talk · contribs · deleted contribs · page moves · block user · block log)

Discuss. Guy (Help!) 20:30, 10 February 2008 (UTC)

Am having a quick look. How long do I get? :-) Carcharoth (talk) 00:03, 11 February 2008 (UTC)

Long as you like. I'm quite keen to come up with a scalable and supportable mechanism for dealing with this kind of user. Guy (Help!) 08:16, 11 February 2008 (UTC)

Here's a peripherally related one: Sugardaddy (talk · contribs · deleted contribs · page moves · block user · block log). Looks like a Utah resident, most contribs are to GoYin (edit | talk | history | protect | delete | links | watch | logs | views), a Utah firm. The article was speedily nuked as spam per Wikipedia:Articles for deletion/GoYin. Maybe I should simply not care about this kind of thing; I think it's massively more common than when I first started sysopping, maybe by now it's so normal that one simply needn't worry about it. Perhaps that's where I'm going wrong. Guy (Help!) 10:05, 11 February 2008 (UTC)

For what it's worth, I've noticed the majority of this sort of editor simply go away after their favorite toy article is taken away from them. So the marginal benefit of dealing with them is likely not worth it, most of the time. Someguy1221 (talk) 10:08, 11 February 2008 (UTC)

AFD nomination was appropriate, as the sourcing in the article makes it a borderline call. SPAs in an AFD is no big deal, just tag the AFD contributions appropriately and keep moving. A multi-level marketing business is the sort of thing likely to get a few SPAs - for mostly the same reasons that any website with a forum will get a few SPAs if it receives an AFD nomination; they will almost definitely have a discussion forum sales reps can participate in, and somebody will mention the AFD. The AFD and article history don't reveal anything worth spending more admin time on, the one clear advertiser/spammer was blocked more than a week ago. GRBerry 21:37, 11 February 2008 (UTC)

Transwiki

Hello, I hope I am putting this in the right place. I would really appreciate these former articles be transwikied to the Lego Wikia because I can't access them. Unfortunently, I was unable to obtain the contents before they were deleted:

• Voya Nui
• Vahki
• Takanuva
• Krana
• Bionicle society
• Rahi (Bionicle)
• Metru Nui
• Bohrok
• Protodermis
• Kanohi
• Mata Nui (Great Spirit)
• Great Cataclysm
• Visorak
• Dark Hunters
• Bionicle timeline
• Barraki
• Kanoka
• Piraka
• Zamor
• Rhotuka
• Karzahni

Thanks. -AMK152^{(Talk • Contributions • Send message)} 22:31, 10 February 2008 (UTC)

We can temporarily undelete them here, but you'll need a sysop on the Lego Wikia to export the pages from here and then import them there. Mr.Z-man 00:48, 11 February 2008 (UTC)

Yup this is feasible. I'll put them in your user space if nobody objects? -- lucasbfr ^talk 10:06, 12 February 2008 (UTC)

Seems reasonable to me (I was pinged about this), did this happen already? ++Lar: t/c 16:29, 13 February 2008 (UTC)

User:Fairchoice, User:Archtransit, User:JzG, blah blah blah ...

Resolved

 – Fairchoice has been unblocked. MastCell ^Talk 18:46, 11 February 2008 (UTC)

Anyone familiar with the Archtransit situation will know that Fairchoice was indefblocked and then unblocked, discussed on Archtransit's RfC, all that good jazz. JzG has now reblocked claiming sockpuppetry. (log) Before this turns into a wheel war, this should be discussed.

I figure maybe we should discuss this user here, because discussing his fate at someone else's RfC is clearly out-of-process. - Revolving Bugbear 23:01, 10 February 2008 (UTC)

We need more details than that. — Rlevse • Talk • 23:07, 10 February 2008 (UTC)

I don't want to turn this thread into another referendum on Archtransit. The user was originally blocked as a vandalism-only account by JzG. Archtransit extracted a promise from the user to contribute constructively, and shortened the block to forty-eight hours. This has already been discussed and is not what concerns me.

The account was reblocked by JzG today, pointing to sockpuppetry and Archtransit's RfC. The section in question is this one, which several people have already pointed out is an out-of-process referendum.

The reason I can't give more is because there isn't more -- the user has no contributions for the last couple days, and I can't find any sockpuppetry case involving him. - Revolving Bugbear 23:16, 10 February 2008 (UTC)

A block summary linking to an RfC on a different user isn't a great idea. Review of Fairchoice's contributions shows no problem editing since his block was shortened by Archtransit and his talkpage has DYK credits. It seems Fairchoice hadn't edited for several days before JzG reblocked him. I can't see the relevance of the RfC to the further block (Fairtrade has not commented on it). That leaves the sockpuppetry accusation, which I think JzG needs to substantiate. WjBscribe 23:13, 10 February 2008 (UTC)

Going on the basis that getting Guy's personal thoughts on this would be the most straightforward course towards demisting this block, I have notified him of this thread. AGK (talk) 23:21, 10 February 2008 (UTC)

• Fairchoice is very obviously not a new user, and has behaved as a trolling single purpose account whose benefit to the project is negligible by comparison with the disruption he's caused; his unblock was problematic, and Archtransit has a history of problematic unblocks. But if others think they can rein him in and persuade him to make productive edits instead of constantly agitating for Expelled to be less honest, they are more than welcome to try. For myself, I have had enough of the obsession with extending near-indefinite good faith to people who are here only to pursue an agenda. Maybe if Fairchoice were to leave Expelled alone for six months and only edit other articles, in a neutral way rather than promoting an agenda, then I would change my mind about him. For now, I wash my hands of the affair. I'll be interested to see how Carcharoth's proposal for dealing with this kind of user shapes up. Guy (Help!) 08:12, 11 February 2008 (UTC)

We don't assume people are abusing socks just because they appear to have experience from their first edits. We have WP:RTV, we have users who edit for months or even longer as IPs, and we have tacit rules allowing people to create separate accounts for privacy concerns, and for a very good reason, we don't ask them to disclose them unless it appears they are doing something inappropriate with them. - Revolving Bugbear 17:12, 11 February 2008 (UTC)

Yes, all of those things are true. We also have quite a few sockpuppets, meatpuppets, and agenda accounts created in response to off-wiki canvassing, particularly on intelligent-design-related topics. The question of how to balance the need to welcome new users and the need to protect controversial areas of the encyclopedia from endless cycles of disruption by single-purpose agenda accounts is a thorny one. MastCell ^Talk 18:39, 11 February 2008 (UTC)

Regarding this particular user, he's been unblocked by User:Mike Rosoft. There are certainly some... oddities... here ([7], [8], etc), but I suppose there are plenty of eyes on the situation now. MastCell ^Talk 18:45, 11 February 2008 (UTC)

I agree with you on all points, MastCell. For now, the user does not seem to be editing disruptively, but of course if he starts being disruptive again, I have no problem with a reblock. - Revolving Bugbear 18:51, 11 February 2008 (UTC)

Edit request

Can someone please make the edit I requested at template talk:main. Richard001 (talk) 23:48, 10 February 2008 (UTC)

What edit? I don't see any recent edits on that talk page. עוד מישהו Od Mishehu 06:03, 11 February 2008 (UTC)

Richard made a request in August, and no one has yet implemented it or replied to it. I'm not familiar with sub-articles, so I'll leave it to someone who is.-gadfium 08:26, 11 February 2008 (UTC)

Subarticles seem to be the same sort of thing, though that template is being considered for deletion (and I have argued it should be deleted, see the discussion). Probably best to leave it for now and see what happens, and change it to {{summary in}} if its deleted. If you don't know what edit I mean, just search for my signature on the page. Richard001 (talk) 06:48, 13 February 2008 (UTC)

Admin request on Commons

Can an admin on Commons please block User:Rappingwonders2? Checkuser revealed User:Rappingwonders is a likely sock of User:98E (who in turn, is a sock of User:Yung6 on Commons). This user has a total disregard for copyright and after being blocked on Wikipedia, Rappingwonders2 continues to upload copyvios on Commons and edit wars over the same templates and images as Yung6. Spellcast (talk) 08:57, 11 February 2008 (UTC)

Wrong site, try Commons. John Reaves 08:59, 11 February 2008 (UTC)

It would be good if as well as the "other languages" interwikis on meta pages like this one, we could also have "other projects" (so this page would have interwiki links to the Commons/Wikibooks/Wiktionary/Meta etc admin noticeboards). Is that technically possible? Neıl ☎ 11:44, 11 February 2008 (UTC)

User(s) blocked. there by Boricuaeddie - for future reference, try COM:AN/B, it's pretty well monitored. ~ Riana ⁂ 15:03, 12 February 2008 (UTC)

See also Commons:COM:RFCU and in particular commons:Commons:Requests_for_checkuser/Case/Rappingwonders2 I missed the link from there to here, (there was one, right? :) ) so... sorry for the delay. Should be sorted now. ++Lar: t/c 18:21, 13 February 2008 (UTC)

Question about Twinkle use

I have never posted a question here so please excuse me if I do it incorrectly I went to an andministrator about an editor's use of Twinkle. I would appreciate some input on this from others.

Here is what I wrote to the administrator and his response:

Is Twinkle allowed to be used for reverting and/or edit wars? I don't have Twinkle but it's being used for this. The latest place was at [9] and this was at least the second time TW was used for this revert. I don't have Twinkle so I don't know if this is right or wrong do I am bringing it to the attention of an administrator, you! :) I thought it was supposed to be used for policy things not things like this. I appreciate your time, --CrohnieGal^Talk 17:59, 9 February 2008 (UTC)

The tool isn't the problem so much as the behavior. Reverting isn't the way to make progress. Jehochman ^Talk 18:41, 9 February 2008 (UTC)

Twinkle has been used multiple times in this manner so my question; is twinkle to be used for reverts? It doesn't seem fair to those of us who do not use it for it to be used in the manner above but I will leave it to all of you to explain or do what ever you feel is appropriate. I will inform both of these editors about this comment. Thanks, --CrohnieGal^Talk 11:57, 11 February 2008 (UTC)

As the editor uses proper, hand-written edit summaries I think it doesn't make any difference whether he uses Twinkle, Undo or manual revert. Kusma (talk) 12:00, 11 February 2008 (UTC)

TW is just a tool, and not the only one available for the job. As with any tool, it can be misused. What TW does, can all be done manually, so it does not make a difference, just makes it easier/faster. Look at the behavior, not the tool. If it is used for vandalism, or unwarranted/unexplained reverts, then the user(s) should be warned, and should they get out of hand, reported. -- Alexf⁴² 13:05, 11 February 2008 (UTC)

I feel as do others that using twinkle gives the user an advantage that those of us don't have. My reading and understanding of twinkle says that it's not to be used for edit changes nor edit war. Am I incorrect on the reading I did about twinkle? Please reread my links. Thanks, --CrohnieGal^Talk 16:44, 11 February 2008 (UTC)

I think your question has been previously answered above. - Philippe | Talk 16:48, 11 February 2008 (UTC)

I'd like to know what advantage Twinkle gives. Making a revert manually seems trivial. --Hyperbole (talk) 17:55, 11 February 2008 (UTC)

Edit warring is a bad thing, so having an unfair advantage in edit warring isn't really a good thing for the person with the advantage. Mr.Z-man 19:44, 11 February 2008 (UTC)

Agree with Kusma and AlexF. I stopped following the Quackwatch article awhile ago, way too contentious, but ... it's not the tool, it's how you use it. :-) regards, Jim Butler (t) 21:02, 11 February 2008 (UTC)

Ok thank you, them will you answer Jericho's question? --CrohnieGal^Talk 17:00, 11 February 2008 (UTC)

http://en.wikipedia.org/w/index.php?title=User_talk:Levine2112&diff=next&oldid=171475643 Levine2112 has been warned before. QuackGuru (talk) 19:40, 11 February 2008 (UTC)

Just like with any tool, when the user abuses it, they can be blocked like any other user, or the tool can simply be removed from the users monobook. But having said that, this tool can be used for whatever the user wants it to be used for, including reverting text from an article that he/she disagrees with, but must always keep in mind WP:3RR and Wikipedia:Edit warring.Tiptoety ^talk 00:22, 12 February 2008 (UTC)

I would like to attest that I have seen this user misuse this particular tool on a number of occasions, in particular in actions against my edits. Also, please note the following page: Wikipedia:Administrators' noticeboard/Levine2112. ScienceApologist (talk) 00:32, 12 February 2008 (UTC)

Oh, well it looks like they violated the terms, maybe a block is due, along with removing the twinkle script from this users monobook. Tiptoety ^talk 00:41, 12 February 2008 (UTC)My bad

Robotics again

We just had a very encouraging conversation about this subject here, but unfortunately, I'm feeling very beat-up and tired at the moment and I need to come back and ask for your help...even though I am not disputing what any one person has done, and this is usually the place for disputes. Please trust that I'm trying to make this conversation just as short as I can, and trying not to whine. You won't understand the argument without context, even though this is not the place for context and Wikipedia is not the place for passion. I have put a lot of time into trying to build up grass-roots community of students, hobbyists, experts and users into a working community of people who take care of each other by solving the tough issues involved in getting home and yard robotics to work right. I find that it's like herding cats. When I discovered how great wiki-values work, I thought I had the answer...and it seemed to me that this suited the purposes of Wikipedia every bit as much, because without these people, your robotics coverage is going to come principally from academics and people who have something to sell...no matter what policies are in place to discourage "spam". You just can't make a silk purse out of a sow's ear, and if we don't find some way to make househusbands and experimenters and students and all the other users of consumer robotics "buy into" the process of creating an encyclopedia, then you're going to continue to have coverage which is created by, well, other people. No offense intended, and there are many great robotics articles here, but if I understand wiki-values correctly, I'm pretty sure that what you've got here so far is not close to what you ideally want. But, after putting a whole lot of effort into this, I don't think at the moment that I should invite anyone else to come edit robotics articles on Wikipedia, I think I should just try to write articles myself (which kind of grates, because that puts me in the position of having to constantly defend against the perception that I think it's "all about me". This is the opposite of wiki-values). I think it's fair to say that I've been taking the lead in reverting spam links in and improving Robot, which is the number one target for robotics spam-linkers. I created my second article last night discussing a particular robot, have only created two articles on Wikipedia about any particular robot...first Lawnbott, which was discussed in that link above, and then last night the Spyder (lawnmower), and it was speedy'd and then deleted as spam, even though it had a reliable source and an explanation that more were coming (and it turned out I jumped the gun, it will be at least another week before I can get a second reliable source, so I'm totally okay with it being deleted, as it turns out). I don't want to go into specifics, I don't have any problem with the speedy, and I'm really not pushing any one product. But then I get this:

Satyr has pretty much covered everything I was going to say, Dank notability, notability, notability. These articles are promoting non-notable products, so I feel that the spam speedy applies. No offense, and without intending to "bite the noobs", I would say that the Wikiproject sounds like its entire purpose is to create articles destined for rejection. Not every clever new concept and interesting project is notable; most of them are bound for the dustbin of history. I would advise a robotics project to switch to improving the existing articles, and not creating all these articles about products that have not even been released. —Preceding unsigned comment added by Orangemike (talk • contribs) 13:58, 11 February 2008 (UTC)

Now, prove me wrong, please. What are the odds that someone who isn't really familiar with Wikipedia, who does not spend two months fighting spam, and who just wants to contribute their personal knowledge to robotics articles won't quickly get their articles deleted and their edits reverted, and how long do you think it will be before they leave? I'm not talking about academics who want to write the definitive article on task planning software...they are quite welcome here. With the result that you have an encyclopedia full of articles about how robotic vacuum cleaners, say, might theoretically find their way around a home, without any users actually having such useful appliances, because the people who are actually getting this stuff to work, and would be interested in sharing their knowledge, by talking about specific consumer robots that are actually available, and strategies to make them work, feel like they run up against a brick wall when they come to Wikipedia. I'm not threatening to take my ball and go home, I'm saying that when I invite someone to leave their protected little community and come over here and particpate in ours, if they feel beat up by the process, even if that's just because they didn't know what they were doing, then they're not going to give me a second chance. I have to be practical. - Dan Dank55 (talk) 16:21, 11 February 2008 (UTC)

Oops, Orangemike actually responded in our Wikiproject 2 hours before I wrote this...sorry about that, I didn't see it. He does make some eloquent points, but I think it's stuff everyone here already knows ... you can copy it here if you like, Orangemike. He also says "I hate to tell you this, especially if you have the evangelical fervor of 'I actually want to see people benefiting from robots in the home...we desperately need them before all the baby boomers retire, and the developing world needs them to help provide food and power' crippling your objectivity." Funny, I don't feel crippled. I think we all agree that a very Buddhist approach to passion works best on Wikipedia. I also think we all acknowledge that none of us would be here if we had passion about absolutely nothing. - Dan Dank55 (talk) 20:07, 11 February 2008 (UTC)

Dan - I think the place you were looking for is WP:RFC, but here's fine :)

"Sharing knowledge" is a totally worthy goal. Having people around who know stuff is extremely useful and desirable! However, that's not what Wikipedia is exactly about... not exactly. Wikipedia is meant to be a tertiary source of information. See WP:PSTS for a full explanation, but in this case what that means is:

1. We don't publish original research, as you probably already know. If you come up with a radical new design for a robot's motor functions, for example, you can't publish that new design here.
2. We don't (usually) publish secondary information - when that new design is published, we don't do reviews or critiques of the motor functions.

I haven't looked through the WikiProject, but I suspect OrangeMike is correct (he often is :) ). Notability is the key. You might be more interested in pointing the primary users you're talking about towards WikiHow, or maybe posting reviews at WikiNews? Or even creating your own wiki at WetPaint or JotSpot? Just some thoughts... -- SatyrTN (talk / contribs) 20:19, 11 February 2008 (UTC)

I want everyone to know that Orangemike and Satyr are both 100% accurate (Orangemike has left me another note) and also eloquent. I think Angela and Jimbo wouldn't be that happy about the references to those other sites (even though wikiHow is indeed great), they're pouring a lot of time and money into Wikia for just these purposes. And that's exactly the problem...I know people at Wikia, I'm comfortable there, I created an attractive site and invited a bunch of people to join...and no one showed up. Not one. Then I found out that I could entice them to participate at Wikipedia...but, seriously, after today's conversation, knowing something about what they need, this isn't going to do it. I do understand what the limitations here are, and it's not good enough. They would have to have some sense that people who do in fact have some idea about what is "notable" and what isn't in robotics...which online reviewers can be trusted and which ones can't, which machines are surviving and which aren't, which ones are selling and which aren't...would be making the decisions about what stays or goes. I do understand the limitations here...you rely on very smart volunteers who have to apply the same rules to everyone, and generally do a fantastic job of that. It's just that, when we can have this whole conversation and the question of which sources can be trusted and which can't is not even relevant, when there's simply a presumption that all our articles in the future are going to be discarded and we're never going to be able to prove our case...well, you might want to go back and refer to the previous discussion in Archive 125. Robotics community can successfully intersect here, but my uninformed guess is that it will never be comfortable here, at least not in the way that I was hoping, which is a shame. - Dan Dank55 (talk) 20:54, 11 February 2008 (UTC)

Sigh. On top of everything else, Jameson and I don't seem to be able to work together, and I've resigned from Wikiproject Robotics. When it rains it pours.

I need to say just a few more things and then go, sorry this is running on. I did understand the rules. I do agree with Orangemike that it wouldn't be easy to accommodate some of the articles that I and other roboticists might like to write. When Orangemike first speedied me within seconds, when I had a valid source and the robot was (I believe) notable, it didn't bother me at all...I assumed I had done something wrong. The more he talked, the more it sounded...and sounds...like a policy decision, that the most popular robots are less notable than, say, any episode of Three's Company. Now, personally, I didn't care what the rules were, I just wanted to know what they were. But I want to work mainly with user communities for popular robots. I had hoped that one thing I would be good for was to help people get up to speed on Wikipedia, but clearly, I'm not the person to offer that service.

I think you guys are great, and it's been a pleasure. - Dan Dank55 (talk) 04:55, 12 February 2008 (UTC)

I want to thank Orangemike and Jameson for being so supportive today. I have agreed to write boring, factual articles here and to start a blog as an outlet for my passion, and I also want to say that having passion in no way gives me permission to make accusations, and I'm sorry about that. - Dan Dank55 (talk) 20:03, 12 February 2008 (UTC)

Broken process management on particular article

Moved to Wikipedia:Administrators' noticeboard/Corey Delaney discussion

Blocking procedures

I was recently blocked - in my opinion, without justification - and it's really killed my enthusiasm for this project. Let me just spell out a couple ways the blocking process seems completely broken:

1. I was never informed that there was a discussion on AN/I. The user who I had a content dispute with obnoxiously dropped a warning template on my page after we had reverted each other twice[10], and apparently the admins at AN/I decided that sufficed as a warning. I disagree. Warnings should always be issued by third parties to a dispute - preferably, admins - and should always include a notification of any AN/I discussion.

2. After a discussion between several admins, where most concurred that I was not edit warring and my edits did not merit a block[11], another admin simply blocked me unilaterally without contributing to the discussion at all. Another admin questioned his block; the blocking admin didn't really defend it, but just said "if you feel an unblock is in order I would appreciate a note indicating your findings." [12] It's a shame that the procedure for blocking someone is to wait to see if anyone wants to hit the block button; discussion and consensus are meaningless.

3. Naturally, I placed an unblock template on my page. A reviewing admin declared that he did not like the block and did not consider it justified, but took no action.[13] Other admins declined to review my request, and one actually deactivated the request, because a glance at the talk page made it appear as though it were being handled.[14]. What kind of appeals process is this? What's the point of addressing an unblock template if you have no intention of undoing another admin's actions?

4. Nobody even bothered to address the behavior of the complaining user, who was making large reverts that consisted mostly of long quotations and marking them "minor edits"([15]), who came on to my talk page to taunt me about getting me blocked ([16]), who specially pleaded to an admin for a longer block, ([17]) who when casually warned by an admin that he was also edit warring, was extremely aggressive and sarcastic in his reply ([18]), and whose talk page actually begins with what amounts to a proud assertion that he is a POV warrior ([19]).

Guys, a block is a black mark on any Wikipedian's permanent record - they shouldn't be handled this casually. Without even addressing whether anyone thinks I'm a good editor who should stay on the project, let me just say that handling blocks this way will drive off good editors who should stay on the project. --Hyperbole (talk) 17:48, 11 February 2008 (UTC)

Interesting thoughts. I would like to remark on a few of your point:

• Point 1. Editors who are the subject of a discussion at any noticeboard are generally notified—both to maximise the value of the discussion, by allowing it to encompass all angles and experiences, and as simple common courtesy. I wouldn't say this is applicable to the theory you are suggesting (that the blocking system is broken), although I do believe that it is rather unusual for your block to be made on the basis of an AN thread, without prior notification. Having said that, it does not appear you have presented any evidence to contradict the viewpoints and conclusions related to the circumstances, both that of the participants in the discussions, and the blocking administrator. Furthermore, blocks are made to prevent damage to the project—or, more accurately, to put an end to, and prevent further, damage. For that reason alone, the primary consideration in the blocking process should be the encyclopedia as a whole, as oppose to an editor's image—although, I grant you, it is only considerate to take it into account where possible. It may well be that the blocking administrator viewed the matter as urgent, and therefore issued the block to prevent further damage, with a mind to following it up after the damage was no longer imminent.
• Point 3. I have found the appeals process to be very efficient, being a participant in it myself. There are several mediums available to blocked editors—they may, as you obviously already know, use the {{Unblock}} template. Alternatively, they may email the Unblock-en-l mailing list for review. Finally, they may directly contact an administrator of their choice, or, for certain blocks, the Arbitration Committee. I don't think you have yet fully experienced the process to such an extent that sweeping generalisations as to its (in)effectiveness can be issued, as I am sure you will understand.

It seems that you have had a pretty bad time with your block. It may be beneficial here to open up the circumstances surrounding the block for general discussion—perhaps in this section of the Noticeboard. AGK (talk) 18:16, 11 February 2008 (UTC)

Is there any point in discussing the merits of a block after it's already run its course? At any rate, if you feel like reviewing it, please use your own judgment: I was blocked for these edits ([20]) , which in my opinion would be considered constructive by any reasonable editor and which are not a revert similar to this one ([21]). --Hyperbole (talk) 18:26, 11 February 2008 (UTC)

You make a very convincing case. I lend you my support. The block was improper and your record should be expunged. The blocking admin should be advised that this block was improper and review the blocking guidelines. Bstone (talk) 19:13, 11 February 2008 (UTC)

As far as expunging records - that may be a nice idea, but - we don't generally do that. The software doesn't provide for the capability - the problem, at the core, is that we're using the same thing (a "block log") both as an audit trail for admin actions and as a "permanent record" for a user. Block logs are only altered or removed in highly exceptional circumstances. —Random832 19:20, 11 February 2008 (UTC)

I realize. Still, the block was inappropriate and should be expunged, with apologies. Bstone (talk) 19:39, 11 February 2008 (UTC)

An apology/note in the block log could be done, but as far as I know, a block log has only been totally expunged (removed from the database) once. Mr.Z-man 19:48, 11 February 2008 (UTC)

Were you edit warring? PouponOnToast (talk) 19:51, 11 February 2008 (UTC)

Twice, I reverted some edits I considered inappropriate and asked for discussion on talk. When that failed, I made a partial revert and partial rewrite, and asked for discussion on talk. When all that failed, I rewrote the lede and two controversial sections simply to bring them in concert with WP:STYLE. As to whether that meets the definition of edit warring, whether such editing is harmful to Wikipedia, and whether such editing merits a block, well, please draw your own conclusion. --Hyperbole (talk) 20:04, 11 February 2008 (UTC)

Hopefully you've learned your lession and will adhere to 1rr like I now do. PouponOnToast (talk) 20:09, 11 February 2008 (UTC)

Yep, my advice would also be to consider WP:1RR, after following this I've been block free, and shockingly enough been given more buttons.--Addhoc (talk) 20:14, 11 February 2008 (UTC)

Frankly, I don't see much merit in the philosophy that if problems with an article are corrected quickly, that deserves a block, but if they're corrected on a slower timescale, that's fine. It's a real shame that, instead of actually determining whether an editor is making a good faith effort to work toward consensus, admins just count up edits on their fingers and hit the block button based on that and that alone. --Hyperbole (talk) 20:17, 11 February 2008 (UTC)

All right. I'm off the project.

Per m:Right to Vanish, I'm asking that my user page and talk page be deleted and protected/salted (as I'd prefer not to have anyone leave me nastygrams after I'm gone). Once this is done, I'll change my password to something I'll never remember and be done with this place. If this is the wrong place to make this request, please let me know. Wikipedia: rewarding volunteer editors with endless antagonism since 2001. --Hyperbole (talk) 20:47, 11 February 2008 (UTC)

Another excellent editor leaves the project due to poor admin actions. When will it end? Bstone (talk) 20:55, 11 February 2008 (UTC)

Unfortunately, it is something we need to live with. This isn't the first, nor will be the last one. While he had some pretty good points, we are all human and bound to fail. Like it or not, this is repeated several times per week, most times without fuss at all. -- ReyBrujo (talk) 21:06, 11 February 2008 (UTC)

Pages deleted and indefinitely protected, as requested. —Wknight94 (talk) 21:02, 11 February 2008 (UTC)

Since when do we delete talk pages? Per "User and talk pages, and their subpages, and other non-article pages that no others have substantively contributed to and whose existence does not impact the project, may be courtesy blanked or deleted." Emphasis mine...RxS (talk) 21:09, 11 February 2008 (UTC)

Really? We haven't messed with this guy enough? —Wknight94 (talk) 21:13, 11 February 2008 (UTC)

I'm not very familiar with him, but I don't see how not deleting a talk page is messing with him. Nothing personal with him...RxS (talk) 21:21, 11 February 2008 (UTC)

It's messing with him because he specifically asked "that my user page and talk page be deleted and protected/salted". So why not? —Wknight94 (talk) 21:23, 11 February 2008 (UTC)

(edit conflict) You may want to take a look at WP:VANISH rather than just the meta page. It's a courtesy that we extend to users in good standing that wish to exercise their right to leave the project, nothing more. I've seen it done many times in the past, and I don't see a compelling reason not to leave it deleted since the user requested it. Ioeth _{(talk contribs friendly)} 21:24, 11 February 2008 (UTC)

User talk pages are a useful record of past events. Also.. I think we encourage people to use dramatic "do what I say or I'll leave" type tactics by doing whatever they say when they claim to be leaving. I think this drama is best not encouraged. Friday (talk) 21:25, 11 February 2008 (UTC)

You mean we encourage tragic goodbyes? Yeah, after all, everyone has the right to leave the way they want. -- ReyBrujo (talk) 21:59, 11 February 2008 (UTC)

I did look at WP:VANISH, up until 2 weeks ago it was a redirect to the meta page. It was recreated at that point, but if you look at the talk page there was very little consensus for that change. Even at that point it was "in some circumstances" and certainly not a right. Deletion of talk pages has had no discussion and what little there was concluded that it was a bad idea. RxS (talk) 21:38, 11 February 2008 (UTC)

The guy is driven off the project by a questionable block - can't we have some decorum and grant his final wish as he walks out the door? How about we trump the WP:VANISH vs. m:RTV nitpicking with WP:IAR? —Wknight94 (talk) 21:46, 11 February 2008 (UTC)

I agree that, since there won't be any "further consequences" about his leaving (like blocking the administrator who blocked him, or a request for comments, or anything else), deleting the talk page is fine. If he returns, everything can be restored. And remember to add him to Wikipedia:Missing Wikipedians. -- ReyBrujo (talk) 21:56, 11 February 2008 (UTC)

Well who knows, in this case maybe not. But in many cases an editor leaves under some sort of controversial circumstances (not necessarily of their own making) and it's a bad idea to put related discussions out of the reach of the vast majority of editors. Diffs needed in RFC's, Arbcom cases, policy debates etc all disappear. RxS (talk) 22:06, 11 February 2008 (UTC)

For what it's worth, even on meta, they don't seem to be saying "go ahead and delete user talk". User talk pages generally have significant contributions from other editors. Individual cases may be fairly unimportant but it'd be good to get folks onto the same page with this issue, as it does come up frequently. Friday (talk) 21:41, 11 February 2008 (UTC)

<outdent> What about a restore, blank, retire message, and protect? RBRP. (reburp?) Solves it, and accomplishes what Hyperbole wants (which I assume is no new messages). Keeper | 76 | Disclaimer 21:43, 11 February 2008 (UTC)

Why are you assuming when he was so crystal clear in what he wanted? Deletion. —Wknight94 (talk) 21:48, 11 February 2008 (UTC)

Because that's what he said on Friday's talk page: "I'd prefer that my user/talk pages don't turn into shrines where editors I've had disputes with over the last two or three years can post celebratory messages as though my exhaustion with this project represents some kind of ideological victory for them." [22] --Kbdank71 21:53, 11 February 2008 (UTC)

For users exercising RTV, I would also prefer archiving and full-protection of their pages instead of deletion, but I don't know what the community consensus is on the issue. There's a discussion at Wikipedia talk:Right to vanish#Meta. — Satori Son 22:15, 11 February 2008 (UTC)

<outdent>To Wknight: Sorry, I wasn't clear, I had a RL moment there and hit save instead of exit. What I meant was "Hyperbole has stated that he is leaving the project". "Hyperbole doesn't want any new messages on his talkpage". If he/she is gone, it isn't up to him/her to decide how the project continues, no? Longstanding consensus is that talkpages are not owned by the user. It isn't his/hers. It belongs to Wikipedia and should be accessible (please note, I'm not saying editable, simply accessible), to those other than admins. Period. I didn't know this was up for debate. Unless there are BLP concerns that need oversight, there is no reason to make this page only available to admins by deleting it. User:Hyperbole did explicitly say delete my talkpage, but I don't believe he/she understands what that means or that he/she doesn't necessarily get to decide. His/her next sentence anyway is as I'd prefer not to have anyone leave me nastygrams after I'm gone" This is accomplished with blanking/templating with "RETIRED", and full protecting. Besides that, he/she said, in essence, I'm leaving. How many editors, every week, say "I'm leaving"? It's drama. He may not be back, and that would be a great loss to the project. I actually sympathize with how he/she was unequivocally treated by admins when he/she came here for help in good faith. We (admins) dropped the ball by doling out a one-sided semi-solutioin to his legitimate attempts to find consensus and he ended up coming keyboard-to-keyboard with a bully that is notorious for pulling the NPOV card whilst never really utilizing the AGF card. All that to say, however, the talkpage should still exist. Reburp it. (RBRP=restore, blank, retire, and protect from further editing). Serves the purpose that Hyperbole asked and preserves the history for anyone outside adminship. Keeper | 76 | Disclaimer 22:52, 11 February 2008 (UTC)

please stay

Hyperbole will be missed.

You will be missed. I enjoyed reading your comments on the talk page and appreciated your NPOV editing. May your keystrokes grace Wikipedia again. You have a way with words. Please, don't leave. Take a break and come back refreshed! QuackGuru (talk) 21:05, 11 February 2008 (UTC)

It's understandable that even a short block can put one in a negative mood about Wikipedia. I don't know about the editing dispute that precipated the block. But Hyperbole has made many valuable edits over the past two and a half years and I hope he'll return refreshed. ·:· Will Beback ·:· 23:39, 11 February 2008 (UTC)

While I don't agree that the talkpage should be deleted for transparency and historical reasons, (see above), I do believe that Hyperbole is a fine editor and was mistreated at the Chiropractic article and mistreated when he/she brought it here to admins' attention. I hope he/she will reconsider and continue to add valuable information to this project. Keeper | 76 | Disclaimer 00:12, 12 February 2008 (UTC)

I, too, hope for your return. Bstone (talk) 07:12, 12 February 2008 (UTC)

Sorry if this question is in the wrong place, but how long was the original block for? Could no-one have explained to the User to wait it out, and then take up the idea of a null-edit or whatever to update their block-log if the undeserved slur was going to cause them to leave the project? Newbyguesses - Talk 01:36, 13 February 2008 (UTC)

deletion reason script

I have updated Mediawiki:Sysop.js with my script to automatically select a reason from the list at Mediawiki:deletereason-dropdown. To update your cached copy, go to this exact url: http://en.wikipedia.org/w/index.php?title=MediaWiki%3ASysop.js&action=raw&ctype=text/javascript and click shift-F5 (internet explorer) or ctrl-shift-R (firefox). —Random832 18:15, 11 February 2008 (UTC)

That's great news. Good work, Random! ;-) AGK (talk) 18:18, 11 February 2008 (UTC)

Note that the list of regexes that it uses for finding deletion reasons is not 100% complete, if anyone finds a tag that it doesn't seem to pick up on correctly, contact me and i'll work on adding it in the next version. —Random832 19:33, 11 February 2008 (UTC)

Indefinitely blocking vanished users

On the "Right to vanish" talk page on Meta there's been discussion about indefinitely blocking and removing all user rights from users who exercise their right to vanish. It seems like a perfectly logical step to take. The right to vanish is a serious thing that should entail serious consequences. The discussion is located here.

I'm thinking that we should adopt a standard practice when someone exercises their right to vanish on en.wiki that includes an indefinite block (including e-mail) and having any user rights removed by a steward. Thoughts? (Feel free to comment below or on Meta.) --MZMcBride (talk) 00:26, 12 February 2008 (UTC)

Strongly oppose. How is this a problem? Admins have left and come back. There's no need to kick them out the door on the way out. Corvus cornix_talk 00:28, 12 February 2008 (UTC)

This is, I presume, only for users who specifically indicate their intention to permanently vanish, and request deletion of their user page, and won't be applied willy-nilly to users who simply haven't been heard from in a while but haven't expressed any intentions regarding the future of their account? *Dan T.* (talk) 00:30, 12 February 2008 (UTC)

Only upon request (regardless of what our blocking policy says) could be considered part of the right to leave. But not when someone just leaves. The same way some choose to leave with wikidrama, others may choose to return in silence. If someone really, really wants to leave forever, he would delete the email preference and choose a long random password which, by all means, is the same. -- ReyBrujo (talk) 00:33, 12 February 2008 (UTC)

Yes, sorry if there's been confusion. This is not for a {{retired}} template applied in a fit of anger or anything like that (inactivity, etc.). This is for the users who have their user talk pages and user pages deleted and have made a conscious decision to split permanently from the project. --MZMcBride (talk) 00:34, 12 February 2008 (UTC)

I would only support this as long as the user makes i very clear that they wish to vanish, and that they understand their account will be indef. blocked and will have all user rights removed. I do not see what harm this can do. Tiptoety ^talk 00:45, 12 February 2008 (UTC)

I agree with this. Vanishing an account is no trivial thing; it should be done only when someone truly wants to vanish forever. If someone's gone (or at least left their account behind) forever, then there should be no issue with a block. And if they don't really intend on being gone forever, then they shouldn't be invoking the right to vanish. -Amarkov moo! 00:48, 12 February 2008 (UTC)

Amarkov, you made my point for me. нмŵוτнτ 01:16, 12 February 2008 (UTC)

I agree, and it sounds great on paper, but I think it happens way too often that vanished users come back. I'm unsure of the need for this, unless it is simply to emphasize that vanishing is a serious thing. I guess this would be ok, given that the user could still log in and request unblock on the talk page, but I personally, I'd have to think about this.. -- Ned Scott 01:20, 12 February 2008 (UTC)

Vanished users should not be coming back under the same account. Period. The process of invoking the right to vanish involves destroying many records of bad behavior; how can we go back and undo all the edits replacing the username with "Former user X"? Vanishing and coming back with the same name looks far too much like a free user history wipe, and that is not good. -Amarkov moo! 01:26, 12 February 2008 (UTC)

(e/c)I would agree with this as well. I think part of the reason is so that if they come back, they don't just start editing again, they'd be fully un-vanished first, to avoid people using RTV as a way to hide something by only vanishing temporarily. Mr.Z-man 01:29, 12 February 2008 (UTC)

Where's the difference between a vanished user returning under his old username (with the bad behaviour records deleted) and a vanished user returning under a new username (with the bad behaviour records deleted)? The latter makes it even harder to spot any previous wrongdoing, actually. We would only make one of the two impossible if we'd block the vanished user, unless we treat such users as banned and block their new accounts as well. And I doubt anyone is trying to propose that. --Conti|✉ 02:26, 12 February 2008 (UTC)

I would definitely support this. As Amarkov says, RTV should only be implemented when a user really is leaving....for good, so I see no reason not to block the account, and remove any user rights. - Rjd0060 (talk) 01:22, 12 February 2008 (UTC)

I don't see any need for or value to taking this step, at least in the absence of very unusual circumstances such as a user vanishing in lieu of an impending indefinite ban for serious case. Otherwise, I see this as a solution in search of a problem, and a deterrent to once-valued contributors, having become temporarily disenchanted with Wikipedia but then changing their minds, returning to us. Newyorkbrad (talk) 01:26, 12 February 2008 (UTC)

Giving people the option for a permanent RTV block could be a good thing, if they want to really cut the cord. If that happens, I can see also deleted their talk page and protecting vs. recreation. Gone with the option of coming back, vs. gone and gone for Good with a big G. Lawrence § t/e 01:28, 12 February 2008 (UTC)

The problem with this is that there is no necessary correllation between how strongly a user believes that he or she wants to cut all ties to Wikipedia at the moment he or she is upset enough to vanish, and whether he or she might want to come back a few days or weeks or months later. People want to come and go for all sorts of reasons, both real-life-based and wiki-related, and if we eliminated from the ranks of Wikipedians everyone who at one point or another announced that he or she was leaving forever, we would be without the services of many, many decided contributors and administrators. Unless the "vanished" user had been a serious problem before departing, I don't see why we would want to add even slightly to the disincentives that face a departed user who is thinking about returning. Newyorkbrad (talk) 01:31, 12 February 2008 (UTC)

Though they can always request a unblock. Tiptoety ^talk 01:29, 12 February 2008 (UTC)

People should not request RTV if there is any possibility that they will be returning. -Rjd0060 (talk) 01:36, 12 February 2008 (UTC)

Many unhappy users simply don't always think that way. People change their minds. We shouldn't discourage people from coming back unless there's a real reason to. Newyorkbrad (talk) 01:40, 12 February 2008 (UTC)

Reading through the RTV page on meta, suggests it to be a permanent solution, as it should be. There are other options aside from vanishing. Users need to weigh them, and if they do decide RTV is the best way to go, that should be the end of it. - Rjd0060 (talk) 01:44, 12 February 2008 (UTC)

Like I said above, there is a real reason to. Vanishing completely destroys records of user history, and that is not good if the user isn't really gone. -Amarkov moo! 01:47, 12 February 2008 (UTC)

Nothing is permanently lost. If a user un-vanishes, the records are easy enough to restore. --Carnildo (talk) 03:29, 12 February 2008 (UTC)

NYB has it exactly right. There are plenty of irritating editors, there are a fair number who vanish, there may even be some who irritatingly vacillate between vanishing and appearing. If there are too many in the last category (which I doubt), send some of them my way and I'll vanish them and resuscitate them as requested (as long as it's merely a matter of bog-standard deletion and undeletion). I'll even welcome them back with personal messages, not tedious boilerplate. And I expect that I won't be alone in making such an offer. Meanwhile, no need to turn up the menacing tone and add to the drama. -- Hoary (talk) 08:30, 12 February 2008 (UTC)

I don't think I like the idea of indefinitely blocking departed users, but I had always assumed that the removal of userrights was standard practice. An unattended account with rights is infinitely more dangerous than a fresh vandal account. I could very quickly do a lot of damage with a compromised bot account; more still with an admin account. A compromised bureaucrat, oversight, checkuser or steward account would, of course, be disastrous. Even something like rollbacker or autoconfirmed is potentially more dangerous than a fresh account; and if the editor was well known in the community, their edits are more likely to pass unnoticed in recentchanges or watchlists than an unknown new account. An unattended account is much more likely to be compromised than a used and monitored account. And if an editor does decide to return, his or her rights can easily be restored if they left in good standing. So if removal of rights isn't already standard practice, it certainly should be. Happy‑melon 10:06, 12 February 2008 (UTC)

Many administrators or other users with higher-level user rights often give up their extra abilities upon resignation from the project, however, there has never been any solid rule that enforces all those with special rights to resign them at once should they decide to leave. I don't particularly see how an unused account has a higher probability to become compromised in any form than one that is actively used, and accounts that have been compromised have often been detected quite quickly, as prior situations have demonstrated. Spebi 10:15, 12 February 2008 (UTC)

Perhaps I should rephrase: of course the chances of someone guessing/cracking the password of an unattended account are no different to that of an account in use. But an unattended account, once cracked, can be used indefinitely by an invader, until it is realised that it has been usurped. The most common reaction to an admin account being compromised is, apparently, to edit the mainpage to a large and usually obscene image: this is a fairly good indication that the account has been compromised, and the time for them to be desysopped is usually extremely short (I've heard 3 minutes). I imagine, although I doubt it's happened, that a compromised bureaucrat account would be used first to desysop Jimbo or similar. This crude and attention-grabbing use of compromised accounts naturally results in quick identification and blocking. There are much more malicious uses a compromised account could be put to if it is possible to have some preparation time. You might be able to put a penis on the mainpage for twenty seconds with a crude edit, but if you took the time to bury it somewhere deep in the transclusion structure it could be five minutes before anyone worked out how to get rid of it. Every time I put my mind to this question I come up with more effective ways to damage the site with a stolen admin or bot account: I can think of ways to irreperably remove all external links from all pages, to slow the loading of 95% of articles to a crawl, or to place a penis at the top of all our featured articles. But with time to prepare, to make a number of edits which don't appear to be nasty until you hit the one edit that drops the lewd image, you can do more insidious damage. The point is that an active account will notice these edits: if you look at your contributions and see something you don't remember doing, and don't understand why, you would get suspicious. If there's no one legitimatley using the account, that's not going to happen.

In fact, it's fairly immaterial whether the compromised account is used or unused by its legitimate owner. We all know that accounts with userrights can potentially be dangerous: that's why we have RfA, RfB, RfBA, etc. If the owner of the account has left, they are not going to be using their userrights for the benefit of the project, so it makes absolute sense for them to be reset. Why have more potentially dangerous accounts lying around than are genuinely necessary? I'm not saying that, once removed, those rights should not be restored if the editor returns - as I said above, if the editor left in good standing there is no reason why they should have to do anything more than ask. But leaving admin/bot/crat accounts lying around when we all know their potential for misuse strikes me as an unnecessary weakness. Take them away automatically as part of RTV, give them back automatically if they return. If their account is compromised in the meantime, they'll have to stick to ordinary vandalism. Happy‑melon 16:04, 12 February 2008 (UTC)

Wow, that was longer than I expected :D! Happy‑melon 16:05, 12 February 2008 (UTC)

It was pretty on the button, though. I would concur that any inactive account with a good pedigree - with extra tools or not - is a prime target. The good faith shown toward a returning editor of some standing may allow some of the less obvious malicious edits to survive much longer. LessHeard vanU (talk) 22:32, 12 February 2008 (UTC)

I'd have to disagree, strongly. We have admins who "leave" the project only to come back within days. Some repeatedly. We have rules against admins blocking themselves or other users for "wikibreaks" - why are we wanting to block a user for a "rtv", which is but the ultimate "wikibreak"? Why is this, of all possible admin actions and consequences, to be the one thing you don't get to change your mind on? Yes, it's a serious act, but it's wholly in the realm of that users concern, not the projects. Achromatic (talk) 01:21, 13 February 2008 (UTC)

• The point in all this that seems to be being missed is that 'right to vanish' is serious. When people use it to make a dramatic, pointy, disruptive exit, they are abusing the right to vanish, and that abuse is doubled if they return later. Please, if you want to leave the door open for a future return, blank your pages and use the {{retired}} template, or leave some other sort of message. There seems to be a need to leave a 'message' by having your user and/or talk page links in signatures turn red. Instead of asking for pages (especially user talk pages!) to be deleted, you can leave dramatic messages and departure essays on your user page, but using 'right to vanish' as a standard departure method is wrong. Most departures can be handled other ways, and there need to be good reasons for exercising 'right to vanish'. The main one being that you really do want to vanish, or you need to disassociate from your real name. The only way, regrettably, to discourage frivolous use of the 'right to vanish' is to make much clearer that if you do return, that everything that was done to enable the vanishing will be undone, except the removal of user rights. Carcharoth (talk) 08:59, 13 February 2008 (UTC)

Angie Y. (talk · contribs · deleted contribs · logs · filter log · block user · block log)

Hello, I am having trouble with User:Angie Y.. Despite many warnings she has continued adding POV and original research to articles. The outcome of her RFC was that she should be mentored. I have been doing that, but it is not working. She ignores warnings. She has also been uncivil at times. Diffs are below, all less than one month old.

[23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45]

I don't know what to do now. Mentoring isn't working. I'd like the opinions of others. Thanks. JetLover (talk) (Report a mistake) 00:51, 12 February 2008 (UTC)

Based upon those diffs (particularly the last couple), I would support a short duration block for incivility - but I think a final warning would be in order first. - Philippe | Talk 00:59, 12 February 2008 (UTC)

ScienceApologist: Trolling or COI On Quackwatch

Resolved

 – I guess the moral of the story is "If you're trying to tell a joke, make sure it's actually funny". --BETA 01:30, 13 February 2008 (UTC)

It is my belief that this diff [46](in the edit summary), is an exercise in very poor judgment on ScienceApologist's part. It probably constitutes either an obvious conflict of interest, or blatant trolling, intended to bait users to list it on COI/N, and disrupt things yet again. But that's just my take on things.

This appears to be yet another battle of an edit war on Quackwatch, so perhaps we should also discuss intervention options for this article as well. It is difficult to decide what category this falls under, so I must defer to your collective opinions on this matter. --BETA 05:12, 12 February 2008 (UTC)

You can't just point to one diff and shout COI, trolling, or whatnot; you claims appear completely unjustified. If you don't like the edit, talk to SA or post on the talk page. I truly don't understand why you came running here, there's no actual edit war going on at the moment. Someguy1221 (talk) 05:19, 12 February 2008 (UTC)

The slow edit war has never stopped. http://en.wikipedia.org/w/index.php?title=Quackwatch&curid=879168&diff=190848338&oldid=190829788 Read the edit summary: qw is for true believers, frauds, cranks and dishonest intellectuals. . . The edit summary is flame bait. --QuackGuru (talk) 07:28, 12 February 2008 (UTC)

Ok, the edit summary in the diff that BETA provides is perfectly kosher. The diff that you provide isn't even from SA. It's not easy to judge the merits of a content dispute when you're not providing a very clear picture of it. Someguy1221 (talk) 07:46, 12 February 2008 (UTC)

The edit by ScienceApologist looks a good one to me. If you don't like it, say why on the article's talk page. -- Hoary (talk) 07:40, 12 February 2008 (UTC)

The group of edits in question seem OK to me also, though there does seem to be a edit war is process about a specialized box that looks like but is not an a template--presumably not a template so it can avoid challenge at TfD. DGG (talk) 13:48, 12 February 2008 (UTC)

Just to provide some context for SA's edit summary: there was a period when the now-banned User:Ilena was active when there was an intense effort to "out" specific editors of the Quackwatch article. In one case, this took the form of claiming that an editor was a relative of Stephen Barrett (Quackwatch's principal). This claim was never proven, and I believe perhaps definitely disproven. I believe SA's comment about "my uncle Barrett" is a sarcastic reference to this prior period of attempted outing. MastCell ^Talk 19:43, 12 February 2008 (UTC)

Ok, but it was still poor judgment. We're building an encyclopedia here. If Nicholas Carr was so openly flip about an article he was editing for Britannica, he'd be out on his ass so fast his bookmarks would spin. --BETA 20:16, 12 February 2008 (UTC)

If Nicholas Carr had to deal with the rich pageantry that is Wikipedia at present, he'd probably be a bit more openly flippant from time to time. MastCell ^Talk 20:26, 12 February 2008 (UTC)

Argumentum ad populum?... Nice... I'll see your argument, and raise you a "two wrongs don't...". --BETA 23:53, 12 February 2008 (UTC)

It's not an argument. More an observation that humor is a universally employed defense mechanism. Are you really suggesting that ScienceApologist should be blocked for that edit summary, or are we having a philosophical discussion at this point? MastCell ^Talk 00:03, 13 February 2008 (UTC)

For those interested, according to the aforementioned banned user (and several others), User:Ronz was Stephen Barrett and I was Daniel Barrett (his son). So heck, SA must be my cousin. We call him Jim-Bob-Billy-Day-Mercedes in real-life, or at least for the amusement of editors-who-should-know-but-don't-better. Shot info (talk) 00:22, 13 February 2008 (UTC)

Alright, I give in. I never actually considered he was trying to be humorous. I thought it was either trolling or conflict of interest. --BETA 01:30, 13 February 2008 (UTC)

Logging of community imposed editing restrictions

The section at Wikipedia:Editing restrictions#Placed by the Wikipedia community looks rather empty. It's a good idea to log all new community imposed editing restrictions here for awareness purposes. I've added the one I can remember, there are bound to be others. Please add them. Thanks. MER-C 08:20, 12 February 2008 (UTC)

VolkovBot damaging templates

The VolkovBot appears to have a bug which is damaging templates.

The bot adds interwiki links to pages, but doesn't use <noinclude></noinclude> tags around it's additions to templates, which in many cases breaks the template. The bot works ok when there are already interwiki links which have <noinclude></noinclude> tags, as it just modifies that text, but if it is inserting the first interwiki links, it fails to use <noinclude></noinclude> tags around the new text. Some example diffs are: [47] [48] [49] [50] [51] [52] [53] ... etc.

I think the bot needs to be blocked from editing templates asap until it is fixed.

--Ozhiker (talk) 14:02, 12 February 2008 (UTC)

I don't think it is that serious, it made one template edit in the last 5 days. Since someone already contacted the bot owner in his Russian page, I think he is already warned. I may suggest him to stop editing templates, but right now they are not that many, can be fixed manually if necessary, and the amount of good edits greatly outnumbers the number of buggy edits. -- ReyBrujo (talk) 14:49, 12 February 2008 (UTC)

Openhousenewyork

Can someone take a look at the recent edits there and the comments by me, 68.167.244.226 and Openhousenewyork on my talk page. I think there might be a copyvio and a coi but I'm too out of it right now to think straight. --ÐeadΣyeДrrow (Talk | Contribs) 16:22, 12 February 2008 (UTC)

How's what I wrote on your talk page? עוד מישהו Od Mishehu 16:42, 12 February 2008 (UTC)

Thanks, that's what I wanted to say, but the sentences weren't coming together. --ÐeadΣyeДrrow (Talk | Contribs) 16:47, 12 February 2008 (UTC)

hmmm?

I've just had my Wikipedia user page vandalized again by a so-called Sadie from freenode (the IRC network.) This is the 4/5th time that this has happened now. http://en.wikipedia.org/w/index.php?title=User%3AEth01&diff=190975851&oldid=190669752 Eth01 (talk) 20:47, 12 February 2008 (UTC)

And I'd like Wikipedia administrators to...? ➔ REDVEЯS has changed his plea to guilty 21:16, 12 February 2008 (UTC)

if possible to take appropriate action? Eth01 (talk) 22:16, 12 February 2008 (UTC)

which would be...? ➔ REDVEЯS has changed his plea to guilty 22:27, 12 February 2008 (UTC)

Redvers, perhaps at least try to be helpful. Majorly (talk) 22:40, 12 February 2008 (UTC)

I suppose Eth01 may be requesting sprotection, or a blocking of the ip, but I doubt any admin is going to do that given that this is a single occurence. LessHeard vanU (talk) 22:42, 12 February 2008 (UTC)

(ec)Majorly, I remain confused as to what the problem is and what the editor would like done about it. Asking questions should help clarify the request into something that administrators might be able to help with. ➔ REDVEЯS has changed his plea to guilty 22:45, 12 February 2008 (UTC)

I've semi protected the user page. If this was hideously wrong, then feel free to unprotect. Addhoc (talk) 22:49, 12 February 2008 (UTC)

LessHeard vanU, it's happened more than once. Addhoc, that's perfectly fine with me. Let's see what happens... :) Eth01 (talk) 22:57, 12 February 2008 (UTC)

<Semi-protection when the last questionable edit was more than 24 hours ago and the one before that was three weeks ago? Well, it's an unusual interpretation of policy, but I suppose it's up to you, Addhoc. ➔ REDVEЯS has changed his plea to guilty 22:59, 12 February 2008 (UTC)

Redvers, please don't confuse this anymore than what it already is. I'm assuming Addhoc applied semi-protection, because he deemed it appropriate - NOT YOU! Eth01 (talk) 23:04, 12 February 2008 (UTC)

Don't know how to revert this

Resolved

Yo mamma seems to have gotten some unusual content. I don't know how to revert it.--293.xx.xxx.xx (talk) 00:00, 13 February 2008 (UTC)

All fixed.—Ryūlóng (竜龙) 00:02, 13 February 2008 (UTC)

Improper RfC

This discussion has been moved to Wikipedia:Administrators' noticeboard/Mantanmoreland.

Protection seems to be gone on Irish phonology and its on the main page 13 feb 2008

Resolved

 – move protected per policy. See WP:MPFAP - Alison ^❤ 17:26, 13 February 2008 (UTC)

Protection seems to be gone on Irish phonology and its on the main page today --Sf (talk) 00:48, 13 February 2008 (UTC)

Its standard practice to have the TFA move-protected only. (and it does not appear to have any previous protections) Mr.Z-man 01:50, 13 February 2008 (UTC)

AIV backlog

Resolved

 – Backlog cleared. MastCell ^Talk 17:21, 13 February 2008 (UTC)

Hi, just wanted to give someone a heads up that AIV has quite a large backlog. Thanks. :) Somno (talk) 04:10, 13 February 2008 (UTC)

As of now, looks like it's down to 2 active reports... MastCell ^Talk 04:15, 13 February 2008 (UTC)

And now it's cleared, yay! :) Somno (talk) 04:33, 13 February 2008 (UTC)

Admin category woes

Amusingly enough (and related to the Rouge Admin category mess on AN/I), I've walked the Administrator category and removed the numerous false {{user admin}}, {{administrator}} and category inclusions. 24, only two of which are leftovers from former admins as far as I can tell. Naughty, naughty. — Coren ^(talk) 04:30, 13 February 2008 (UTC)

I'm surprized it was so few. I would have expected at least a few socks of banned users....--Rocksanddirt (talk) 05:12, 13 February 2008 (UTC)

Is this proper? (page recreation rather than move)

Resolved

 – Pages have been rejuggled correctly.

The content of Harry Kelly was cut and pasted into the brand new page Harry Kelly (politician), and the former page turned into a disambiguation page. That means the page history of the content at Harry Kelly (politician) is still in the history at Harry Kelly, without connection between the two. I *think* this needs fixing, but a) I'm not really sure of the policy, and b) I'm not sure how to actually go about "fixing" this.

To be clear, I'm *not* in any way objecting to the existance of the disambiguation, just wondering about the divorce between content and history. Andrew Jameson (talk) 11:02, 13 February 2008 (UTC)

It was dealt with by PeaceNT. Yes, this is completely against policy. For GFDL reasons we need the edit history and cut and paste moves are not allowed. If you notice one please list it at the Cut and paste move repair holding pen. Thanks for letting us know. Woody (talk) 11:21, 13 February 2008 (UTC)

Thanks to the both of you; I think this is resolved. Andrew Jameson (talk) 14:51, 13 February 2008 (UTC)

Major CAT:CSD backlog

Resolved

 – backlog now cleared

CAT:CSD is severely backlogged - and being populated faster than cleared. I'd like more admons to help me clear it. עוד מישהו Od Mishehu 12:49, 13 February 2008 (UTC)

Working on it. Tks. Alexf⁴² 12:58, 13 February 2008 (UTC)

Ditto. Neıl ☎ 13:04, 13 February 2008 (UTC)

"Admon?" Is today "Talk Like a Reggae Musician Day?" (Man, I have to get a new wall calendar... I keep missing these holidays.) I'll pop in there for a bit today. Caknuck (talk) 14:35, 13 February 2008 (UTC)

Due to a lot of hard work, it's looking a lot better, with the exception of the perenial beast of Category:Images with the same name on Wikimedia Commons. Pedro :  Chat  15:21, 13 February 2008 (UTC)

Wikipedia:Requests for arbitration/Matthew Hoffman

This arbitration case has been closed and the final decision is available at the link above. The Arbitration Committee finds that the blocks on User:MatthewHoffman were unjustified. It also states that Vanished user's adminship will be waived at this time. Vanished user may regain his sysop access by application to the Committee, upon demonstration of six months editing in compliance with communal norms and conduct standards. If regained, he will then be placed on parole with regard to both conduct and admin tool use for a further period of six months. For the arbitration committee, Cbrown1023 talk 13:50, 13 February 2008 (UTC)

Commons admins: deletion backlog

People who are also admins on Commons, please help with the backlog to October at http://commons.wikimedia.org/wiki/Commons:Deletion_requests/Older_Discussions -- SEWilco (talk) 16:15, 13 February 2008 (UTC)

Overlink crisis

Does anyone know if Wikipedia:Overlink crisis really is a crisis? --Kralizec! (talk) 19:57, 13 February 2008 (UTC)

Asking on wikitech-l is likely to get an answer from the devs about whether all the links are causing problems. (Personally, I doubt it, but I don't know.) On another note, I'm fighting a strong urge to tag that page with {{wikify}}... --ais523 20:17, 13 February 2008 (UTC)

Sounds a load of nonsense to me, using exagerated suggestions of navboxes with 500 links in them, it also seems to suggest that if that same box was included on 500 pages that would be 500*500 links stored on disk somewhere. Since the box is likely transcluded it in actuality is just 500. It's also very little to do with links a typical link might take an extra 4 characters over the raw text [[]] so the issue in the case of 500 links would be the real need for the 500, the additional 4 per link would be a fraction of the space used. I could of course be missing the point, clarity doesn't seem to be a factor in the page. --81.104.39.63 (talk) 20:35, 13 February 2008 (UTC)

TfD to be closed

Resolved

 – closed

Please would an admin close Templates_for_deletion#Template:Great_Western_Main_Line. The nominator has withdrawn the nomination. - Neparis (talk) 20:30, 13 February 2008 (UTC)