Wikipedia:Administrators' noticeboard/Δ

This a centralized page of all discussion pertaining to User:Betacommand and his bot User:BetacommandBot. These discussions were previously on the Administrators' Incidents Noticeboard, but haved been moved here in the interests of space and navigation. This page is for participating and discussing in a civil manner all matters that relate to Betacommand and his bot.

Please do not use this page to attack these users instead of participating in civil discussion.
This is not a complaints department. If you have a private matter that concerns operation of BetacommandBot, please discuss it with Betacommand on his talk page.

All BetacommandBot looks for is the name of the article or a redirect to that page. That is required by WP:NFCC#10c. It doesn't need to be a wikilink, just the name of the article. It doesn't need a template-style rationale, it doesn't need the "article=" parameter, all it needs is to meet one basic part of our non-free policy.

—Betacommand's description of BetacommandBot

Proposal: Release the code, or shut off the bot

There is no excuse in a free encyclopedia to have a secretive, closed-code bot causing deletion of multiple perfectly valid images that have been a part of the project for many years. I have much less of a problem with this bot if Betacommand would do two things: (1) Open up his code to community scrutiny. Failure to do so demonstrates a clear distrust of the community that is completely incompatible with this type of project. (2) Stop with the incivility toward complaints. Some are from new users, who simply don't understand the sub-section of a sub-section that BCB is using to tag for deletion. Some are from more established users overwhelmed by the massive talkpage spam geerated by the tagging of older images that they'd uploaded long before.

If these two conditions are met, I think the drama generated by BCBot's tagging would be minimized exponentially. Bellwether ^B_C 00:40, 26 February 2008 (UTC)

• Another call to block. Yawn. --Hammersoft (talk) 00:59, 26 February 2008 (UTC)

• If you have nothing to contribute, don't. I've seen your equally incivil comments on BC's talkpage. You're part of the problem, not the solution. Bellwether ^B_C 01:05, 26 February 2008 (UTC)
• And for the record, I'm calling for the release of the code, and a change in BC's (and his supporters, such as your) demeanor. Bellwether ^B_C 01:07, 26 February 2008 (UTC)

• And for the record, I find this umpteenth call to block the bot to be tedious, repetitive, and unlikely to yield anything positive. I'm sorry you disagree with my tone, but I'm calling a spade a spade and find no reason to not call it that. --Hammersoft (talk) 13:37, 26 February 2008 (UTC)

• Your tone is "on the record" in many places, defending BCBot--no matter the issue, no matter the problems--often in BITE-y ways. Bellwether ^B_C 15:07, 26 February 2008 (UTC)

• Factually incorrect, as I have specifically voiced disagreement with some of Betacommand's actions before. --Hammersoft (talk) 16:07, 26 February 2008 (UTC)

• Please step away from the corpse. Seriously, you want code that can run at 35000 edits an hour opened up to all and sundry? Anyone who would be confident enough with such code to modify it should be able to write their own version of BCBot and have it approved. Black Kite 01:10, 26 February 2008 (UTC)

• It's not a "dead horse." There are many people who feel that on a free project, code for such active bots should be open to criticism and improvement. Such secrecy, combined with the incivility present in the communications of BC and his supporters, does not improve this project. Bellwether ^B_C 01:14, 26 February 2008 (UTC)

• Whilst BC is occasionally incivil, that aimed towards BC is many times worse. Try reading the archives of his talk pages sometime, or try enforcing WP:NFCC on articles. You'll soon get the idea. Seriously. Black Kite 01:16, 26 February 2008 (UTC)
  • I've heard this claim many times, but to date the evidence has been pretty thin. Can we have some diffs where BetaCommand receives abuse over and above that which is normal for an administrator doing CSD patrol? I know he no longer has the bit, but the parrallel is obvious. - 152.91.9.144 (talk) 01:23, 26 February 2008 (UTC)

• I still don't get the "The bot can do lots and lots of edits a minute and therefore shouldn't be made open source" argument. As I've found out above, normal editors can't make that many edits anyways, whether they program a bot to do it or not, and it's highly unlikely that an admin would suddenly start running the bot (without any kind of approval) once the source code would be out in the open. So.. I honestly don't see the harm if the code that can run at 35000 edits an hour is opened up to all and sundry. --Conti|✉ 02:18, 26 February 2008 (UTC)

• Conti, I contacted some MediaWiki devs shortly after that was posted. there is no edit throttle on users. β^_command 02:21, 26 February 2008 (UTC)

• Hmm, I stand corrected, then. Maybe the devs should think about that, tho, an editing limit for non-bots sounds quite reasonable to me. Anyhow, I still think it should be rather easy to program a bot that would do lots of edits, so if someone with enough malice would be around, he would've already programmed and used that bot. I doubt someone will suddenly turn up and use that bot to do any harm if the source code would be released. I therefore still don't really see this as a reason not to reveal the code to the community. --Conti|✉ 02:28, 26 February 2008 (UTC)

• Look, the editing rate in a bot is a number in the code. There's no super-secret trick to making a bot edit 35,000 times per minute. Anyone can take a bot with open source code, like any of the Pywikipedia bots, and crank up its editing rate if they want. Then they'll get blocked. There is nothing magic about BCB's ability to edit quickly, so that is a pointless argument for why its code should be hidden. rspeer / ɹəədsɹ 06:23, 26 February 2008 (UTC)

• For the love of all things: This is getting ridiculous. - Rjd0060 (talk) 01:13, 26 February 2008 (UTC)
  • Per WP:BOT he must inform us what framework/language he is using. Since the bot was approved as a Pywikipedia bot he either a) must answer my question above or b) request approval again. In addition to that, it is not possible for non-admins or non-bots to run it at that speed yet Betacommand now shows that he doesn't even trust OTRS let alone the admins. And we are suppopsed to respect the wishes of a user who acts like this? No thank you. He wasn't desysopped for nothing and he needs to either listen or leave. EconomicsGuy (talk) 01:19, 26 February 2008 (UTC)

I don't think that is your choice; to decide who has to leave because you deem them "not listening". He is listening, as obvious from his edits to this page. It is no secret what the bot does, as it is approved for every function it serves. - Rjd0060 (talk) 01:42, 26 February 2008 (UTC)

Also, since this has been brought up a couple times, out of context. BC did not jab OTRS or any admins. What he said was that he doesn't trust OTRS to decide who is trustworthy to receive the code for the bot. His failure to trust all admins is completely legitimate. An example of why that distrust is justified was given by him: Archtransit. The community makes mistakes. BC wants to avoid this mistake from being made. Lara❤Love 02:53, 26 February 2008 (UTC)

• He doesn't even need to release the code, just inform other editors in a sensibly located place, what it actually does, beyond the innacurate and mis-leading statement 'the bot enforces policy'. MickMacNee (talk) 01:22, 26 February 2008 (UTC)
  • let me restate myself for the 100th time ALL BETACOMMANDBOT LOOKS FOR IS THE NAME OF THE ARTICLE OR A REDIRECT TO THAT PAGE. THAT IS REQUIRED BY WP:NFCC#10c IT DOESNT NEED TO BE A WIKI LINK JUST THE NAME OF THE ARTICLE. IT DOES NOT NEED A TEMPLATE STYLE RATIONALE, IT DOES NOT NEED THE ARTICLE= PARAMETER ALL IT NEEDS IS TO MEET ONE BASIC PART OF OUR NON-FREE POLICY β^_command 02:31, 26 February 2008 (UTC)
    • • Betacommand basically just doesn't get it, he cleary thinks the above a sensible reply to the request; well, take this as yet another attack if you want, I know you will, but the above reply is just another example of what he thinks is being helpful or informative, blowing his stack on a talk page that will be archived and never seen again in a few weeks time, helping no-one who has to deal with his bot, or any third party who wants to save the mass-deletions his bot threatens. The sentence isn't even all that informative for the unninitiated who have not been following this whole sorry teeth-pulling epic, which is the real hoot, red bolded text or not. MickMacNee (talk) 12:49, 26 February 2008 (UTC)
      • MickMacNee, you asked what BCBot checks for, I used simple English to explain that. If you cannot understand English well Ill be happy to translate it into what ever language you do speak. If English is your native language, that means that you have not even bothered to read our non-free media policy. I am not sure how else I can say it any simpler. I made it big and red in hope that you might actualy read it, guess not, yet another example of where you ignore what I say, and attacking me. β^_command 15:17, 26 February 2008 (UTC)
        • What does providing the policy link have anything to do with giving a sensible reply to a request to state what the bot does? Unless you are saying everyone can deduce what it does from the policy, which is plainly not the case because the bot cannot read English as you point out, and hence cannot assess whether there is a separate fair-use rationale for each use of the item, presented in clear, plain language, and is relevant to each use, because as we all now by now bar some die-hards, for NFCC10c enforcement, the bot is just a link checker when all is said and done, barely even relevant to the spirit of the 10c statement. And again, you seem to think posting replies here is what I and others want, it is not, as stated repeatedly all over the place and in the original discussion page now obscured (although any communication is gladly received at this rate). What happens when this page gets archived and someone again asks here how the bot "enforces the policy" as is often stated, or someone reaches the bot page with the same issues I and others have had? Do you just carry on ignoring them, blowing your stack once in a while, because you can't see the basic logic in heading off enquiries at source? Seriously, why do you think it is not necessary to provide a stable link from the bot page to a toned down and informative, contextualised version of the red text above, with the quite important extra explanation of why the image description auto-page link(s) is/are not sufficient. — Preceding unsigned comment added by MickMacNee (talk • contribs) 03:23, 27 February 2008 (UTC)
        • A fair-use case for using non-free media must identify the article in which it is to appear. It follows that using the presence of the name of the article as a proxy for compliance is a useful automation step. I suspect that this is more difficult for some people to understand than for others, and perhaps some rewording of the BetacommandBot FAQ might be in order. --Anticipation of a New Lover's Arrival, The 03:30, 27 February 2008 (UTC)
          • It is difficult to understand when there is a rationale stated in clear plain language "for the film Foo", but it's tagged because it is actually supposed to be for the article Foo (film), I have corrected two of those just today, dating from 2006. But that is by the by, the point being stated by me in the last sentence is the fact that there is a reason why the often quoted defence on BCBtalk and the copydesk that presence of the article links at the bottom of the image description page are not sufficient to meet NFCC10c and the usage needs to be stated in effect twice on the same page; the reason behind this was finally explained by betacomand somewhere, but like I say, as it was just a talk page reply it is now not easily findable again, and I had never seen it explained before, nor posted as alluded to above in a better place since. MickMacNee (talk) 04:02, 27 February 2008 (UTC)
            • Actually I did notice that you edited Image:House of Wax poster.JPG and Image:Hostage poster.JPG, but they don't seem to have any fair-use justifications at all, for any article, so I tagged them appropriately. Without the bot's work, these instances of inappropriately used non-free material would not have been spotted. I don't think your edits, though well intended, helped. --Anticipation of a New Lover's Arrival, The 04:26, 27 February 2008 (UTC)
              • Actually you are both wrong. The correct approach is to upload a smaller size version and add a proper fur template with appropriate parameters. Both the size issue and lack of a proper fur template can be detected by a script. The size issue can be automatically fixed and fur templates inserted by an assisted script at fairly high speed. The only reason I'm not simply doing this myself is that the last guy who automated this had 5000 edits indiscriminately reverted. This isn't rocket science, we know what the problems are (size, decorative use and missing fur) and we know how to fix them. For images used in articles with standard infoboxes this can be done at high speed. For other images we can do this semi-automated at fairly high speed. All that is holding this back is the unwillingness to allow anyone else to do automated edits on image pages. I notice that one prominent member of BAG has a "Death to fair use" message on his user page. How are we suppopsed to work constructively on this when this is the attitude we are being met with? Folks, this a collaboration - not a one-man or one-group show. Let's work together. EconomicsGuy (talk) 07:27, 27 February 2008 (UTC)
                • I happen to be human and I could write a fair use case for an image of whose provenance I was aware, where I could convince myself that such a case could be argued (I know this because I've done it). What you seem to be doing is confusing the planting of a template on an image page with the writing of a fair use justification. The former could be done by a computer, but as talented a programmer as I know many of my fellow Wikipedians to be, I don't think any of them would be capable of writing a computer program, to run at reasonable speed in Wiki conditions, and synthesize fair use justifications for random images uploaded by wikipedians. To take Image:House of Wax poster.JPG, how does this fit into the Foundation's policy? We have a "minimal" exception, "to illustrate historically significant events, to include identifying protected works such as logos, or to complement (within narrow limits) articles about copyrighted contemporary works." How does this poster complement the work, and is it of such importance that it merits use? Where, for instance, was this poster used? Did its imagery become an important part of the distribution or illustrate the theme? Who holds the copyright in the original picture? How was the current copy obtained, and does it contain any intermediate transformative material? Who designed the poster again? We cannot yet write programs to do this, at least not at reasonable speed, in wiki conditions. --Anticipation of a New Lover's Arrival, The 07:45, 27 February 2008 (UTC)
                  • The copyright holder is obviously the studio - this can all be retrieved from the infobox on the page where it is used. I'm talking about semi-automated fixing of these. As for the template vs a rationale issue we do have fur templates that we prefer people use with appropriate parameters. Betacommandbot does not do any work related to the rationales. Itt merely looks for a back link. For all we know there are thousands of non-compliant images out there even after BetacommandBot is done. If you want to clean this up a fair speed you need to allow the use of automated tools to help with the trivial bits such as resizing, back links and retrival of relevant fur parameters from infoboxes. How many fair use rationales do you know of that don't simply use those already in the fur template? Look at Image:Talk_Show.jpg. How do you want such rationales individualized to each image? The rationale is the same for all of them and the info easily retrieved. EconomicsGuy (talk) 07:57, 27 February 2008 (UTC)
                    • It may be possible for a human to deduce that the studio holds all copyrights in a poster (and usually he'd be right--but not always, see Dalek, for instance, for an example of a famous property whose image is not owned either by the programme maker or the distributor). A computer cannot make or defend such a deduction. It is not even capable of checking that the image in question is at all related to the work it is supposed to illustrate. Semi-automatic work won't do. If we're going to use somebody else's work, we're going to have to do some work of our own. --Anticipation of a New Lover's Arrival, The 08:14, 27 February 2008 (UTC)
                      • Why is it debates on Wikipedia always end up involving Doctor Who ;). More seriously though, suppose I write a plugin for AWB. You are presented with the page where the image is used and asked to confirm some simple facts retrieved from the infobox, alternatively correct the info. In addition to this the size is checked. This is easy to do. Like I said, semi-automated but still at fairly high speed in most cases. Sure, we aren't getting bot-like speeds but if enough people participate this would make the process more efficient. EconomicsGuy (talk) 08:25, 27 February 2008 (UTC)
                        • Perhaps you would be interested in FurMe. It is a javascript implementation that does practically exactly what you are saying with the exception of image size checking (I would actually think that you could get a bot approved that reduced the size of images that were tagged with a certain template, which could be added by FurMe, so that could probably be automated). - AWeenieMan (talk) 16:45, 27 February 2008 (UTC)
                          • Actually that looks like something we could get a fair number of additional folks to use. We really should give that a try, I'll check it out later tonight! EconomicsGuy (talk) 16:59, 27 February 2008 (UTC)
    • There's cognitive dissonance here: You won't release the super-skerit code, but that's all it does? Further, I think it's fairly clear that that's a hopelessly reductionist explanation. you could almost as easily have said "It edits wikipedia." - 152.91.9.144 (talk) 03:15, 26 February 2008 (UTC)
      • without getting into complex programming speak that 90% of the people here will not understand, its what the bot does for dummies. also do not edit my comments. β^_command 03:42, 26 February 2008 (UTC)
        • Oh please enlighten us. What super super sekrit algorithm is used to analyze the very few words on the image description pages? Why can't you just admit that your bot is a simple script and nothing more than that? Why does it mean so much to you that everyone has to think you are some sort of programming genius for having written a freakin' Mediawiki bot? This secrecy and belief that you have some sort of super special powers is ridiculous. We are grown-ups here, not kids. At least describe in pseudo-code what your bot really does and per WP:BOT what framework/language you are using. That is not too much to ask if you want people to trust you given your history with scripts prior to losing your bit. EconomicsGuy (talk) 10:02, 26 February 2008 (UTC)
          • It's becoming increasingly clear that this is a pride thing. He feels like he's smarter than the rest of us, and the we couldn't possibly understand (or even be trusted with) the truth about his bot's code. I have been distinctly reminded of a certain Col. Jessep in the responses of BC and his supporters. YOU CAN'T HANDLE THE TRUTH!!!! Bellwether ^B_C 15:17, 26 February 2008 (UTC)
            • Yet another personal attack ^^ β^_command 15:22, 26 February 2008 (UTC)
              • You have a very liberal definition of what constitutes a "personal attack" toward you, while you maintain a near-constant state of incivility toward anyone who dares question your actions in regards to your bot. The above was my observation of how you and your supporters have behaved in response to simple requests for your bot to be open-source on a free project. Nothing more, nothing less. Bellwether ^B_C 15:49, 26 February 2008 (UTC)
• Three times in one week oppose - They say that once you do something ten times it becomes a habit. We're on our way. Lara❤Love 02:58, 26 February 2008 (UTC)
  • Sorry, what exactly is it that you're opposing? (1) Open up his code? So far the only objection is that it's too fast for "normal" users. (2) Stop with the incivility? You're for more incivility? Or are you objecting to the straw man argument raised by Hammersoft, what this is another threat to block. In the same spirit of my asking several of the members of the "other camp" to step back, can I ask that you, LaraLove, take these pages off of your watchlist for a while? I personally am not finding your input helpful. It's growing increasingly repetative. For example, in the above comment you've "voted" without providing much to go on. - 152.91.9.144 (talk) 03:15, 26 February 2008 (UTC)
    • Oh, let me clarify. This specific oppose is to the specific proposal of this specific thread. And yes, you can ask that I take these pages off my watchlist, but no, I'm not going to. I'm only repetitive because I'm responding to the same things over and over and over again. If people would read more comments on the page than their own, they wouldn't have to keep asking the same questions and I and BC and others wouldn't have to keep making the same, repetitive comments. I didn't elaborate on my oppose because I just opposed a similar proposal a few inches up, which followed another similar proposal from like two days ago, maybe... and I think there was one a day or so before that... and one last week. Maybe I'm wrong... but it's something like that. Lara❤Love 03:36, 26 February 2008 (UTC)

• Comment If you want to have a policy that all bots are open source, then are ways for you to do that. Arbitarily deciding that one bot needs to be open source because you don't like it (or whatever) despite being unable to explicitly explain what the problem is however is not acceptable Nil Einne (talk) 10:29, 26 February 2008 (UTC)

The bot shouldn't be open source because it's disliked by people, it should be open source because, hopefully, this would reduce the drama by quite a bit.

1. People would stop asking what the bot does. And if they would, we could simply show them the code or tell them exactly what they want to know.
2. Wikipedians could improve the bot in many ways. (Hey, that's almost like this weird wiki thing, isn't it?)
3. The bot could, finally, be split up, so if one part messes up, we wouldn't have to block all other functions as well.
4. If the bot owner one day decides not to work on Wikipedia anymore, we wouldn't have to create a whole new bot for all these tasks.

Well, and it's kinda ironic that a bot that removes lots of non-free content is itself not free. But that's a minor point, of course. On the other hand, I see no reason whatsoever not to publish the code of the bot. --Conti|✉ 13:17, 26 February 2008 (UTC)

Quite simply, BetaCommand is under no obligation to release his code. He's already stated that he's not going to make it open source and he's not releasing it to anyone he doesn't explicitly trust. Calls to block the bot have failed repeatedly, several times in the past few days, so it's really a pointless issue to even discuss. By all means, please continue, as it's your right. I'm just pointing out how much of a complete waste of time it is to keep making the same requests over and over again. Lara❤Love 14:55, 26 February 2008 (UTC)

I, for one, am not trying to say "Release the code or be blocked!". Instead, "Releasing the code would be a really peachy thing to do, and there's really no reason not to, even though you are of course completely free not to do so anyways. Releasing the code would solve quite a number of problems, tho." is what I had in mind. --Conti|✉ 16:05, 26 February 2008 (UTC)

As a programmer, I can see why he is loathe to release his code to the public. I do, however, have faith in Betacommand's programming ability - I think it's other things that cause a lot of ire - for the very reason that it has tagged thousands of images correctly. x42bn6 Talk Mess 16:31, 26 February 2008 (UTC)

I don't really see why he wouldn't want to release his code. But that's probably because I am not a programmer. Could you explain what reasons there might be not to release your own code to the public? --Conti|✉ 16:37, 26 February 2008 (UTC)

Somewhere in one of these huge discussions, Betacommand said that the last time he wrote useful code and released its source, somebody modified it to be disruptive code.  I guess I can understand why he's not tremendously eager to do that again (especially nowadays, when so many people seem to be scrutinizing his every move).  For an exhaustive treatment of such issues, see here.    Xeriphas1994 (talk) 15:18, 27 February 2008 (UTC)

Not only that - imagine writing a book and then releasing it to the public domain. He may be using similar code in his other projects and he may not want the ideas copied. Put short, it's his code and it may well be in his own interests not to release the code. I don't feel he needs to release any algorithms or pseudocode either - personally, I think the bot does work that is quite easy to understand - go through a list of images, do some rule checking, and do stuff, then move on. x42bn6 Talk Mess 21:47, 28 February 2008 (UTC)

That's the usual argument to keep code to yourself, yes. But this is Wikipedia. Wikipedians are supposed to support free content, right? --Conti|✉ 22:27, 28 February 2008 (UTC)

Free mainspace content, of course, but certain parts of the MediaWiki code are closed source for the same reasons.  AFAICT the community's apprehension about "sorcerer's apprentice" bugs has created a sort of don't-ask-don't-tell attitude among many non-coding editors, especially w/r/t admin functions.  (This should obviously be taken with a mole of salt, since I've been reading the archived debates by subject area, not chronologically.)

This may be getting somewhat off topic because, even if a strong community mandate arose that all bots be open source, existing bots would probably be grandfathered for a while.    Xeriphas1994 (talk) 03:22, 29 February 2008 (UTC)

Thanks for the link, Xeriphas. I even !voted on that RFA, actually. I can see how the code of ProtectionBot might be used to vandalize, but I don't quite see that possible danger in BetacommandBot. I mean, I can kinda see Betacommand's point, but I still think the benefits greatly outweigh the possible drawbacks. --Conti|✉ 22:27, 28 February 2008 (UTC)

Conti, for those who need to need to see the code its open to them. Due to the nature of the language that I program, I dont have excitable(.exe) files My end user files are pure source code. I write my programs to get the job done I dont have safety features that a lot of other people build into their code. For those that I trust getting viewing access to my code is not an issue, (lar just got a copy). AGF does apply to wikipedia, however it does not apply to programming. My philosophic view is never assume any thing, verify facts, and prove your theorem before making a judgment. Until you prove your theorem of you being trustworthy, I dont trust you with the tools. (similar to being an admin) β^_command 02:24, 29 February 2008 (UTC)

Thanks for letting Lar see the code, Betacommand. I hope we have you with us for a good, long time, but it's good to know that if ever you were unable to continue devoting so much of your time to this excellent, valuable and necessary work, it would be possible for a suitably knowledgeable coder to continue maintaining and running the bot. --Anticipation of a New Lover's Arrival, The 14:04, 29 February 2008 (UTC)

Just want to second what Tony just said. Also, I did actually manage to get a bot to edit at extremely high speed on a test server. *Cough*multithreading*cough*. There's nothing wrong with having a bot that gets the job done with the simplest possible code so just want to apologize to Betacommand, especially per Tony's very well-worded comments on my talk page. That said, what is the general mood about getting the remaining images checked for other non-compliance issues like images not being sufficiently low resolution? EconomicsGuy (talk) 14:24, 29 February 2008 (UTC)

Personally, I would have no problem with a bot going through and fixing obvious cases of non-free use images that are of high resolution (album covers sounds like a good place to start to me). A few things to keep in mind though: 1) image size may not be the only factor to consider, as a higher resolution jpeg with high compression settings might not really need to be resized, for example; 2) I am sure there are cases where people give good rationales for having higher resolution images (perhaps they are discussing in specific one detail of the image); and 3) what is the maximum allowable resolution of a non-free image (or what should the standard be). Some non-free use templates set this mark at 300px x 300px (and completely ignore any other factors), but I never found a compelling consensus for this number.

I would think that having images tagged by human as being too large and fixed by bot would probably go over best. Perhaps you could get your bot to produce a list of images it would want to alter for review by humans. Or semi-automate it to review changes. Or perhaps there is a way to have the bot be smart enough to ignore high resolution images that are high rsolution for a reason.

I guess my question would be, how automated would you want your bot to be? - AWeenieMan (talk) 19:18, 29 February 2008 (UTC)

about the most a bot can do reliably without much error is list images over a certain dimension. other than that the risk of error becomes too great. β^_command 19:36, 29 February 2008 (UTC)

On image size, I would say that we discuss what is reasonable for most purposes. At nominal compression, JFIF (JPEG) files of around 20kb maximum combine fast download and optimal caching with good display characteristics. I've obtained excellent results from files as small as 3kb. Free images may be much larger, it doesn't matter because the Mediawiki software chooses a suitable conversion for the download context. But it may sometimes happen that a larger non-free image is justified (I can't think of such a circumstance off-hand, but there might be such). This is a matter for the image upload guidelines. --Anticipation of a New Lover's Arrival, The 19:39, 29 February 2008 (UTC)

First of all the bot I wrote should not be editing Wikipedia as I wrote it from scratch only to see how fast I could edit on a test wiki on a spare machine (yes I have no life - this is what I do for laughs). Far too unsafe for Wikipedia editing. Pywikipedia seems like the way to go (far safer) and I could have that working fairly soon. Is this issue covered by the deadline? In response to Betacommand I would think that the resizing was the most reliable task for the bot as the tagging would need to be done at least semi-automated to catch false positives as AWeenieMan describes. I'm open to ideas or a non-negotiable no before I spend too much time on the code. EconomicsGuy (talk) 19:51, 29 February 2008 (UTC)

Entirely serious question

Since BetaCommand felt free to write, and is allowed to run, a bot that flagged images for deletion under "NFCC 10 c"... would I be allowed to write and run a bot to revert its flaggings, under "Assume Good Faith"? Both are policy... I would definitely like some serious replies to this, thanks! --75.5.176.16 (talk) 07:11, 26 February 2008 (UTC)

you would need to file Bot Request for Approval and any such bot would not be approved. β^_command 15:27, 26 February 2008 (UTC)

But it would be my goal to improve and have a better-maintained Wikipedia to have such a bot, and the "Ignore all rules" policy seems to DIRECTLY support me, and in fact to go ahead with the bot despite a lack of approval. --75.5.176.16 (talk) 02:38, 27 February 2008 (UTC)

The problem with invoking "ignore all rules" to justify such actions is that we have about 1000 admins, one of whom might just "ignore all rules" and block you. Ignore all rules is not a tool to be trifled with. --Anticipation of a New Lover's Arrival, The 02:53, 27 February 2008 (UTC)

I wouldn't recommend that you try this. --Anticipation of a New Lover's Arrival, The 07:16, 26 February 2008 (UTC)

Well, a bit of an addition: If revertings would be 'too extreme', what about removing the deletion-flaggings and replacing them with user-or-article talk page notifications? Assuming good faith, a notification of some sort whould be a much more friendly and useful solution to images without a rationale. I don't see the use in the rampant deletion of images via a time-delay that users may not have time to respond to, or even understand. --75.5.176.16 (talk) 07:11, 26 February 2008 (UTC)

The procedure for speedying was, I seem to recall, agreed months ago. If you now have a proposal for changing it, it wouldn't be a specific Betacommand matter but a matter of policy, so you should discuss your proposal there. --Anticipation of a New Lover's Arrival, The 07:25, 26 February 2008 (UTC)

I'm not too familiar with the back-end of Wikipedia (yet)... would you be able to point me to a proper area to ask about such a policy question/proposal? Thanks --75.5.176.16 (talk) 07:31, 26 February 2008 (UTC)

I believe the procedure is outlined in Wikipedia:Non-free_content#Enforcement and the deletion criteria at Wikipedia:Criteria_for_speedy_deletion#I7. If you want to change the former I think you would therefore discuss your proposal on Wikipedia talk:Non-free content. --Anticipation of a New Lover's Arrival, The 07:43, 26 February 2008 (UTC)

Thank you. --75.5.176.16 (talk) 02:30, 27 February 2008 (UTC)

You don't seem to understand 'assume good faith'. No one is saying that the people who uploaded the images are trying to harm wikipedia. They are probably trying to improve wikipedia. I.E. No one is assuming that these images weren't uploaded in good faith. However they have also in most cases failed to comply with NFCC policy which we consider serious enough that it needs to be fixed ASAP or the images deleted. Nil Einne (talk) 10:06, 26 February 2008 (UTC)

Entirely serious answer: You are unregistered and have four edits, the first of which for a very odd comment on Beta's talk page and the other three to this page. So I'm going to go with the wild guess of "no" on this one. Now, from that odd comment and your comments here, it's obvious you're not a new user, so what account do you normally edit under? Lara❤Love 14:35, 26 February 2008 (UTC)

I have no account, mainly because of crap like BCB. Finally got annoyed by it enough to (maybe) pursue something. Also, why did you focus on my registration or lack thereof? Does that make my opinions or questions less valid? --75.5.176.16 (talk) 02:30, 27 February 2008 (UTC)

And another serious answer; policies don't work in conflict with each other. They work in harmony. Using one policy to undermine another policy isn't going to be supported. They work together, not apart. --Hammersoft (talk) 20:27, 26 February 2008 (UTC)

It seems pretty obvious to me that they do not. (Ref: any discussion on BCB.) --75.5.176.16 (talk) 02:30, 27 February 2008 (UTC)

None of this is helping anything. Why is anybody responding to this? If it's a legit threat, it won't last, and if it's not, you're just fanning the flames. —Torc. ^(Talk.) 03:30, 27 February 2008 (UTC)

How do you figure this to be a threat? I see something I dislike about Wikipedia and am fed up enough to do something about it. It might seem hostile to some eyes, and not to others, but it is not a threat, just something I am looking into at the moment. --75.5.176.16 (talk) 02:09, 28 February 2008 (UTC)

Outsider's perspective

I hesitate to dip my toe into such murky waters, but here is an outsider's perspective:

• BetacommandBot is performing a valuable service to the community by helping us work towards our non-negotiable deadline.
• People who work on copyright violations certainly come in for a lot of flack.
• All Wikipedia editors have a responsibility to remain civil, notwithstanding provocation.
• Bot operators have a responsibility to be responsive to questions and suggestions about their bot's behaviour.
• The notice at the top of User talk:Betacommand is definitely uncivil, especially the point about "whine and complain".
• The message left by BetacommandBot (example) could be more friendly, but I'm not seeing any concrete suggestions here for how to fix it.
• The periods of 48 hours for new uploads and 7 days for old uploads do seem a little short, and could usefully be lengthened, say to 7 days and 21 days.

My suggestion would be that we assemble a short FAQ that explains why BetacommandBot tags images, and outlines a workflow for fixing them. Betacommand can then link to this page in the bot's messages, and at the top of his talk page. Bovlb (talk) 21:18, 26 February 2008 (UTC)

• Wikipedia:Image copyright help desk/BCBFaq. --Hammersoft (talk) 21:21, 26 February 2008 (UTC)
  • That's excellent, and exactly what I was after. The question remains: Why am I not tripping over links to it everywhere? Why it it not in the message that BCB leaves? Why isn't in the box at the top of BC's talk page? I have added it to Wikipedia:Media copyright questions. Bovlb (talk) 01:09, 27 February 2008 (UTC)

Wikipedia:Image copyright help desk/BCBFaq says to use {{Template:Non-free_rationale}}, but the latter has no documentation or discussion, and a warning that essentially says "Don't use this". Guidance, please? —johndburger 15:46, 27 February 2008 (UTC)

Please see the TfD. It should not be used. The right hand doesn't know what the left hand is doing anymore. Assuming good faith per the sensible editor just below. EconomicsGuy (talk) 16:44, 27 February 2008 (UTC)

I would say this was a simple typo. I have changed it to recommend using {{Non-free use rationale}}. - AWeenieMan (talk) 16:52, 27 February 2008 (UTC)

Ah, got it, thanks. I added a pointer on the talk page for any future head-scratchers. —johndburger 16:57, 27 February 2008 (UTC)

I have asked the user to make his talk page notice more friendly, and to link to the FAQ, but from his response he does not seem inclined to do either. Bovlb (talk) 07:05, 1 March 2008 (UTC)

• We shouldn't be surprised by this at all. Bellwether ^B_C 03:47, 3 March 2008 (UTC)

Moving forward

• Looking at the conversation in #A Modest Suggestion above - MickMacNee's concerns, Hammersoft's response, and my summary - I have a few questions which I hope someone can answer, not only about the desireability of doing these things, but also the practicality of doing so (which I presume willl have to be answered by Betacommand):

• • If suggestions can be generated for improving the interface with the user, i.e. the messages left by the bot, can these be easily implemented? If so, what specific changes would be suggested to be made to help the typical user be less confused about what they need to do?

• • If some consensus can be reached about a better categorization scheme which would increase the chances of images not being lost at the deadline, can this be implemented?

• • If it is important that the bot run on a more consistent schedule, to minimize its impact, is there someone Betacommand would trust to run the bot at those times that Betacommand is not available, under Betacommand's direction, of course?

• Let's start with those questions, which I hope are not totally naive. Ed Fitzgerald (unfutz) ^{(talk / cont)} 21:25, 26 February 2008 (UTC)

Point 1: Maybe I've missed it, but the bot haters haven't come up with suggested changes that I know of that have not already been implemented.

Point 2: Probably. But, that has nothing to do with the bot really. It has to do with where Template:Dfu places images, as it is responsible for the categorization, not the bot.

Point 3: I can't respond, but past responses from Beta indicate (1) no, and (2) the initial tagging runs are all done, so there's not much point to this, nor is there much point in re-writing the bot from scratch as the work's been done. Now, it's just routine scans for new non-compliant images, and an additional phase that Beta's not coded to remove non-compliant images from articles (I'd guess this would run after the March deadline). --Hammersoft (talk) 21:31, 26 February 2008 (UTC)

Hammersoft: Thanks for the response. One caveat, though, since it's possible that everyone's taken a deep breath at this point, it would probably be best to keep things level and avoid using pejoratives whenever possible, so the discussion can continue at a lower decibel level.

About point 1, I guess the import of my question was if changes were suggested and agreed to, could they be implemented? Thanks. Ed Fitzgerald (unfutz) ^{(talk / cont)} 21:40, 26 February 2008 (UTC)

Ok, redoing:

• Point 1: If something was suggested that was reasonable and accurate, probably. Beta's been amenable to such changes before.
• Point 2: Sure. Anyone can update the template. It's not a bot function. The bot just uses that template.
• Point 3: Moot point. Work's done, and now routine scans are all that will happen.

Better? --Hammersoft (talk) 21:43, 26 February 2008 (UTC)

Indeed - thanks. Ed Fitzgerald (unfutz) ^{(talk / cont)} 22:06, 26 February 2008 (UTC)

Good suggestions, I think I touched on a few of them above in My Thoughts. As the deadline is approaching, how about just giving everyone until then to fix them. That is not a viable longterm solution, but as we are within a month of them all being gone, I don't see a problem with that time table. Also, I believe that BCbot should use a different template than a human user would for tagging 10.c violations, as they operate in a fundamentally different way. And finally, yes the scan schedule seems to be a moot point but perhaps we can learn from that in the future. Random89 (talk) 08:24, 27 February 2008 (UTC)

I wonder about the notion that the bot's irregular schedule causes any serious "impact". Carcharoth's page on compliance seems to suggest strongly that this and other bots have little or no impact on overall deletion rates. It's just a small, though necessary, part of the work Wikipedia does that results in image pages being improved to the point of compliance or deleted. --Anticipation of a New Lover's Arrival, The 14:08, 29 February 2008 (UTC)

I am not a "bot hater", but in the section above, I made two concrete suggestions: that the bot's talk page be more civil; and that the bot's messages and talk page link to the bot's FAQ. Bovlb (talk) 17:54, 3 March 2008 (UTC)

Categories

I've just blocked BetacommandBot (talk · contribs · deleted contribs · logs · filter log · block user · block log) because it looks like it's doing an unapproved task. It's approved to do CfD work and remove red linked cats after a discussion. What it's actually doing is removing all red linked cats, even those which have simply never been created. This has included removing them in userspace, where users have specifically wanted them to be red linked. I don't believe there is any consensus to remove all red link cats from all pages, and BrownHairedGirl made a very good point on Betacommands talk page saying that a lot of the cats are red linked because of small typos - this is easier to spot if the cat is red linked on the page rather than completely removed. Ryan Postlethwaite 18:03, 4 March 2008 (UTC)

Is there any reason not to permanently removal whatever approval Beta has to run any bots for any reason? He's unable to respond meaningfully to concerns about bot behavior, and they do seem to come up a lot. Perhaps he could find a way to help some other, more responsible, bot owner, short of running one himself? Friday (talk) 18:07, 4 March 2008 (UTC)

Ryan Postlethwaite please note that I stopped the bot over 10 minutes ago, shortly after your post to my talkpage. β^_command 18:09, 4 March 2008 (UTC)

I do appologise, it was still functioning a few mintues after I posted to your talk, and in the time I went to block, you must have stopped it. Ryan Postlethwaite 18:11, 4 March 2008 (UTC)

you posted at 17:52 the last edit by the bot was 17:56, and your block was 17:58 β^_command 18:13, 4 March 2008 (UTC)

Yup, exactly what I said - you must have stopped it in the time I went to block. Ryan Postlethwaite 18:15, 4 March 2008 (UTC)

Ryan I just considered this category house keeping part of the CfD work. Ive been removing redlink cats for a while. prior to this there have not been any issues. What I have seen from prior runs is that pages often get improper categories removed. and ones with typographical errors get fixed. in general a lot of good comes from these runs. β^_command 18:50, 4 March 2008 (UTC)

I am somewhat worried over Betacommand's operation of this bot, this is the second time in a month that it has been known to run a unapproved task, remember making a few thousands of null edits in a sub page to prevent deletion on the main page? - Caribbean~H.Q. 18:14, 4 March 2008 (UTC)

Indeed, there was a two minute gap between the stop of the bot and the block, which is not an unreasonable amount of time to elapse. Had you stoped your bot before you replied to the concerns on your talkpage and invited discussion about the subject rather than carrying on, it would have never been blocked at all. ➪HiDrNick! 18:18, 4 March 2008 (UTC)

(ec) It looks more to me as if it were running an approved task in a manner indicative of programmer error (setting which redlinks to remove), not necessarily anything nefarious. I also note that Betacommand had stopped the bot prior to the block - a four minute response time, quite impressive in any context, and certainly not indicative of an unresponsive bot operator. I'm inclined to tag this as Resolved, since it is unclear what further admin action is necessary. I presume the bot has been unblocked? UltraExactZZ ^Claims ~ Evidence 18:22, 4 March 2008 (UTC)

Obviously the bot isn't unblocked yet, given there's no assurances that the bot is fixed and won't continue to do this task. I actually don't think this was a bug - Betacommand seemed to think this came under the scope of "CfD" work which it clearly didn't. Ryan Postlethwaite 18:26, 4 March 2008 (UTC)

Not unblocked yet. And the response time wasn't 4 minutes. BHG posted about this at 14:48 (others may have posted earlier), edits resumed at 16:56, block took place at 17:58. In total, BetacommandBot has made over 2000 edits removing categories since 00:32. Why did Betacommand resume using the bot even after BHG had raised her concerns? I'm also concerned that BetacommandBot got approval to do category work in the first place. Category work is tricky and not at all easy to undo if a bot goes wrong. I'm hoping there won't be future problems, but this is the second time I've seen Betacommand get something basic wrong in category work. The first was a (non-bot) attempt to fully populate high-level categories without much discussion that I could see. This failure to distinguish between deleted categories and red-linked categories is another. And I thought Cydebot did this sort of thing anyway (and better as well)? If it had any chance of succeeding, I'd urge WP:BAG to withdraw approval for this bot to carry out any category work until more extensive testing and discussion, and well, some demonstration that Betacommand actually understands how categories work, but getting bot approval withdrawn, even in a specific case like this, is actually rather difficult. Carcharoth (talk) 18:45, 4 March 2008 (UTC)

I think BC's attitude towards criticism is unhelpful. There is absolutely no justification for describing the removal of a category that has never had a CfD as "CfD work". None. —Random832 18:18, 4 March 2008 (UTC)

I think BC probably did that in good faith, since he was delinking the category, removing the other non existent categories sorta makes sense, database-load wise (I was not aware that red cats were asked for). -- lucasbfr ^talk 18:26, 4 March 2008 (UTC)

• Re blocking a bot for running an unapproved task - just based on the policy, only completely new tasks require explicit approval. In the case of completely new tasks, it requires the approval of a member of BAG. This isn't a completely new task (as you say, BCBot does similar work already) and Betacommand is a member of BAG (currently, with Voice-Of-All, the member with the earliest join date). Blocking him a few minutes after leaving a note is sort of an assumption that he will refuse to discuss it, or change the operation of the bot. Given past experience, this doesn't seem like something that should be assumed. The bot should be unblocked, and input should be sought from another member of BAG. Avruch ^T 18:47, 4 March 2008 (UTC)
  • Well, this was a completely new task - he was removing all red linked cats, not ones that have been deleted through CfD like his approval states. There is no consensus to remove all red linked cats, so he shouldn't be performing the task. Other people left notes earlier with no response - I gave him one final opportunity before I blocked soon after myself. He was continuing to use his bot at first dispite my warning. I see no reason to unblock the bot until BC makes it clear he will stop the task. Ryan Postlethwaite 18:53, 4 March 2008 (UTC)

I'm fairly disturbed that BCB seems to come up here on a regular basis. After a lengthy discussion about the last episode on-wiki (tagging thousands of images for deletion), more discussion followed on en-wiki-l and it doesn't really seem that apparent that BC is understanding some of the core issues the community has here. What I see instead is some kind of repeated justification for doing things that were never specifically approved by the community. Does there come a time, when we've had enough, that's the question. Wjhonson (talk) 18:54, 4 March 2008 (UTC)

As a positive-action follow-up, I'd like to recommend that any changes to BCB code should go through an approval process by an unaffiliated set of eyes. That is, the code itself should be reviewed, not merely the textual explanation of what it does. Wjhonson (talk) 18:57, 4 March 2008 (UTC)

The WikiEn-l discussion is mostly irrelevant to on-wiki issues, IMHO, but even if not it should be stated that the thread was mostly you arguing your point against everyone else who disagreed with you. Not exactly proof of consensus for your position. Avruch ^T 19:13, 4 March 2008 (UTC)

Then you weren't paying close attention. Many posters agreed that the situation was not acceptable. And the discussion of wiki related material is never irrelevant. That's why we have the en-wiki-l list. I wasn not trying to prove consensus by my remark, only pointing out that there is a significant amount of discord over the actions of this bot. Wjhonson (talk) 23:45, 4 March 2008 (UTC)

I'll post here what I posted on BetacommandBot's talk page: This bot needs to follow {{nobots}} and {{bots}} per Template:Bots. I've about had enough of this bot removing things and when I revert, it coming back to do it again. And what about this bot edit where it removed one cat but said in the edit summary it removed another? Betacommand replied to me on BetacommandBot's talk page that he "doesn't have to follow" {{nobots}}/{{bots}}. Agreed, you don't "have to", but you should. All bots should - it should be a requirement for having a bot. If Category:Rouge admins still existed, I'd put this bot in it and add {{nobots}} to its own userpage! - ✰ALLSTAR✰ ^echo 19:04, 4 March 2008 (UTC)

Funny you should say "If...still existed" and have it show as a blue link. I agree that he should follow the bot/nobot template in userspace in most cases. Avruch ^T 19:15, 4 March 2008 (UTC)

Yeah, well, as you noted below, it's since been restored. lol - ✰ALLSTAR✰ ^echo 19:36, 4 March 2008 (UTC)

• Comment - This campaign has been very damaging to our users' navigation. Many, if not most of the cats removed were simple misspellings of actual cats that do certainly exist. These misspelled cats that were blanked should have been replaced with the properly spelled ones, by hand.

Examples of such poor editing include this edit and this edit.

The editor needs to go back and do that. If s/he does not, I recommend a "forever" block and complete dissolution of this bot, which has shown such disruption so many times in the past. Badagnani (talk) 19:38, 4 March 2008 (UTC)

User:BetacommandBot is removing misspelled (and thus redlinked) categories from thousands of articles, always without taking the few moments it would take to substitute the properly spelled category--thus often leaving the article without a proper category. This is very wrong and needs to be stopped. The redlink will be noticed eventually and experts in the particular subject of the article will substitute the proper cat. But removing the cat entirely without substituting the proper cat hampers navigation for all of our users. Please have this person go back and do this by hand, as should have been done in the first place. Badagnani (talk) 19:35, 4 March 2008 (UTC)

Examples include this edit and this edit. Badagnani (talk) 19:35, 4 March 2008 (UTC)

This is already under discussion, see Wikipedia:Administrators'_noticeboard/Incidents#BetacommandBot. Friday (talk) 19:37, 4 March 2008 (UTC)

I'd like some attention to this. Will the damage be undone? If so, when, and by whom? Badagnani (talk) 19:39, 4 March 2008 (UTC)

While it is true that Betacommand could do that, a bot cannot. One only has to see the difficulties related in spell-checking and related words to know that a bot cannot do all things a human finds obvious. x42bn6 Talk Mess 19:44, 4 March 2008 (UTC)

I (as well as the rest of the community) would like to see some attention to this. I do not care if it will be difficult to do by hand. The bot is operated by a real person, and that real person should edit articles in a thoughtful manner, not "blanking and running." Misspelled/redlinked categories need to be replaced with properly spelled ones. Attention to this would be greatly appreciated, more than excuses for the individual who operated his/her bot in such a damaging way. Will this damage be undone? If so, when, and by whom? Badagnani (talk) 19:49, 4 March 2008 (UTC)

Please exclude me specifically from this "the rest of the community" you claim to be speaking for. Thanks! ➨ REDVEЯS dreamt about you last night 19:51, 4 March 2008 (UTC)

I would like to go on record as stating that I _agree_ that there are some tasks that {{nobots}} does not and should not affect. BC's image tagging task even has two separate components: tagging images ignores nobots, but warning users (at least, IIRC) respects it. Of all the problems that BCBot has had, I don't think this is one of them. —Random832 19:54, 4 March 2008 (UTC)

Will this damage be undone? If so, when, and by whom? Clearly, User:Redvers has not actually examined the edits made by the bot, such as this edit and this edit, as well as a thousand others, which are clearly damaging to our project. Badagnani (talk) 20:12, 4 March 2008 (UTC)

Breathtaking, and wrong, assumption. I don't see a problem with losing unprofessional redlinked categories with misspellings in them, and I wish to be excluded specifically from when you are speaking for "the rest of the community". But assumptions beyond that are assumptions. Please refrain from making them. ➨ REDVEЯS dreamt about you last night 20:16, 4 March 2008 (UTC)

Agree with Redvers, on the cats, and, "the rest of the community" doesn't include me either. Please, speak for yourself. SQL^{Query me!} 20:24, 4 March 2008 (UTC)

Uncivility is un-Wikipedian and should be avoided. Please avoid it. Regarding the bot's edits, the damage needs to be undone. A "professional" bot operator would have used the bot to identify redlinked (misspelled) categories, and replaced them with properly spelled ones. What the bot actually did was highly damaging and the damage needs to be undone. Please do not attempt to hinder this, for whatever reason you are doing so. Badagnani (talk) 20:20, 4 March 2008 (UTC)

WP:SOFIXIT. Avruch ^T 20:38, 4 March 2008 (UTC)

Will this damage be undone? If so, when, and by whom? I am very familiar with bot misbehavior, which nearly always goes along with admin inattention, with the end product always being the user noticing the damage (rather than the one who did the damage, who always claims they were "just doing their job" and "it would be too hard" to fix it) being the one asked to undo it. It's time to break this cycle. Let's have some attention to this. Badagnani (talk) 20:41, 4 March 2008 (UTC)

Since simply not agreeing with you is "uncivil", it's hard to reply. Nevertheless: {{sofixit}} ➨ REDVEЯS dreamt about you last night 21:25, 4 March 2008 (UTC)

I provided diffs showing that the bot's behavior was improper and damaging to our project. I'm quite familiar with the policy of asking editors who notice bot damage to undo such massive damage (in this case amounting to literally thousands of blanked categories, for which the spelling should simply have been fixed. Such a request is wrong and it is time to break this cycle. Let's have some attention to this. Will this damage be undone? If so, when, and by whom? Badagnani (talk) 21:28, 4 March 2008 (UTC)

Badagnani, I see no incivility towards you here. Also, just as a piece of advice re: interacting in this forum: repeating the same questions, concerns, or information in reply to several different comments generally casts a less-than-ideal light on your ideas...in short, it makes you look argumentative, which I'm sure is not your intention. Just a thought. Gladys J Cortez 21:29, 4 March 2008 (UTC)

• Comment - I am also very familiar with comments on the comments made by editors pointing out bot damage rather than actually seeing to it that such damage is repaired, and the bot operator censured. Let's break this cycle. I didn't see that you addressed my actual comment. Badagnani (talk) 21:33, 4 March 2008 (UTC)

Will this damage be undone? It's not damage, per se. If so, when, and by whom? As soon as you like, by you. I can only put this so many ways: if you want to add redlinked and misspelled categories to articles, then you should do so yourself. There is unlikely to be a queue forming to do it for you. {{sofixit}} ➨ REDVEЯS dreamt about you last night 21:36, 4 March 2008 (UTC)

• Comment - I'm very familiar with the editor noticing bot damage (who usually does not have a bot at his/her disposal--and is just a highly productive contributor to our project) being the one asked to repair bot damage, often numbering thousands of damaged pages. I have provided diffs and the pages certainly are damaged, if the properly spelled cat was not substituted, as they should have been. It's time to break this cycle. Attention to this damage, and repair thereof, is needed. Badagnani (talk) 21:39, 4 March 2008 (UTC)

{{sofixit}} ➨ REDVEЯS dreamt about you last night 21:47, 4 March 2008 (UTC)

• Comment - I think I see what the problem is--the three editors above who seem to be defending the bot operator don't seem to have actually examined the diffs, which are typical of many hundreds of the thousands of articles damaged by the bot (in that slightly misspelled cats were blanked rather than repaired). In this edit, the redlinked/misspelled cat "Beijing Cuisine" was blanked, and in this edit the redlinked/misspelled cat "Botswana people" was blanked. In both cases, one letter would have fixed it. The articles were left without cats and, without the redlinked cat to alert editors in each subject area that the cats needed to be fixed, now they're simply not there, hampering navigation for all our users. This was an improper use of a bot and needs to be fixed. Simply defending the bot operator (perhaps because the other editors are also bot editors and strongly wish that such a precedent not be set, that bot operators who make mistakes actually have to go back and fix their own mistakes, by hand, as the rest of WP editors do), and asking the non-bot-operating editor who noticed the damage to do to, serves to support such damaging actions in the future, with no accountability for such widespread and huge damage to our project. This is a cycle that needs to be broken. Let's see to it that this fixed--and that we don't shift the onus from the actual editor who caused the damage in the first place (in this case, an editor who has achieved widespread notoriety for such damage). Badagnani (talk) 21:50, 4 March 2008 (UTC)

That's the second time you've made the breathtaking assumption that I'm commenting without reviewing the evidence, and the second time you've been startlingly wrong. As you say: "one letter would have fixed it" - go on then: fix it. Nobody is going to re-add misspelled, redlinked categories to articles. A bot cannot add them with the correct spelling - bots cannot spell. If this is as big a problem as you believe it to be, you are going to have to fix it. {{sofixit}} ➨ REDVEЯS dreamt about you last night 21:57, 4 March 2008 (UTC)

<sigh>This bot is worse than Britney Spears or Michael Jackson about always being in the public eye.</sigh> - Philippe | Talk 22:00, 4 March 2008 (UTC)

• Comment - A bot may be used to identify improperly spelled categories, then the correct ones may be located and added. This is now impossible, as the cats were blanked entirely. This was improper. The fact that this damage was done to thousands of articles makes it improper to ask the editor noticing the damage to repair it, rather than the bot operator who did the damage--and also encourages the bot operator to keep doing such damage in the future. I'm very familiar with the onus being placed on the editor noticing the damage, shifted away from the actual bot operator who did the damage. This is a cycle that needs to be broken. Let's have some attention to this and the damage corrected. See Ryan Postlethwaite's first comment above, which says "BrownHairedGirl made a very good point on Betacommands talk page saying that a lot of the cats are red linked because of small typos - this is easier to spot if the cat is red linked on the page rather than completely removed." Badagnani (talk) 22:04, 4 March 2008 (UTC)

It is unlikely that any bot or program can guarantee a correct decision especially regarding spell-checking. It is difficult enough for a bot to realise that "Micheal Jackson albums" [sic] is really "Michael Jackson albums" or that "Michael Jackson Albums" is "Michael Jackson albums". To me there is no real issue with redlinked categories being removed although HTML commenting would possibly be better but I do not feel there is any lasting damage especially as one has the history tab. x42bn6 Talk Mess 22:11, 4 March 2008 (UTC)

(ec)Then write a bot or request someone with bot-making experience makes you one. Nobody is going to re-add misspelled, redlinked categories to articles so they can be fixed later. We are better off without them. If you want, go through the bot's contribs and make any changes you want yourself. But there will be no wholesale return of misspelled, redlinked categories, there is no way a bot can add categories with the correct spelling, and there is no point continually arguing for these things or playing the martyr on this noticeboard. ➨ REDVEЯS dreamt about you last night 22:12, 4 March 2008 (UTC)

Arbitrary section break

• Comment - The damage is extensive and makes navigation impossible for thousands of articles. Please read the first comments in this thread, and also the comments at Betacommand's page for more background on this issue. The two diffs presented above show that this massive blanking has damaged thousands of articles, by leaving them with no cats. A bot may certainly be used to identify redlinked cats, then the properly spelled cats can be put in. It simply takes care. In this case, from Betacommand's responses, s/he simply didn't care to take care. Supporting his/her actions is improper in light of this massive damage, and encourages future damaging campaigns of this nature on his/her behalf. We must be reasonable, and asking the editor who noticed the damage, who does not operate a bot (as has happened many times in the past) is improper. It is time to break this cycle, and to finally ask that the bot operator who perpetrated this damage to our project correct him/herself. Badagnani (talk) 22:18, 4 March 2008 (UTC)

Redlinked categories mean that navigation is impossible in the first place. Redlinked categories are no more damaging than having no such redlink category on an article. BetacommandBot acted in good faith although it would have possibly been better to think of other solutions but I still do not see the point of continuously attacking Betacommand. x42bn6 Talk Mess 22:22, 4 March 2008 (UTC)

(ec)There's no communicating with you, is there? This is the third time you've made the breathtaking assumption I've not looked at the evidence before commenting, and the third time you've been wrong. ➨ REDVEЯS dreamt about you last night 22:22, 4 March 2008 (UTC)

I don't understand how the removal of a redlinked category makes navigation impossible. Badgnani, could you explain what you mean by that? - Philippe | Talk 22:26, 4 March 2008 (UTC)

• Comment - Please see the diffs I provided above. The blanking of the misspelled/redlinked "Botswana people" rather than replacing it with the correct "Botswanan people" (as a thoughtful, reasonable editor would have done) means that this individual cannot be found in such a manner. There are thousands more unrepaired articles of this sort. When will this damage be fixed? Badagnani (talk) 22:32, 4 March 2008 (UTC)

• Time to set up Wikipedia:Administrators' noticeboard/Betacommandbot and get the constant argumentation off the noticeboards. Guy (Help!) 22:28, 4 March 2008 (UTC)

How about WP:AN/B? Its nicely bluelinked already, and full of delicious arguments. Avruch ^T 22:31, 4 March 2008 (UTC)

• There appears to be a record playing here, on continuous repeat... But I can't recall the album name. Any ideas? Avruch ^T 22:30, 4 March 2008 (UTC)

What about this one?. Or even just the the disam page would suffice... Keeper | 76 | Disclaimer 22:37, 4 March 2008 (UTC)

• Comment - Asking that damage caused to our project by a bot be corrected is not an "attack." A bot certainly can be used to identify redlinked cats and replace them with the correctly spelled ones, although it would take care to do so. Let's see to it that this damage is corrected. Covering for the mistaken bot operator (likely by other bot editors who fear a precedent being set, that bot errors will have to be repaired by the bot editors themselves) and insisting that s/he not repair such damage is illogical and equally damaging to our project, in that it encourages future damaging campaigns of this nature. The diffs provided show that thousands of articles are still without cats, whereas the redlinked cats would eventually have alerted editors active in those particular subject areas. There are many ways to proceed in correcting redlinked cats, and this was not the proper way to have proceeded, as witnessed by the earliest comments in this thread, and blocking. When will this damage be repaired? Badagnani (talk) 22:30, 4 March 2008 (UTC)

Can you point to a precedent for requiring affirmatively that an individual make or unmake a particular action? We have prohibitions, but affirmative requirements I think are some what in contradiction with the "volunteer" aspect of Wikipedia. Avruch ^T 22:33, 4 March 2008 (UTC)

I can. The bot policy is very clear about this: "The contributions of a bot account remain the responsibility of its operator. In particular, the bot operator is responsible for the repair of any damage caused by a bot which operates incorrectly." Those who show themselves unwilling to accept this responsibility should not operate bots; this entire issue could have been avoided if the operator had sought consensus for the change before making it. ➪HiDrNick! 22:49, 4 March 2008 (UTC)

That is a very good point (the bot policy note). Avruch ^T 23:40, 4 March 2008 (UTC)

• Comment - Your comment above (i.e., Avruch's comment about the record) makes it difficult to respond to you, as it appears to demonstrate a lack of seriousness regarding this issue. Badagnani (talk) 22:35, 4 March 2008 (UTC)

• Look, it's a non-starter to demand Betacommand himself go back and fix all of the broken categories his bot removed. It's not like things were fine until his bot came along; things were broken, and his bot came along and broke it in a different way. However, I'd also like to ask people to re-consider the underlying suggestion we would be better off with no category than with a misspelled category. Please consider the message, not how annoying the messenger is. I'd like to suggest one of these two options:
  • (a) Preferred option (IMHO): Could someone with AWB or script or bot skills revert all the edits BetacommandBot made on this red-category run, so (as some point out) the red category is back on the page, hopefully catching the eye of interested editors who will fix the spelling when they see it? Also, I assume Betacommand got this list of empty categories from somewhere; if you point out to category-challenged people like me where that list is, I might find it interesting to go thru and change the spelling from that end too.
  • (b) Alternate option: Could someone with AWB or script or bot skills make a list of all the edits BetacommandBot made on this red-category run, and stick them on a page somewhere (User:Barneca/BCBot's redcat run if necesary, or better yet some appropriate wikiproject's space) as a click-able list? Then, those who want to can look through them, and relatively easily fix the misspelling.

Option (a) is better IMHO. I'm just afraid that the antagonistic approach some people have taken to BetacommandBot have pissed enough bot-people off that no one would be willing to do it. But either way, we'd have a list to work off of, and could stop arguing and start fixing. However, it seems clear to me that in the future, this isn't the greatest task for a bot to take on; I think the pre-bot broken situation was better than the post-bot broken situation.

Finally, there's a certain amount of Boy Who Cried Wolf syndrome here. So many people scream about BetacommandBot so often for so many incorrect reasons, that many bot-type people automatically tune out what might be a legitimate request. --barneca (talk) 22:36, 4 March 2008 (UTC)

• Comment - This is unhelpful and unreasonable. The damage needs to be repaired. Leaving an article about a Botswanan person with no cat--to one article, let alone thousands--then moving on to other tasks, is simply unreasonable and unacceptable. Badagnani (talk) 22:39, 4 March 2008 (UTC)

Jesus Christ, Badagnani, it looks to me like I'm the closest thing you've got to an ally in this argument, and I'm "unhelpful and unreasonable"? Saying the same thing 1000 times doesn't help your case, it hurts it. Read my lips: I 100% guarantee Betacommand is not going to be forced at gunpoint to replace the broken links with unbroken links. Due to your behavior here, I think the odds are pretty similar that no one is even going to undo his bot's edits. If you keep it up, no one is going to do anything. Please. --barneca (talk) 22:47, 4 March 2008 (UTC)

So far, and not to badger you but, your position here has found no support - and for good reason. If you think that the problem of having the non-categories of articles removed is big deal, you are able to fix it. There is no precedent or principle that would allow us to mandate Betacommand take a set of actions (particularly if its laborious, and yes I know he would only be reverting his own activity). So, since no one else is going to do it right away, and we can't make BC do it... Feel free, eh? Avruch ^T 22:45, 4 March 2008 (UTC)

Badagnani, the English language means that basic spelling corrections are unlikely to be always correct - incorrect category selections, in my opinion, do more damage than redlinked categories. Take Category:Iraq, for example. Say an article was categorised as Category:Ira. Should this article be categorised under Category:Iraq or Category:Iran? A bot cannot be carefully written to categorise articles without a lot of work - until today, not even Microsoft Word has spell-checking to the extent that the word you intended is always the one on top of its suggestions. Also, take Category:Guantanamo detainees known to have participated in their seconnd ARB hearing which is a redlink but has an article categorised under it. It has a spelling error yet even if this was corrected, it would be put into a redlinked category. There are some things humans must do because bots cannot do everything. There may come a time when a perfect bot is able to guarantee spellings are always correct but today's world does not allow for that. x42bn6 Talk Mess 23:00, 4 March 2008 (UTC)

• Comment - We must always be reasonable in everything we do at WP. Such arguments, propping up the massive damage conducted by this bot operator, insisting that s/he not repair it, and ignoring the lack of navigation caused by this blanking of cats, does a disservice to our users, and shows contempt for the proper care of our content. I've encountered it many times before--in fact, every time a bot has similarly damaged our project. The other bot operators come out to defend the improper actions of the bot, insist that the editor noting the damage (who is usually simply a highly productive contributor without access to a bot) be the one to correct it, and fail to show any concern for our users or the data we should be maintaining with great care and reasonableness--presumably because they are other bot operators who fear a precedent being set (that is, that a bot operator conducting a similarly misguided, damaging, and unapproved campaign be asked to repair such damage after it is discovered). Rather than spend many dozens of paragraphs to insist the improperly behaving bot editor not do such work, it would be best to actually see to it that this massive damage be promptly repaired, as would be the most reasonable course of action. Badagnani (talk) 22:58, 4 March 2008 (UTC)
  • "and ignoring the lack of navigation caused by this blanking of cats" - If there are redlinked categories, how is this any better for navigation purposes? If there is damage then the onus is on the operator to fix it; however, I still do not see the damage caused by cleaning up purposes. BetacommandBot, as I have said above, cannot think like a human and therefore cannot choose at its own free will (or, rather, lack of it) what is correct and what is not. I think you are arguing that if there were redlinked categories then it would be easier to spot and fix. Yet there are cases where if there is a redlinked category then it should be removed (i.e. Category:Wikipedians who believe that a;sfjdoiausdof8iausf is a correct word because THEY SAY SO). But if a legitimate category is removed, then someone will inevitably categorise it properly, just as inevitably, a user will fix a spelling mistake for a category, or a user will remove superfluous words to make a category correct, or a user will remove an incorrect category. For a bot, I believe it will only ever be able to remove redlinked categories and to me there is no damage - but if you feel there is, you have a list of contributions and you have a history tab - and this is fixable. x42bn6 Talk Mess 23:08, 4 March 2008 (UTC)

Hi x42bn6, I agree with almost everything you're saying about BCBot's inability to figure out what the cat "should" have been; when it evolves to the point that it can do that, I'd also like it to make me a nice cup of tea. However, I agree with just one of Badagnani's points: I think a misspelled category is better than a missing category; a misspelled category will hopefully catch the attention of someone who knows enough about the article to fix it, whereas the fact that there is a missing category, while likely to get fixed sometime, is less likely to be noticed soon. I think it would be marginally better if BCBot's red category edits were reverted. And going forward, for the very reasons you mention, I think this kind of removal is not really appropriate for a Bot. As you say, some things are better done by a human. What do you think of my 2 suggested options i made earlier? Would either one of those be acceptable? --barneca (talk) 23:19, 4 March 2008 (UTC)

I prefer b) although I would not be opposed to a). But until we know something about the "damage" BetacommandBot has done nobody will know which option is best. That said, it may probably best to do a) because there is no guarantee that a list of 1000+ category checks will be fixed quickly. Choices, choices... x42bn6 Talk Mess 23:26, 4 March 2008 (UTC)

The question is, where to have this discussion, since it seems to have been hijacked here. --barneca (talk) 23:42, 4 March 2008 (UTC)

• Comment - Diffs for representative categories (slightly misspelled ones) that have been blanked have been given above. I certainly do not object to the removal of nonsensical cats such as "XESSESSLDFKSJDF Wikipedians" and this is why editors (even those who operate bots) must always act with the utmost care and reasonableness. Badagnani (talk) 23:29, 4 March 2008 (UTC)

Well, I think this part of this discussion has reached the limit of its usefulness, and successfully derailed the rest of the discussion as well. BCBot/BC is not going to be required to change everything back. Period. You might ask him, but good luck. I wouldn't do it if I were him, even though I am actually not a bot operator at all. I'd suggest, Badagnani, that repeating yourself over and over again with little variation is probably not going to convince anyone to your point of view. Again, though, good luck. Avruch ^T 23:22, 4 March 2008 (UTC)

Hi Avruch. OK, but where would you suggest I go to make my comment/suggestion if this thread is no longer useful? So far, the only response I've got is from Badagnani... I don't think I'm being so shrill that I should be ignored too. --barneca (talk) 23:29, 4 March 2008 (UTC)

Sure, somewhere else other than this section (or even, this issue) would probably work. HiDrNick has made a good point above that the bot policy appears to require bot operators to undo damage caused by their bot. Since the bot policy also specifically prohibits bots from correcting spelling errors, BCBot would have no other option than mechanically replacing all the broken cats with new and identical broken cats. Avruch ^T 23:42, 4 March 2008 (UTC)

• I agree DrNick's point is a good one, but we haven't reached a consensus that what BCBot did was "damage". It was broken one way, now it's broken the other way. People with the best interests of the project can disagree on which is worse (as it appears you and I do). But I don't know how to have this conversation (anywhere) in a calm organized way with all this loud background noise. --barneca (talk) 23:50, 4 March 2008 (UTC)

• Comment - Assisted spell-checking is acceptable (as is doing it by hand). The bot can be used to simply identify redlinked cats, whose spelling can then be fixed in an assisted manner, or by hand. The claims that bots can't do this or that fail to acknowldge actual solutions that have already existed for years. Badagnani (talk) 23:46, 4 March 2008 (UTC)

Agreed. But Betacommand took a slightly blunter approach - not wrong, but possibly not the best solution. That said, tool-assisted or manually done are unfeasible if Betacommand wishes to do it alone, but it is one of the solutions presented below. x42bn6 Talk Mess 23:51, 4 March 2008 (UTC)

• Comment - We must be reasonable in everything we do. Inserting more paragraphs about how such damage is fine, and insisting that a bot operator causing such damage not be asked to repair it, is unreasonable, does a disservice to our users, and our content. Let's see to it that this is fixed, promptly. Badagnani (talk) 23:27, 4 March 2008 (UTC)

• Comment Is there any chance that you could stop repeating yourself? I think anyone who was listening to begin with has turned off by now. Black Kite 23:31, 4 March 2008 (UTC)

• Comment - If the problem were fixed after the first (or second, or third) request rather than the repeated (and repetitive) comments insisting that the editor noting the damage be the one to repair it (and repeatedly insisting that the bot operator not repair it, and instead move on to other tasks), repeated requests for such a prompt repair would not have been necessary. I think that's quite clear. Badagnani (talk) 23:34, 4 March 2008 (UTC)

Badagnani, you're missing the point. A bot cannot be careful. It cannot say to itself, "Hold on a second, that looks like an obvious misspelling - I'll run it through my in-built spell-checker and test it against the "obvious" capitalisation errors." It is not necessarily a spelling error, capitalisation error or anything - it could be an extra word, an extra space, a word that is a synonym to another word in the category, and so on. A bot cannot be careful. Please remember that. x42bn6 Talk Mess 23:43, 4 March 2008 (UTC)

• Comment - This is incorrect. A bot can identify errors, which can then be corrected in an assisted manner (or by hand). Bots are operated by human beings, who must always edit with great care. This was not done in this case, and this damage needs to be repaired. Badagnani (talk) 23:48, 4 March 2008 (UTC)

I agree, but Betacommand cannot possibly do all of it alone - there are several thousand things to fix. But it still does not mean there is damage that he needs to reverse. If a person builds a house but it turns out there is a better way of building a house, then there is no need to unbuild the house to build it the new way unless the old way is completely wrong. But Betacommand's method is not wrong - it is, however, heavy-handed and for some cases possibly makes things harder to fix. x42bn6 Talk Mess 23:56, 4 March 2008 (UTC)

• Comment - Leaving a page for a Botswanan person with no cat (at all) was an improper action. There are up to several thousand other pages like it now, with no way to know which need to be fixed (even by editors skilled in this subject, who would normally attend to such things). Even a bot set to add a notice regarding such a misspelled cat would have been preferable, as it would have allowed such editors to identify and correct the problem. As Barneca noted, now nobody knows there is a problem, whereas the redlink shows it clearly. It was necessary to insist that this be fixed because there have been many prior instances where the damage was as great or greater--and the insistence that the bot operator not correct the mistake was, despite the very clearly stated bot policy, similarly as great or greater. We must have the utmost care and respect for our data, and users. Badagnani (talk) 00:01, 5 March 2008 (UTC)

• Comment Yes. You are correct. You were correct the first time, and you are still correct this time, which I make to be approximately the 22nd time you've posted something similar on this page. Please stop now, this is starting to verge on the disruptive. Black Kite 00:16, 5 March 2008 (UTC)

• Badagnani, you're being quite disruptive here. I ask you that you contribute here much more constructively here, so please treat this as a final warning for disruption. I suggest you refrain from editing this page for say, 12 hours. If you continue your disruption, I can formally ban you from this page from 12 hours and if you violate it that ban, I'll block you for 12 hours. How does that sound? Maxim_(talk) 00:20, 5 March 2008 (UTC)

• Comment - I appreciate the constructive criticism of my postings. Can you outline specific elements of my comments to which you object (so strongly as to block one of the most productive contributors to our project?), so that I may not make them again? Thanks, Badagnani (talk) 00:24, 5 March 2008 (UTC)
  • You're making the same demand over an over again. See your posts on this page at 20:20, 20:41, 21:33, ... 00:01 I hope you get the gist of it. You've posts with the same point around 20 times; that's really disruptive. Maxim_(talk) 00:27, 5 March 2008 (UTC)
    • I even largely agree with the point you're making and even I'd support this block if it was enacted. Saying or making the same point several times under "Comment" sections is close to blatantly disruptive, and when your behaviour itself is derailing the discussion, it actually (and somewhat sadly) ceases to matter whether you have a valid point or not. Orderinchaos 13:37, 6 March 2008 (UTC)

Discussion got swamped

I like to address here how the useful discussion above got swamped by Badagnani and those responding to him. Badagnani had some useful points to make, but overdid it by incessant repetition. Equally though, those incessantly responding to him were feeding that sort of behaviour, and were contributing to the swamping of the discussion. Please, in future, just respond a few times and then refer them to your previous answers and then ignore. Carcharoth (talk) 05:21, 5 March 2008 (UTC)

• Some specific points. Redvers said "I don't see a problem with losing unprofessional redlinked categories with misspellings in them" - the correct approach is to correct the spelling, if the category is needed. This cannot be done if the category tags are simply removed. I am puzzled as to why Redvers and others cannot see this basic point. I agree with Barneca that in the case of "requested" categories (ie. ones where someone put a category tag down without creating the page) and mis-spelt categories, BetacommandBot has changed the situation from one broken one to another broken one, but with less chance (in fact almost no chance) of being corrected. Bot operators should be prepared to revert their edits or clean them up manually if need be. It is an important part of the process, as it makes them more careful next time. Carcharoth (talk) 05:21, 5 March 2008 (UTC)

Constructive step

If anyone's interested in actually starting to fix things (and note, whether or not BCBot's edits are reverted, a human has to fix these things), My User:Barneca/Sandboxen/Sandbox page has a list of all of BCBot's red category removals. It isn't formatted perfectly, and I don't have the time or inclination to perfect it, but since a decison about reverting the edits, one way or the other, is going to be made in April or May at the earliest, I did what I could now. I'm going to tackle three of them and call it a night. Anyone else? --barneca (talk) 00:35, 5 March 2008 (UTC)

I did a few, 50-75 or so - maybe a bit more. I'm not doing the other 1925 though ;-) Avruch ^T 00:54, 5 March 2008 (UTC)

If you want to find and fix uncategorized articles, have a look at Special:Uncategorizedpages and Category:Uncategorized pages. If you want to find and fix red category links, see Special:Wantedcategories. Bovlb (talk) 01:03, 5 March 2008 (UTC)

(e/c with Bovlb) I saw that, thanks Avruch. I see you went the reverting route. I misunderstood and thought we disagreed on that, so I went the hard way instead. I'll revert a few myself, but then need to go do real world stuff.

From a little research, if all of BCBot's removals are reverted, it appears the place they would show up, and where humans can go to correct misspellings, remove obvious mistakes, recategorize, etc. for nonexistant categories with members is Special:Wantedcategories, which is formatted much better than my sandbox! G'night. --barneca (talk) 01:05, 5 March 2008 (UTC)

I disagreed on making Betacommand do it, mostly. I don't think its really necessary to revert them, but if it makes people happy... Its fine with me. If later consensus says the cats should be removed, BCBot can pretty simply do it again. Avruch ^T 01:07, 5 March 2008 (UTC)

To be safe, I'm assuming I'd get lots of people mad at me if I used rollback on this, right? --barneca (talk) 01:12, 5 March 2008 (UTC)

p.s. We never disagreed on making BC do it. --barneca (talk) 01:13, 5 March 2008 (UTC)

Who knows. I used it. If someone wants to remove it from me for this, oh well. The whole "Only ever use this for vandalism" argument is a pain in the butt. Its a tool, use it when its useful and don't when it isn't. That said, I hardly use it anyway. Avruch ^T 01:15, 5 March 2008 (UTC)

I agree, the distinction seems silly to me (especially when reverting a BOT for pete's sake), but I'd also like to hang on to my adminship for longer than 20 days. I think I'll be unbold --barneca (talk) 01:37, 5 March 2008 (UTC)

What would be useful is running a spell-check over the list of categories, and looking for incorrect capitalisations, and seeing if the corresponding categories exist. Barneca's sandbox is useful, but a bit messy now, so I'll start my own list and analysis. I'll do this off-wiki, but it will be interesting to see how many red-linked categories were correctly or incorrectly removed. Carcharoth (talk) 04:48, 5 March 2008 (UTC)

My admittedly small sample showed very roughly 1/4 should just be removed, 1/2 were spelling or capitalization or similar problems, and 1/4 were complicated cases requiring more category experience than I have, usually creating the category, or using what would have been a parent category. FYI most of the "spelling" errors weren't actual spelling errors, they were either capitalization errors, or mistakes like using (as a made up example so I don't have to check spelling) Category:Municipalities in Oregon instead of Category:Cities in Oregon. --barneca (talk) 05:04, 5 March 2008 (UTC)

I'm going to see if I can tackle this tomorrow and try to fix some of the deleted categories. It's very disappointing that people are having to clean up for what the bot did. Enigma ^msg! 07:11, 6 March 2008 (UTC)

Fairyland, FieldTurf, The Barber, Eleanor Boardman, User:Astronouth7303, New Era University, PAOK F.C, Nakoruru, User:Kmweber,Fred De Luca,Casshern,Tracey,List of candidates in the United States presidential election, 2004,Odawara Station.

It's very time-consuming to edit the sandbox and cross out all the ones I did. Oh well. I don't think anyone will ever be able to completely fix this, unless someone designs a bot to fix what another bot did.

BCBot indefinitely blocked

So, BCBot remains indefinitely blocked. The areas that we might focus on are:

• Should the bot remain blocked, and for how long?
• Should a request for approval to BAG have been made, despite Betacommand's status as a member?
• If yes, and this is a repeat violation of the requirement to have task approval ahead of time, should Betacommand again be removed from the bot approvals group?

Anyone have any opinions on this, rather than the secondary stuff above about reversing all the removals of broken cats? Avruch ^T 23:26, 4 March 2008 (UTC)

• Comment - Yes. And the bot operator should undo the damage done (i.e., the blanking of slightly misspelled categories). Badagnani (talk) 23:35, 4 March 2008 (UTC)
• Comment - Status as a member does not mean you can do any code change you wish, anytime you wish, and then run it without notification.Wjhonson (talk) 23:37, 4 March 2008 (UTC)

• Well, I'm aware that this isn't the only thing BetacommandBot does so it should be unblocked if it wants to do something else, but it should leave this task alone until something has been fixed.
• It isn't required at present but such a thing would be preferred, to me. Not necessarily to seek approval but to see if there are alternative solutions or overriding criteria, or even if there is an existing task already, amongst other reasons.
• No opinion. Mistakes can happen. x42bn6 Talk Mess 23:34, 4 March 2008 (UTC)

• As I said above, I don't think this kind of knowledge-intensive task is a good thing for a bot to be doing. If BC says he won't have his bot do this particular task anymore until a consensus forms that he should resume what he was doing, unblock; I certainly don't think unblocking should depend on whether the edits are reversed.
• It doesn't matter if a BAG request should have been made; BC's actions were in good faith, and looking back like that serves no purpose. I do think that a consensus here or at BAG should be acheived before the same task is resumed.
• It would damage the project to remove BC from BAG, and I can see no upside to it. --barneca (talk) 23:37, 4 March 2008 (UTC)

• (a) it'll need to be unblocked for the next time it needs to run its core work. (b) it would probably have been a good idea. (c) No. Black Kite 23:47, 4 March 2008 (UTC)

Yes to all three points:
a) BetacommandBot should remain blocked until i) Betacommand acknowledges that the task was run without approval, and demonstrates that he he understands the difference between implementing decisions agreed at CfD and simply removing redlinked categories, ii) the bot's approved list of tasks are submitted afresh for approval, so that the list can be checked carefully for any possible ambiguities
b) Yes, approval should have been sought. "CFD" clearly refgers to WP:CFD, "Categories for discussion", and these edits are nothing whatsoever to do with WP:CFD.
c) Yes, Betacommand should definitely be removed forthwith from the BAG. Quite apart from the serious issue of repeatedly running a bot for an unapproved task, there is a clear conflict-of-interest between Betacommand's presence on the BAG and his status as the owner of the bot which generates the most controversy. I quite accept that some of that controversy is over his legitimate work wrt to images, and that most objections to that work are simply wrong ... but some separation of powers is needed here, and Betacommand's position presents a clear conflict-of-interest. --BrownHairedGirl (talk) • (contribs) 04:32, 5 March 2008 (UTC)

Unblock

As the blocking admin, I think I should clarify what I'd like to see before an unblock. Just a simple statement from BC to say that he will stop removing red linked cats, and should he want to do the same in the future, he will firstly seek consensus that it's what's wanted first, and then get BAG approval. It's not difficult for him to say that and it's my only caveat for unblocking. Ryan Postlethwaite 23:57, 4 March 2008 (UTC)

And I'd just like to see the damn bot follow {{nobots}}/{{bots}} on userpages. Seriously, how hard is it to implement that in the code? I won't support ever being unblocked unless this is done. It's called "bot with manners", and this one doesn't have any. - ✰ALLSTAR✰ ^echo 01:03, 5 March 2008 (UTC)

As I said this task will not be run without further discussion. β^_command 04:19, 5 March 2008 (UTC)

Then I'd agree to the unblock. But the fact that the task ran into difficulty either means that there was not enough oversight at WP:BAG, or it was incorrectly implemented. Which was it? Carcharoth (talk) 05:02, 5 March 2008 (UTC)

Apart from the commitment to stop removing redlinked cats, I want to add one more: that before any unblocking, Betacommand should be removed from the Bot Approvals Group. There is a clear conflict of interest between Betacommand's presence on BAG and BCB's status as a bot which has repeatedly run unapproved tasks. BCB's work in images is important and needed, but it is highly controversial among editors who don't understand why it has to be done, and it is basic good practice that there should be separation of powers between running such a bot and approving it. --BrownHairedGirl (talk) • (contribs) 04:39, 5 March 2008 (UTC)

The discussion where Betacommand rejoined WP:BAG can be seen here. There is a bit of history behind WP:BAG, its operation and membership, including the proposed deletion of WP:BAG itself at one point. I recently proposed that people who weren't bot programmers but were knowledgable on policy and guideline issues would be helpful as part of WP:BAG to bring a wider perspective to the table and ensure a stronger voice on non-technical issues. All too often bots are approved without requiring further discussion in specialist areas of Wikipedia (eg. WP:MoS or WP:CfD), or further community discussion. Also, like admin flags, it also hard to get bot flags or specific bot task approvals withdrawn through WP:BAG. One proposed community-drive process is at Wikipedia:Requests for comment/User conduct#Use of bot privileges, which is untested so far. Carcharoth (talk) 05:01, 5 March 2008 (UTC)

• Comment - It's being corrected, at User:Barneca/Sandboxen/Sandbox. Please take a look, and assist if you would like. I note that the human behind the Beta Command bot has not lifted a finger to assist yet, apparently preferring to simply move on to other projects, as bot operators generally do when they make an error (despite WP's very clear bot policy, which states that they must be the ones to make the repair of damage they cause). Regarding my comments, had you contributed to the discussion earlier, I would not have needed to reiterate my request that our bot policy be adhered to, and Betacommand undo his/her damage. Lord knows, it's been so many times that the same thing occured, and no one like you ever contributed with similar comments to the discussion (the bot operator generally either ignoring my request or stating in no uncertain terms that it "wasn't their job" to make the repair, and that it would be "too difficult," and they're not doing it--but they always invited me to undo it, even if it was into the thousands of articles damaged), and, in fact, there are articles that have been without the proper cats for over a year (and still are without cats) because of this. Badagnani (talk) 06:04, 5 March 2008 (UTC)]

I really like BrownHairedGirl's idea of seperation of powers. Perhaps BC would voluntarily step-down from BAG? Either way, I'd agree with this proposed policy. I also very much like nobots on userpages. And can we add no bots on user talk pages as well? I see no point for bots in non-article space at all to tell you the truth. I went to the RfC but I don't see anything there yet. Wjhonson (talk) 06:09, 5 March 2008 (UTC)

What do you mean by "separation of powers"? It's not like BC claims he's self-approving his tasks. Gimmetrow 09:30, 5 March 2008 (UTC)

• Bot has been unblocked by User:Krimpet at 04:31, 5 March 2008, with block log summary "bot has stopped". Carcharoth (talk) 09:41, 5 March 2008 (UTC)
  • That's a real pity. There seems to have been attempt by Krimpet to discus the issue (see [contribs), and no consensus here to unblock. I will ask Krimpet to reinstate the block. --BrownHairedGirl (talk) • (contribs) 11:31, 5 March 2008 (UTC)
    • I seriously don't see anything wrong with an unblock. BC's said he wont run the task without discussion. Doesn't stop its other tasks from running (e.g. Wikiproject tagging). If the task runs, reblock. Will ^(talk) 11:55, 5 March 2008 (UTC)
      • Me either. The unblock seems perfectly sound to me. SQL^{Query me!} 14:42, 5 March 2008 (UTC)
        • I too support the unblock to allow the bot to carry out other tasks, but I am concerned that Betacommand hasn't actively engaged in the discussion here (he has made a total of four edits). There is a direct question from me about whether this was a misunderstanding on his part, or incorrect programming of the bot, that he has not yet answered. Does he not care about the effects his bot has had? Do either of you (SQL and Will) have comments on what MBisanz has said below? The bit about how BCB has less "mission critical" tasks and so that argument against blocking is becoming less relevant? MBisanz points out three recent incidents. Is there any indication that Betacommand learnt anything from any of them? Carcharoth (talk) 16:06, 5 March 2008 (UTC)
          • • Carcharoth, as I stated before, this is/was a task that I have done multiple times in the past without issue. I thought category maintance was covered under my CfD work. Prior runs have been un-eventful and very helpful for getting the missing/improper categories addressed. β^_command 16:24, 5 March 2008 (UTC)
              • Thanks for the clarification (good thing i edit conflicted here, erm...). I support the unblock too, since the task stopped. That being said, we should think of a way to fix the red categories in the future, only (very) popular misspellings would appear on wantedcats. -- lucasbfr ^talk 16:28, 5 March 2008 (UTC)
                • here is a thought, I can generate a list of red linked cats, (I posted about 1/5 in my sandbox), there are ~26k red cats. Weekly I post them to wiki and users remove/create cats that are needed. once the list has been up for a week any remaining redlinks will be emptied. and start doing this weekly. that way we can get wanted categories created and get rid of the useless red link cats. β^_command 16:43, 5 March 2008 (UTC)
                  • We're getting closer to a solution here, except I think many here are very uncomfortable with the last step you propose. Identifying red cats for users is a good move though and I think it could become a maintenance task like faulty ISBN numbers and no-cat articles and such things. Orderinchaos 13:44, 6 March 2008 (UTC)
              • Sorry, Betacommand, I missed that. Let me quote you from above:

                "Ive been removing redlink cats for a while. prior to this there have not been any issues. What I have seen from prior runs is that pages often get improper categories removed. and ones with typographical errors get fixed. in general a lot of good comes from these runs."

                Could you clarify how long you have been doing this task for and how many runs you have done on this task? I know your bot does lots of other tasks, but if you use edit summaries consistently, it shouldn't be hard for you to split your contributions log up into dated sections according to which task the bot was carrying out at that time. So, do you have such a list? That way, an analysis can confirm or deny your statement "in general a lot of good comes from these runs". Would you be able to do that before doing any more runs? Carcharoth (talk) 17:26, 5 March 2008 (UTC)
• As for the proposal: "once the list has been up for a week any remaining redlinks will be emptied" - this sounds like redlinks could be removed that would be useful if corrected, and that is a bad thing. In general though, this is not the right place to propose and get approval for such a proposal. You really, really, need to get the attention of people doing this sort of work and discuss it with them before charging ahead with bot work like this. The discussion here should have shown you that. One further point "the useless red link cats" is not the same as "the red link cats no-one has got around to fixing yet". This is another bot task with wide-ranging effects that needs further discussion in the right venues, and that doesn't mean WP:BAG - that group doesn't have the non-bot knowledge necessary to approve something like this. Carcharoth (talk) 17:26, 5 March 2008 (UTC)

Carcharoth, I was just throwing an Idea up, seeing what the responce would be before making further discussions. It has been a few months since going redlink removal, and Im not sure when it was done last. I can confirm my statements about how it improves things. I know Ive been doing this for at least 9 months. β^_command 17:45, 5 March 2008 (UTC)

Just to be clear, you realise that removal of redlinked categories and removal of categories deleted at CfD are two different things? And that you have approval for the latter, but not the former? And that you need to get WP:BAG approval before doing the former? Yes, yes, and yes is what I think. Though I'm still not sure that deciding whether to remove category redlinks is something that a bot can help with. In my view, it should only remove redlinks from a list of approved removals, but in most cases the number of removals per category is small enough to do by script or manually. Carcharoth (talk) 18:10, 5 March 2008 (UTC)

BAG approval is not entirely on point - its a generally technical review, and requires the approval of only one member - of which Betacommand is one. Avruch ^T 18:53, 5 March 2008 (UTC)

Well, putting aside for a moment the strange appearance of someone approving their own request... (is it not more logical to require a different BAG member to do the approval?) Who approves the non-technical parts of a bot proposal? Carcharoth (talk) 23:15, 5 March 2008 (UTC)

To quote WP:B:

• is harmless
• is useful
• does not consume unnecessary resources
• performs only tasks for which there is consensus
• adheres to relevant policies and guidelines

So I'd say the task's relate policy talkpage would nee to give consensus. A bot reverting vandalism for instance should follow the consensus of the AIV page or a bot signing for users should follow the consensus of the SIG page. MBisanz ^talk 23:28, 5 March 2008 (UTC)

Or in this case, consensus at some categorization policy talk page. Carcharoth (talk) 00:07, 6 March 2008 (UTC)

• Avruch said: "BAG approval is not entirely on point - its a generally technical review, and requires the approval of only one member - of which Betacommand is one."
  • BAG approval is not simply a technical approval. Tasks need to be in accord with wikipedia policy in general (not just WP:BOT), and have consensus. Also, I have never seen BC claim he is approving his own tasks. Gimmetrow 03:42, 6 March 2008 (UTC)

My good faith is stretched very thin

I'm glad that BC has joined in the discussion, but he hasn't allayed my concerns. I don't see sign of an offer by BC to undo the damage, but my main concern is that BC says "I thought category maintance was covered under my CfD work".

I'm sorry, but this is stretching my good faith a very long way. WP:CFD is "Categories for discussion", a process which makes precise decisions about removing, merging or renaming particular categories, and whose outcome is a very specific set of instructions at WP:CFD/W, which are usually implemented very thoroughly by CydeBOT, which creates very clear edit summaries so that there can be no doubt why category links are being removed.

CfD is not "general category maintenance", or removal of redlinks. If I accept BC's statement in good faith, then I find myself rather more worried than if I thought he was being untruthful in his reply ... because I am very worried by the idea of a fast-running bot being operated by someone who can't distinguish between CfD and other things which could be done with categories. This is sort if boundary issue is very important in a bot owner, where I would have thought that the key principle was "if there is any doubt, check" ... and after all the trouble BCB has run into before, I would have thought that BC would be very acutely aware of the importance of seeking clear approval for bot tasks.

I'm also wary of BC's defence that this has been done before. It seems to me to be very unlikely that it was, because this run was spotted very quickly, and while I don't intend to try to scan BCB's hundreds of thousands of edits, a defence that nobody objected last time he ran this unauthorised task[1] is not encouraging.

Something is not at all right here. I can't decide whether BC is a fool or a knave, but I can't see any third possibility and those two both alarm me. --BrownHairedGirl (talk) • (contribs) 17:36, 5 March 2008 (UTC)

I honestly think he means well and is project-minded, but is sometimes on a different plane from the rest of us - at times when I've dealt with him he genuinely does not seem to comprehend the concerns. Orderinchaos 13:44, 6 March 2008 (UTC)

BrownHairedGirl, <OMIT RANT ABOUT FACTLESS PERSONAL ATTACKS> dont call me a liar without proof. I take great offense to that. <OMIT MORE RANTS>. β^_command 17:54, 5 March 2008 (UTC)

Betacommand, please re-read what I wrote. --BrownHairedGirl (talk) • (contribs) 18:01, 5 March 2008 (UTC)

I agree with what BHG has said. I've also read through Wikipedia:Bots/Requests for approval/BetacommandBot (last edited in July 2007), and I don't see approval there for this task. I know others have said this already, but it is alarming that Betacommand still says that he has done this task before. I think he is still (despite people trying to explain this to him) confusing the removal of genuinely deleted categories with the removal of redlinked categories. I would ask that BetacommandBot not be used for any tasks until Betacommand provides a list of the previous runs done with this task. We need to see how widespread the damage is. Carcharoth (talk) 17:52, 5 March 2008 (UTC)

• (through EC) I too agree with what BHG has written. I also view BC's angry response accusing BHG of "personal attacks" as symptomatic of what I view as the root problem: a user who is far too thin-skinned to run such a powerful tool that will, from time-to-time raise the ire of other users. That and the fact that BC treats the code like it's a trade secret, refusing to allow the community at large to view and work to improve the bot's functionality. These are my main issues, as I outlined above, in my previous commentary. Bellwether ^B_C 18:08, 5 March 2008 (UTC)

I second that request: if Betacommand has indeed done this before, then let's see the evidence and decide what to make of it. I also agree that BCB should not be used further until this has been reviewed. --BrownHairedGirl (talk) • (contribs) 18:04, 5 March 2008 (UTC)

Carcharoth here is a list of that the bot has done. β^_command 18:11, 5 March 2008 (UTC)

No. That is a list of all the contributions of a multi-task bot. We are asking for the contributions done under a specific task. How can anyone be expected to find a few thousand edits when they are swamped among other ones. The answer is that the bot operator should make careful records of the start and end of each run under each task, and deposit the records on wiki. So, do you have such records? If not, will you start making such records? Carcharoth (talk) 18:19, 5 March 2008 (UTC)

I was reviewing bot policy and noticed this rather interesting clause Wikipedia:Bot_policy#Restrictions_on_specific_tasks.

As explained at Wikipedia:Categorization of people, certain types of person categories should not be filled or emptied using a bot. Addition of sensitive categories to articles requires per-article manual review, and should not be carried out simply by generating a list from existing Wikipedia content.

So as far as I understand it, Bot are specifically prohibited from emptying categories relating the people, as I believe happened in this matter as a side-effect of de-populating ALL redlinked categories. If BCB is actually approved somewhere to do this, then we need to change the policy to reflect this change in consensus. Betacommand, were you aware of this part of policy when you coded the redlink run? MBisanz ^talk 23:11, 5 March 2008 (UTC)

No unblock until the damage is repaired

As noted above, WP:BOT says that "The contributions of a bot account remain the responsibility of its operator. In particular, the bot operator is responsible for the repair of any damage caused by a bot which operates incorrectly."

I have yet to see any attempt by Betacommand to undo this damage, or any offer to do so. Quite apart from the other issues raised above, it seems to me to be wrong to unblock the bot until Betacommand has at least started to repair the damage. WP:BOT is not a guideline, it's a policy, and it states quite unambiguously that leaving the tidyup to others is not acceptable. --BrownHairedGirl (talk) • (contribs) 08:27, 5 March 2008 (UTC)

The trouble is, as people have pointed out before, that BetacommandBot does a rather large number of tasks. Betacommand has said he is working on separating them out, but progress appears to be slow. I think the best you will get is an agreement (which Betacommand has already given) not to run the bot on this categpory-related task again. I'd like WP:BAG to go further and withdraw approval for the specific task until the bot and/or operator have shown they can do the task properly. As for repairing the damage, it is possible from the contribs to examine the edits manually (there are between 2000 and 2500, so a large but not impossible task). Reverting all edits the bot carried out on this task was one possibility, but now that manual correction has started, this is not really possible anymore. I'll try and help out, as I am interested in a tally of how many of the edits were off-task or inappropriate. Carcharoth (talk) 08:59, 5 March 2008 (UTC)

How can he repair the damage when his bot is blocked? If he attemt to do it from his main account, he will be blocked as unauthorized bot. MaxSem^{(Han shot first!)} 09:11, 5 March 2008 (UTC)

This way. Badagnani (talk) 09:12, 5 March 2008 (UTC)

Demanding to fix thousands of automatic edits manually is completely unfunny. MaxSem^{(Han shot first!)} 09:22, 5 March 2008 (UTC)

That is an interesting opinion. Because, in fact, I was asked no fewer than 6 times today to do this (because Betacommand did not wish to do it, believing it would be too hard). I didn't find it funny either. Nevertheless, we are doing it. I invite you to read the discussion archive and take a look before commenting further about this matter. Badagnani (talk) 09:27, 5 March 2008 (UTC)

One solution to the alleged problem was to revert all the edits. That could be done quite fast with rollback. Gimmetrow 09:30, 5 March 2008 (UTC)

What do you mean "alleged problem"? There clearly was a problem, one that a bot can't fix, and one that people are now manually fixing. You could say the bot action drew attention to a problem (mis-spelt redlinked category tags), but it tried to brush those mis-spelt category tags under the carpet by removing them from articles. Carcharoth (talk) 09:37, 5 March 2008 (UTC)

Noe that it's uncovered, there's no significant difference between manually reviewing Special:Wantedcategories and bot's contibs at User:Barneca/Sandboxen/Sandbox. MaxSem^{(Han shot first!)} 09:49, 5 March 2008 (UTC)

I don't understand what you are saying here. What has been uncovered? Carcharoth (talk) 09:51, 5 March 2008 (UTC)

That BCB was removing redlinked categories blindly. MaxSem^{(Han shot first!)} 09:52, 5 March 2008 (UTC)

So long as the BCB edit is the last edit, it can be reverted with rollback. Even by hand it's possible to do 50 rollbacks/minute. If these 2500 or so edits are such a problem, one grueling hour would restore everything as it was. Gimmetrow 09:57, 5 March 2008 (UTC)

The trouble is that some of the removals were categories that had been deleted after a CfD discussion, though it is unclear whether Betacommand got a list from CfD or whether these were redlinks that had been re-added by others after the normal CfD procedure had removed them. And I still don't see why you are saying "If these [...] edits are such a problem". It is clear there are problems there. It is equally clear that Betacommand is doing nothing to help fix the problems. It is also clear that rollback could fix some of the problems, but would in any case need to be applied with an edit summary (you can do rollback with an edit summary, but you need to set things up properly to do this), something like "rollbacking indiscriminate removal of redlinked categories by BetacommandBot after discussion here - manual checking still required" (with a link to this discussion). As has been said above, this will send most of these category redlinks back to Special:Wantedcategories for manual checking. Carcharoth (talk) 10:52, 5 March 2008 (UTC)

(undent)I think it would still be better to revert any of the bot's removals that are still the last edit to the articles. As I said somewhere above, that way people that are interested in individual articles will help fix redlinks from that direction, while those interested in attacking it from this direction can just go to Special:Wantedcategories, which is much better formatted than the list in my sandbox. If BCBot isn't the last edit, either someone's already reverted, which is good, or someone's already fixed the category, which is good, or someone's made another edit in an unrelated area, which is imperfect, but we can live with imperfection. --barneca (talk) 12:59, 5 March 2008 (UTC)

As the guy who created the proposed Bot RFC process, I had realy hoped BCB would not be the first bot subject to it, given the long-running and wide-ranging span of commentary on its actions. Given that this page exists, I think we could sidestep the RFC given that its still only a proposed process and continue this existing discussion.

In the last month (?) BCB has now been used three times to do un-approved controversial edits. One was the spamming of MickMacNee's talk page, which is it had been done by a human user through special tools like Twinkle or AWB would have resulted in the automatic revocation of those tools and probably a decent block to boot, without the benefit of such a lenghty review and discussion. The second was the revision counter run-up at the mainpage. Bots are given a flag that lets them operate far faster than regular users, because its trusted that they will only perform approved tasks.

Now this third incident of the categories removal. Over the last month or so, there has been a slow-running debate over user categories spread over many pages. I can only assume that BC follows some of the discussion as these boards and would've noticed this as a controversial topic. And given that there was already a discussion at WT:RFAR about whether or not an arbcom was warrented over this bot's actions, I'd say he should've trended towards adhereing to both the letter and spirit of the BOT policy.

In the past I've questioned if blocking BCB was a good idea, given the mission critical tasks it performs. But I now understand that MessedRocker's RFC Bot can be run, albeit inefficiently, as a replacement to BCB. And the Fairuse image backlog has been dealt with well before the March 23rd deadline. So I'm withdrawing my prior comment that any definitive action against BCB should take into account mission critical tasks. MBisanz ^talk 14:33, 5 March 2008 (UTC)

If there's an available replacement for Betacommandbot, it probably should be used instead, especially if the replacement is open source, the tasks it does are well defined, and the code can be audited by others. Once the functionality is correct, the performance problem can be dealt with later. --John Nagle (talk) 18:59, 5 March 2008 (UTC)

Mbisanz, there was a fourth misuse of the BCB: the mass-adding of the WikiProject Interfaith banner to over 3000 articles, discussed late February on WP:AN here. It would seem that whenever someone asks for some automated edits, BetaCommand jumps over himself to help, no questions asked. Not a sign of good judgement. -- llywrch (talk) 00:35, 6 March 2008 (UTC)

That wasn't misuse. It was a good-faith attempt to do an approved task, but came across some false positives. Will ^(talk) 00:39, 6 March 2008 (UTC)

Bot task made redundant

The bot task of clearing those 26,000+ redlinked cats is now being done by humans (with the help of the bots) at User:Random832/WantedCats. Carcharoth (talk) 18:37, 5 March 2008 (UTC)

Fairuse image tagging per NFCC 10c appears to be handled by a new bot per Wikipedia:Bots/Requests for approval/Non-Free Content Compliance Bot. MBisanz ^talk 23:45, 5 March 2008 (UTC)

MBisanz, I am just handing operation of the bot off, I am still the programmer. β^_command 23:46, 5 March 2008 (UTC)

Correct, thats why I said "by a new bot" and not "by a new bot op". MBisanz ^talk 23:47, 5 March 2008 (UTC)

I'm very excited about Wikipedia:Bots/Requests for approval/Non-Free Content Compliance Bot. Hopefully this will result in the best of both world's. Betacommand's bot skills and teams of people to be slightly more organised about actually running the bot. Good luck guys (ST47, Martinp23, and SQL), and I hope it works. I'm just wondering where the name came from? Was it inspired by Wikipedia:Non-free content criteria compliance or not? Carcharoth (talk) 00:06, 6 March 2008 (UTC)

Yeah, it was. --MZMcBride (talk) 00:39, 6 March 2008 (UTC)

I have proposed an ammendment on the approval pageMickMacNee (talk) 00:48, 6 March 2008 (UTC)

Well so much for trying to reach a consensus

And what's with all this "not a vote" crap? The way I had it set up is the same way used for VOTING at RfA and RfB. So, wtf?? Nothing but lip service going on on this page anyway. Let's stop talking about it, and do something about it. This page has gotten so big, no one knows what's going on. A consensus discussion was ongoing about unblocking the bot and what has that lead to? Disregard as the bot was unblocked anyway. I mean, seriously here. Get off the duffs and take action. - ✰ALLSTAR✰ ^echo 18:51, 5 March 2008 (UTC)

RFC is that way, which is the traditional way of actually geting something done about these sorts of things. I think that many admins are relunctant to take these problems seriously since BCB does most of the NFCC image tagging, which is an extremely important task. If we create another bot that does the same thing (with open source code that anyone can review and implement, that groups notifications together on user pages and takes other feature requests, etc.) then BCB would been less indespencible. I'll start working on the problem tonight, and I hope some experienced bot operators will take a look at it as well. There's nothing technically hard about non-free image tagging, we just need to get cracking on it. ➪HiDrNick! 21:00, 5 March 2008 (UTC)

User:BetacommandBot/Non-Free Template Useage may be of use. Or ask Betacommand himself for the list of non-free templates he uses to track non-free use. Carcharoth (talk) 23:26, 5 March 2008 (UTC)

Actually, scrub that, see Wikipedia:Bots/Requests for approval/Non-Free Content Compliance Bot. Hopefully that will help smooth things over, and maybe it will even see some regular schedules being planned for future runs on other operations. Carcharoth (talk) 00:01, 6 March 2008 (UTC)

First thing, concensus is very much against role accounts. "Shared with some members of the BAG that BC approves" is not the same thing as open source. Any sufficently motivated vandal can find scripts to edit Wikipeida in a hurry; making the source availilble is not WP:BEANS, the cat is already out of the bag. And "as long as the users in question follow my directions" does not bode well for reason. ➪HiDrNick! 00:39, 6 March 2008 (UTC)

HiDrNick, its not a role account. it is a bot that is operated by trusted users. its just a group of users who all press the same start button. β^_command 00:54, 6 March 2008 (UTC)

I have proposed an ammendment on the approval pageMickMacNee (talk) 00:48, 6 March 2008 (UTC)

and that amendment will rot in hell. β^_command 00:54, 6 March 2008 (UTC)

I'm getting the distinct impression you actually want people to take you on at Rfc rather than try and deal with the bot as a bot without consideration of your people skills. I did have an article project in mind for tomorrow but if you really push me I can set it aside and draw up some evidence. MickMacNee (talk) 01:15, 6 March 2008 (UTC)

And you seem to have some sort of vendetta against Betacommand. How would you like if a user was constantly following you around and being a jackass about it? Maxim_(talk) 01:21, 6 March 2008 (UTC)

• Check your facts, I haven't entered a BC discussion for days, ever since his preferred method of communication on WP, or lack of it, became quite clear, yet it hasn't stopped the regular traffic about him, nor did it not occur before I came along with a good faith attempt at a change of a bot. You will also find when checking that separation of responsibility for NFCC tagging from his bot was one of the first proposals I made in the page that was derailed, stalled at Mfd as an 'attack', and now courtesy blanked because it somehow emotionaly disturbs betacommand to discuss his bot's operation, only for much of its original suggestions to resurface here again without any poking from me. Ever wonder why? This never was about him personally, but he has been determined to move it that way whatever it takes, from day 1. MickMacNee (talk) 01:33, 6 March 2008 (UTC)

• The only one "being a jackass" in the above conversation is BC. I mean "and that amendment will rot in hell"?!? And you take his side? Wow. Bellwether ^B_C 01:26, 6 March 2008 (UTC)

Umm, and that amendment will rot in hell ? Nice one Betacommand. Seriously. - ✰ALLSTAR✰ ^echo 05:04, 6 March 2008 (UTC)

Not nice but I feel that the amended proposal is not nice either. It seems to suggest that if this bot starts screwing up, either Betacommand loses his BAG status, he ends up releasing his code, or both. MickMacNee, I do not feel any of these requests are reasonable at all. One has not even seen this bot in action yet to see if this causes the same amount of ire as BetacommandBot. I do believe Betacommand is quick to fix problems with his bot in terms of debugging - need we demonise him further? x42bn6 Talk Mess 05:10, 6 March 2008 (UTC)

It is intended to allow code changes agreed by the operators to be implented, otherwise, what is the point in handing over operation, and hence the point of contact for communication about the bot, to these operators. They are then basically just button pushers, just a front for betacommand. In that case what's the differene between this new bot and bcb just creating a separate NFCCbot run by himself? Nothing was said about removing his status, just a recognition that should betacommand stop cooperating with the operators of the new bot, then as the declared code owner, the existance of the separate bot becomes moot. MickMacNee (talk) 12:11, 6 March 2008 (UTC)

Hmm, it suggests to me that if the group of ops running it request a change, the change be made or the bot be taken offline. I don't see what's wrong with that? Is there a situation where this group and BC would differ so much over a change that a showdown like this would even happen, and if so, do we want one person overriding the will of the people trusted to complete this work? I think the text can be changed such that if the group votes to share the code with somebody whom he refuses to share, he can simply take the bot offline. The question is whether or not this group will be just a group of yes-men or an actual independent body; if it's the former, it's really not going to fix anything.

I also think striking out some of the sections is absolutely warranted given the tone it takes towards editors who have criticized the process; several people here here would freak if "Betacommand and his bot are vilified for the work they do tagging images" was written from the opposite POV: "Unhelpful messages and dismissive behavior have poisoned public opinion of Betacommand and his bot." And "Anti-fair-use people would be shielded from attacks. People will find it much harder to demonize or blame if there's a trusted group of BAG people versus a single individual" is needlessly confrontational and simply invites conflict. Editors critical of the BCB's operation already face a group of people whenever they comment. It wouldn't be anything new. I also don't think it's accurate to categorize them as "anti-fair-use people". These are simply unnecessary to the proposal and would not change anything if removed. —Torc. ^(Talk.) 05:39, 6 March 2008 (UTC)

Community proposal

This does not address article space, only removal of redlinks/deleted categories from user space, including user page and user talk page.

Please read Template:Bots, and then after having read it: Should recognition and heeding of the {{bots}}/{{nobots}} tags be implemented into BetacommandBot's code for user pages and user talk pages? In other words, if BetacommandBot comes across a user page or user talk page that has the {{nobots}} or {{bots|deny=BetacommandBot}} tag on it, do you think BetacommandBot should have the capability to recognize those tags and if so, actually recognize those tags and perform its functions based on those tags and do you think this should be implemented?

There's no need to say "in userspace only" or "user and usertalk pages only", etc. because that's what it says above in the description of this proposal.

Support

1. Support - ✰ALLSTAR✰ ^echo 05:54, 6 March 2008 (UTC)
2. A no-brainer. -- Ned Scott 06:01, 6 March 2008 (UTC)
3. Support For user and usertalk pages only. MBisanz ^talk 06:17, 6 March 2008 (UTC)
4. Support - Bot actions on user and user talk pages should probably be opt-in anyway. —Torc. ^(Talk.) 07:05, 6 March 2008 (UTC)
5. In userspace only. Will ^(talk) 09:00, 6 March 2008 (UTC)
6. I can see the predictable "But no-one TOLD me there was anything wrong with the image!!" from some, but it's their choice to place nobots, so Support. The template could be more precise about images (per Zscout below) Black Kite 10:30, 6 March 2008 (UTC)

Questions

1. Will this affect on his notifications of fair use images? If so, then there should be a way people know about changes to their images (I assume having images on auto-watchlist when uploaded would be it). User:Zscout370 ^{(Return Fire)} 07:08, 6 March 2008 (UTC)

   Well users do have to add the nobots tag, so in theory they should know they are foreclosing the option of a bot notifying them of things related to their editing. Auto-watchlisting is good, but I suspect those who care about their images will just forgo nobots and those who don't, won't follow the watchlist closely. MBisanz ^talk 07:24, 6 March 2008 (UTC)

   As long as users who do that understand that point, then I will support the proposed idea. User:Zscout370 ^{(Return Fire)} 07:39, 6 March 2008 (UTC)

   I added some rather specific language at the template page reiterating that. Others feel free to re-word. MBisanz ^talk 07:44, 6 March 2008 (UTC)

   Users should be able to set permission for addition of material (i.e. notifications) and deletion separately. —Torc. ^(Talk.) 11:04, 6 March 2008 (UTC)

2. Sometimes, users doing mass-bad uploads appear at WP:AN/I but it is usually only though the number of BetacommandBot warning messages on their talk pages. {{nobots}} or similar will prevent those messages. Is there a way to find out the number of bad uploads by user, or problematic users, or something similar without their talk page? (I appreciate that some users simply remove these warnings, but {{nobots}} prevents any edits in the history in the first place). x42bn6 Talk Mess 11:45, 6 March 2008 (UTC)

Block of BetacommandBot reinstated

I have just reinstated the indefinite block on BetacommandBot (talk · contribs).

The owner, Betacommand (talk · contribs) (aka Betacommand2 (talk · contribs)), remains in breach of an explicit point of WP:BOT policy, viz. "The contributions of a bot account remain the responsibility of its operator. In particular, the bot operator is responsible for the repair of any damage caused by a bot which operates incorrectly." Betacommand has not asisted in any way in undoing the damage caused by the bot's mass removal of red-linked categories, and should not have resumed category-related work before fixing the daamge already caused.

The bot has resumed editing of articles to remove categories, without clearly indicating what it is doing and why. Some of the category removals contain no indication of why the bot is doing this (e.g. [2]), while others say "per CfD" without indicating which CfD.

The lack of explicit labelling by BetacommandBot of category-related edits is a notable contrast with Cydebot. Cydebot regularly does CfD work, and uses much more useful edit summaries, such as this one which is part of a series clearly labelled "Speedily moving". Where there is a normal, non-speedy CfD, Cydebot uses an edit summary which explicitly links to the discusion, as in this list of edits.

Now that Non-Free Content Compliance Bot (talk · contribs) hasd been authorised to do the NFCC work (see WP:BRFA/Non-Free Content Compliance Bot), there is no longer any reason to consider BCbot's work as being so critical that the bot should not be required to comply with policy. Cydebot does the CFD-related work more effectively, so BetacommandBot is not required for that role — it may be regarded as a useful backup, but it is not essential.

An indefinite block is not. of course, an irrevocable block. A discussion has been underway above on conditions for BCBot's unblocking, and I suggest that discusion should be allowed to continue until a consensus is reached. --BrownHairedGirl (talk) • (contribs) 13:41, 6 March 2008 (UTC)

BrownHairedGirl, you have zero ground for blocking. those categories in question where renamed/deleted per valid CfD/U request. I see nowhere in WP:BLOCK that admins are allowed to block for no reason. β^_command 13:50, 6 March 2008 (UTC)

Re-read para 2 above: The owner, Betacommand (talk · contribs) (aka Betacommand2 (talk · contribs)), remains in breach of an explicit point of WP:BOT policy, viz. "The contributions of a bot account remain the responsibility of its operator. In particular, the bot operator is responsible for the repair of any damage caused by a bot which operates incorrectly."

The bot is being used in breach of a fundamental part of bot policy. --BrownHairedGirl (talk) • (contribs) 13:53, 6 March 2008 (UTC)

Bullshit, this this was very standard WP:CFD/WU work. your just pissed that your previous block did not hold water, so you make some un-founded bullshit up to block the bot? β^_command 13:56, 6 March 2008 (UTC)

Erm, how is he to repair this damage if the Bot is now blocked? Was it causing any further damage in your opinion? WjBscribe 13:54, 6 March 2008 (UTC)