Jump to content

Wikipedia:Bots/Noticeboard/Archive 7

Add links
From Wikipedia, the free encyclopedia
Archive 1←Archive 5Archive 6Archive 7Archive 8Archive 9Archive 10

Blackout January 18th 05:00 UTC

As many of you know, English Wikipedia will be blacked out during January 18th 05:00 UTC - January 19th 05:00 UTC to protest SOPA and PIPA. As a result, editing via the API will be disabled (see here). This thread is not meant to air opinions on the blackout, but rather to gather information on how bots may be affected. Shubinator (talk) 06:19, 17 January 2012 (UTC)

Also, there's a related thread over at Wikipedia:Village pump (technical)#SOPA blackout. Shubinator (talk) 06:25, 17 January 2012 (UTC)
Bots that edit on a scheduled basis throughout the day (possibly read-only bots too, but it depends on implementation) will need to be able to handle related errors so that they don't crash and are thus available when the site reopens. --Rschen7754 06:25, 17 January 2012 (UTC)
The Toolserver Mirror will stay up, right? Tim1357 talk 22:19, 17 January 2012 (UTC)

Wolterbot cleanup

Since Wolterbot is dead with no indications it will return and we now have a toolserver tool to do the task can I use my bot to cleanup all the {{User:WolterBot/Cleanup listing subscription|banner=''Insert your favorite WikiProject here''}} cruft still left lying about the WikiProjects> --Kumioko (talk) 03:01, 23 January 2012 (UTC)

Since knowone seems to have any problems with this I will go ahead and submit a BRFA to get these cleaned up. --Kumioko (talk) 17:50, 24 January 2012 (UTC)

Call for Participation: Looking to Interview Bot Community Members

Greetings-

I am a graduate student at the University of Oregon, currently collecting data for my dissertation on Wikipedia editors who create and use bots and assisted editing tools, as well as editors involved in the initial and/or ongoing creation of bot policies on Wikipedia. I am looking for members of the bot community to interview regarding their experiences on Wikipedia and opinions of technical and governance issues on the site. The interview can be conducted in a manner convenient for you (via an IM client, email, Skype, telephone, or even in-person) and should take approximately 30-45 minutes.

Your participation will help online communication researchers like me to better understand the collaborations, challenges, and purposeful work of Wikipedia editors and programmers like you.

My dissertation project has been approved both by the Institutional Review Board (IRB) at the University of Oregon, and by the Research Committee at the Wikimedia Foundation. You can find more information on the project on my meta page.

If you would like to participate or have any questions, please contact me directly via email or by leaving a message on my talk page. Thank you in advance for your interest.

Randall Livingstone

UOJComm (talk) 00:05, 25 January 2012 (UTC)

I would urge those who can take part in this but haven't yet to do so if possible, it is an interesting project but the response rate, from what I understand, has been rather low. It really would help Randall if more people would reply, and he's made himself available on several ims, email, etc, shouldn't be too hard to answer a few email questions at the very least, if possible :) Snowolf How can I help? 00:10, 25 January 2012 (UTC)

ClueBot III is going nuts

And I hope I finally found the right place to tell someone about this. It's been nuking the talk page over at Talk:Johnny Otis and archiving discussions that are just days old and should still be up. It started doing this about a week and a half ago when some of the discussions were less than a day old. From its contribs list I think it's been doing it on other talk pages too. I don't know if there's an error with something on the page, or if it's got a bug, but someone needs to find out what is going on. If it does it to the pages I'm watching before I hear back from someone I'm going to take it to the vandalism noticeboard. I know that's not proper procedure, but it's disrupting the site, or at least portions of it. Evanh2008, Super Genius Who am I? You can talk to me... 08:12, 29 January 2012 (UTC)

In the absence of an {{{age}}} parameter, the bot was archiving as many sections as it was allowed to, as soon as it was allowed to. This would have been happening since October 2011. Happy‑melon 08:48, 29 January 2012 (UTC)
Thanks for clearing that up! That was immensely frustrating, especially seeing as how the only code I speak is BBCode. Evanh2008, Super Genius Who am I? You can talk to me... 08:50, 29 January 2012 (UTC)

Wikipedia:Bot Approvals Group

This isn't really a big deal but I just had a perception of the Wikipedia:Bot Approvals Group page that I wanted to bring up. There are several members under the inactive section that should, IMO, be moved up to at least the Semi-active list. Several make at least occassional Bag related edits and others are still active in Wikipedia even if not in bag itself. Regardless of the criteria it appears the list might need to be updated a bit. --Kumioko (talk) 15:03, 30 January 2012 (UTC)

I agree that those who edit BRFAs (with their BAG hat on) should be moved up the list. However, editors who are active elsehwere should not; it is not the purpose of the list. - Jarry1250 [Deliberation needed] 18:59, 30 January 2012 (UTC)
I agree with Jarry. Inactive should be defined as inactive within the Bot Approval process. Snowolf How can I help? 19:00, 30 January 2012 (UTC)
Fair enough, Kinda what I thought too actually. I just noticed that there are 3 or 4 that make an occassional appearance (such as yourself Snowolf) that might be worth putting in the Semi list. Not a big deal really I just thought I bring it up and ask. --Kumioko (talk) 19:24, 30 January 2012 (UTC)

How many BRFA's can I hve pending at one time

I just wanted to ask how many BRFA's I can or should have open at one time. I had previously thought that bundling like items into one BRFA would be better but after seeing how one of mine has been dragging on I think that was an incorrect assumption. I also tried to do a general one along the lines of Tagging and assessing aarticles but I was told I needed to be specific. So, since it appears that submitting individual tasks is more likely to get approval in a timely manner, and given that I have a list of about 40(and growing) that I am sitting on. I thought I would ask before deluging the BRFA process. Just for clarification, several of these are group items. For example: I have about 30 groupings of articles that need to be tagged with WikiProject banners for various projects (1 for most US Supported States, a couple cities and several others), 5 tasks relating to Main page edits, 3 that relate to cleaning up some items in the File namespace and a few more miscelaneous ones relating to Infoboxes, persondata, categories and portals. --Kumioko (talk) 18:38, 2 February 2012 (UTC)

Since knowone has responded then I will assume there is no limit so I am going to start submitting them per previous discussion in my previous BRFA's that its better to submit multiple requests than to grant a blanket BRFA to do things like WikiProject tagging. --Kumioko (talk) 23:38, 5 February 2012 (UTC)

Im a bit disappointed

I must confess that I am finding myself a bit disappointed in the bot process. I am sure some will think I am just a jerk and out of place but I believe I am only going to state a perception that many have of the process here. It should not take a week to get a response nor should it take a month or more to get approval or denial for a request. If the process is that much trouble then maybe we need to just submit bot requests through the Village pump and go by community consensus. I have more than 40 (and growing) tasks I want to submit but it has taken more than a month for the 2 tasks I have open and there are at least 10 BRFA's currently open. I am trying not to swamp the process and only do a couple at a time but if it takes months to get approval I may as well submit them, swamp the process and wait. If the BAG wants to continue to force users to go through this process then they need to be more responsive or we need to find a better way to handle it. We shouldn't be forcing users to wait weeks and months to get approval for, in some cases, a few hundred entries. I am even more troubled in that several members of the BAG do dozens to hundreds of edits per day but they never find the time to contribute to any of the BRFA's or if they do its a casual comment. Not even counting mine there are a couple of tasks that are simple and should have probably been speedy approved once a minimal trial run was done.--Kumioko (talk) 02:29, 6 February 2012 (UTC)

Sorry, I actually had intended on trying to help fix this trend in BAG, but real life takes over, and it is hard to get to things in a timely manner. In my experience BAG is one of the most unrewarding jobs on wiki, and one of the hardest approval processes to get right. Some Brfa's can be very hard to judge, particularly because they tend to suffer from wall of text syndrome, and complex tasks generally require a significant amount of reading (not just the Brfa it's self, but also the policy pages, and other discussion pages). And that is not even getting into the technical side of things. Even then, there is no guarantee you will get it right. BAGers tend to err on the side of caution and avoid the long and difficult Brfas, and just deal with just the easy ones. Moving to a village pump system would just create even bigger problems. If you want to help, I'd suggest you join BAG and dive in.
I would also like to note, that BAG has improved significantly at keeping the Brfa process running smoothly. Currently we only have 11 Brfas open. At points last year we had 30 Brfas open. This is an awesome change, and I'd like to thank all the BAGers (and non BAGers) who have been actively involved, because even small edits or comments make a huge difference to the approval process.
Go ahead and submit all your tasks at once, it will be faster that way. --Chris 10:07, 6 February 2012 (UTC)
Sorry, at one point we had 49 active Brfas. I think we've improved quite a bit since then. --Chris 10:13, 6 February 2012 (UTC)
Everyone's a freakin' volunteer, and has real lives. There's not a time limit on this project. Chillax, and don't insult people like this. (talk→ BWilkins ←track) 12:11, 6 February 2012 (UTC)
I agree its unrewarding and hard but it shouldn't be a one person job either. Frankly I don't really think I'm qualified to be on the bag group. I'm not that great of a programmer and some of the tasks amaze me at how they work. I'm also not an admin. I also don't intend this to insult anyone editor but the process itself. I realize everyone is a volunteer, but so are those of us that submit these tasks and are trying to get stuff done. To spend to time in submitting them and writing the code and even in identifying the task to have to wait 2 months to get approval to fix the problem is disrespectful of our time. BTW I truly hate the term there's no time limit. It gives me the perception that people don't care and that I'm just wasting my time. Your right there's not a limit but when I have about 40 or so tasks that I want and need to get done to support WikiProject US so that the project runs smoothly and the articles get maintained and cleaned up (particularly the less edited ones) then it affects not just me but the projects I am trying to help support and the articles they support. --Kumioko (talk) 12:19, 6 February 2012 (UTC)
From what I have seen most of your tasks so far could be done by you on your own account with AWB. If you are that desperate to get to them fast why don't you just do that? Frankly I would be upset if BRFAs only took a week or less. Most complex bot tasks should take a month or more so everything can be sorted out so there is less risk once it goes live. -DJSasso (talk) 13:17, 6 February 2012 (UTC)
In addition to the insightful comments already made, I would add that operators seeking a faster resolution of their BRFAs should (in general) make as much of an effort as possible to demonstrate consensus for their tasks. BRFAs suffer most from a lack of outside input; when it does come at the moment, it tends to come in the form of a vitriolic argument between two opposing sides that haunts the BRFA for weeks. By comparison, generally speaking, coming to BRFA with a demonstrated consensus makes the whole exercise far easier and quicker. BAG members are generally very reluctant to approve even seemingly simple tasks if that consensus has not been demonstrated, and that is what holds most BRFAs up IMHO. - Jarry1250 [Deliberation needed] 14:50, 6 February 2012 (UTC)
Agreed. Kumioko, regarding the current request you have that's open, it's being held up a bit in my opinion due to the task's failure rate, the lack of clear evidence of consensus, and the fact that some of your tasks have caused problems in the past. We have to determine that an automated task is going to be harmless and that takes time, ironing out bugs, and sometimes multiple trials to demonstrate.
I absolutely would not support any sort of time limit on BRFAs because the longer they stay open, the better we can gauge consensus. In addition, sometimes we need to discuss our concerns about a task and sanity check each other, and that takes time as most of us are in different time zones and thus are not active at the same times. In point of fact, that's why I stepped back from handling your request myself, because I would be inclined to deny it, but I'm willing to discuss it with the rest of the BAG and/or let another member handle it.
Finally, the fact that we have 10 BRFAs open means little when most are in trial or extended trial. Our response time once an operator indicates a trial is complete is generally very good (even given that it takes time to check the results of the trial). — madman 19:58, 6 February 2012 (UTC)
Reply to Djasso - In general you are right, I could do them from my account using AWB and I have in some cases, however, it is my understanding and experience that to do several thousand edits should be done as a BRFA and frankly I have a lot going on myself and don't have the time to hit enter 10, 000+ times in a shot.
Reply to Jarry - I believe that with the exception of one all the tasks I submitted have been approved by other operators so the consensus is already established. I have quite a few in the hopper that aren't and intentionally waited to submit them until I got consensus and until I established the bot and the tasks I thought were were easy to get approved.
Reply to Madman - First I partially agree and I believe that I also tried to do too much with one BRFA. I did that in the thought that if I grouped several like items I could get several done at once. I won't do that in the future and will submit each task individually so that many aren't held up by one. I admit I had a few hickups with the coding but most of them are minor, were easily fixable and most are due to the inherent complication of the structure of the WikiProject templates. Which by the way all the other bots with approval to add or modify WikiProject templates also have. I have yet to see one that doesn't have some problems such as wiping out additional parameters, removing the wrong project, adding the wrong project, breaking parameters, etc.
Just to clarify to all, knowone expects an immediate response but when 5 or 10 days go by without anyone stopping to check on things and make a comment its gets very frustrating. --Kumioko (talk) 23:34, 6 February 2012 (UTC)
As I said, I wasn't responding with regard to your BRFAs specifically, just BRFAs in general. Thanks. - Jarry1250 [Deliberation needed] 00:16, 7 February 2012 (UTC)
I do recognize that parsing templates is one of the most complex tasks a bot can do due to their "loose" syntax (it's like HTML in that respect; it cannot be parsed perfectly using regular expressions). I didn't mean any of what I said as criticism. :) When I wanted to make sure that my bot, when faced with that task, would perform it harmlessly, it took me about a week to rip out MediaWiki's parsing code and adapt it to my bot so it'd parse it the same way MediaWiki would and couldn't break anything. I definitely don't expect that of most bot operators. :P More complex tasks are going to take a little more time to test and approve though. Once it's clear you've gotten the bugs ironed out of the process, related requests are going to be approved much more quickly.
Also, you can combine multiple tasks into one request; I don't think anyone meant to imply you couldn't. You just can't have really open-ended requests like "Modifies WikiProject banners" because there's no way for us to gauge consensus of what tasks that will include in the future. — madman 02:27, 8 February 2012 (UTC)
  • As a comment to BWilkins about "volunteers". Yes BAG members are volunteers and so are bot operators. Therefore we should not demand the impossible from BAG members but we should also not make bot operators wait too long. On Commons there is a rather "free" bot policy. If you are an experienced and trusted bot operator you can do almost everything as long as the task is not expected to break anything or make someone upset. On enwiki you have to ask permission to almost everything before you start your bot. Perhaps the bot policy on enwiki could be a little less strict?
As an example: I nominated 1.100+ files for deletion and I wanted to add a ffd on the file page and leave a note to the uploaders. I was told that I could not do that without asking for a permission. Later a few hundred of the files was withdrawn and I could strike the files from the DR but I could not use my bot to remove the ffd template from the files without asking for a permission. And if I got a permission to do the tasks then what about next time I nominate a lot of files for deletion? Would I need a new permission or could I use the old one? But it was just examples and the problems were fixed so lets just keep it to a general discussion.
My point is that if it is fast and easy to get a permission to do a task there is a much better chance that someone volunteer do do the tasks requested here or anywhere else. --MGA73 (talk) 14:26, 7 February 2012 (UTC)
To followup on that, Many of the tasks I am asking for currently and will be asking for in the future are relatively straightforward wikiproject tagging but I have been told I need to resubmit for each new one. Even if its speedily approved it adds additional delays each time. So assuming I submit 2 per week and they are speedily approved it will take me at least 6 months to get them all approved. --Kumioko (talk) 15:17, 7 February 2012 (UTC)
If you want to promote a relaxation of bot policy then do so, but be aware that the tight policy we operate here on en.wp was borne incrementally from literally dozens of massive flare-ups where fairly obvious (it seemed at the time) tasks turned out to be highly controversial. (Also, there's a degree to which best practice is helpfully shared by BRFAs that would be lost.) - Jarry1250 [Deliberation needed] 15:21, 7 February 2012 (UTC)
I know there is a reason not to allow "everything" and I'm not suggsting to do major changes. I'm just suggesting that perhaps it would be a good idea to take small steps in direction of a more "loose" bot policy on some areas. --MGA73 (talk) 15:41, 7 February 2012 (UTC)
Yeah, I'm not unsympathetic, but one has to realise that the current bot policy started off relatively lax, and has been tightened incrementally following massive arguments. Thus to "loosen" bot policy, one has to roll back the conclusions of those set-pieces. - Jarry1250 [Deliberation needed] 15:43, 7 February 2012 (UTC)
There's a slightly unwritten rule, that you can run tasks under your main account, as long as you don't get caught (i.e. small task, slow edit rate, something with an already established consensus, NO error rate). Likewise, you can to some extent test bots under your main account as long as you monitor them. The main factor in this though, is that if you screw up majorly, you'll be in a bit of hot water (especially, if it becomes a habit). --Chris 15:48, 7 February 2012 (UTC)
I'm not sure that's an unwritten rule, nor that we should say it is; it's covered by the bot policy. Any task that's not fully automated (e.g. assisted editing) can be performed under your main account. (You could create an alternate account but I recognize that that's a bit of a bother.) All of the conditions you list make such a task more acceptable and lessen the chance of conflict with other editors. — madman 02:27, 8 February 2012 (UTC)
Speaking as someone who was here for The Bad Old Days™, BAG used to follow the model MGA proposes of fairly permissive and broad approvals of experienced operators. Then we had Betacommand, ST47, Geo-tagging bot, Date delinking bot, Plant bot, Bot deletions and a host of other situations which almost led to the disbanding of BAG. To Kumioko and MGA, if you would like to submit a general BRFA for "wikiproject talkpage tagging" or "template normalizing" or "re-categorization of wikiprojects," I would be very likely to grant a broad approval once you had done a small test. Also, even if the rules or convention say 2 BRFAs per person at a time, for the time being, I would not object to upwards of 5 BRFA per person at a time. MBisanz talk 15:49, 7 February 2012 (UTC)
Thanks to all for comments. I have gotten a much better understanding of the process, why it is what it is and a broad understnading of the BAG and BRFA process that I didn't have benfore. --Kumioko (talk) 16:20, 7 February 2012 (UTC)
Yeah like Kumioko I would also like to say thank you :-) As I understand it it is possible to get a more general permission if I/someone want to do a few similar tasks. That will help. --MGA73 (talk) 20:41, 7 February 2012 (UTC)

Kumi-Taskbot and collapse parameters

I'm bringing this issue here because repeated requests to Kumioko (talk · contribs) to stop removing collapse parameters. This still continues and I don't see anywhere in Kumi-Taskbot approvals that collapse parameters should specifically be removed. It happened most recently today after I had asked Kumioko for the second time to knock it off. The original notice pointed out why the collapse parameter, when it has been set to either yes or no, means there is some reason for it's existence. For that matter, why would an unused collapse parameter bother someone so much that they have to systematically remove them? When an article has many project tags, collapsing the banner shell saves a lot of visual space. Looking at this same edit the only corrections made were to white space which I fail to see any approval for. This most recent behavior is only carried over from months of my commenting to Kumioko about what his AWB obsession has been doing to article talk pages and IIRC at one time Kumioko lost his AWB privileges for the very same reason of making mass inconsequential edits to article talk pages. Brad (talk) 02:36, 15 January 2012 (UTC)

Well fme, here are more issues:

Kumi-Taskbot is only approved for work with wiki project templates. Brad (talk) 02:50, 15 January 2012 (UTC)

I stopped removing the collapsed parameter after the second notice although I still contend that it should be removed. The collapsed parameter completely hides all of the WikiProject templates effectively making them invisible unless the user knows to go looking for them. Second, since I do not have access nor do I use IIRC, IMO, if the comments happened in IIRC about me or my bot, frankly they don't exist to me. I had no opportunity for comment nor discussion there. Thirdly, as with any bot there were a few glitches mostly do to the absolute and complete lack of standardization of the WikiProject templates that I had to work through. Which I would like to say are fixed but I continue to make adjustments for as I find them. --Kumioko (talk) 12:16, 17 January 2012 (UTC)
"IIRC" = "if I recall correctly". :) —  HELLKNOWZ  ▎TALK 12:25, 17 January 2012 (UTC)
Yeah I am not sure the collapsed=yes should be used. However I don't think it should be changed by a bot. The boxes are still collapsed without that parameter set...they just aren't completely collapsed so that the names of the wikiproject tags can still be seen. Collapsing them completely defeats the purpose of the tags. -DJSasso (talk) 14:06, 17 January 2012 (UTC)
Thanks H3llkn0wz for clarifying. I was thinking about IRC. To Djsasso, I already removed the code anyway. I am going to bring up the use of the collapsed parameter on the WikiProject bannershell for discussion and see where it goes. --Kumioko (talk) 14:33, 17 January 2012 (UTC)
And you still ignore the fact that you're doing work with your bot that has not been authorized. You removed the parameter from the FAQ template after I asked you to stop removing collapse parameters. There is a particular reason for setting |collapse=no on the FAQ template so that the questions stand out and can be seen. You only have authorization to work with project templates; not make white space changes or changes that you feel should be done. My politeness is wearing thin. Can you not just leave the f-articles alone? Brad (talk) 00:23, 18 January 2012 (UTC)
Well to answer the question at the end, yes I could, but that sorta defeats the purpose of having an online encyclopedia that anyone can edit don't you think? Your dwelling on a dead issue at this point. I already told you I stopped doing that edit. But just for clarification some of the edits that you seem to have trouble with are general edits generated by AWB. Not the collapsed parameter mind you but the white space edits that seem to be so troubling to you and the only reason they went is because I didn't check the little box in AWB that says skip if white space/minor edit. --Kumioko (talk) 00:35, 18 January 2012 (UTC)
It should be the encyclopedia that anyone except Kumioko using AWB can edit. Since you're a responsible bot owner I'm sure that by now you are planning to go around and replace the parameters you removed. I saw another one of your edits over at George Washington that clipped a working collapse parameter. I'll have to start putting that deny tag on more articles as I'm sure this won't be the last of your escapades. Brad (talk) 18:08, 19 January 2012 (UTC)
Well its ok that you think that, doesn't really matter to me that you personally don't like my edits, but its ok. I learned along time ago that you can't, and shouldn't try, to please everyone on here. Since I don't know precisely every edit that parameter was removed from, and in most cases it shouldn't have been there anyway, theres no way I can go back and fix them. My guess is that most of the ones you have a problem with relate to US Presidents so I might go back and check those over. If you see one I missed please let me know. --Kumioko (talk) 19:07, 19 January 2012 (UTC)

What's is the benefit of having collapsed=no in every page? Collapsed is set to "no anyway. -- Magioladitis (talk) 18:58, 2 February 2012 (UTC)

If you compare before and after you can see that the default is to hide the questions in the FAQ, and "collapsed=no" makes the questions visible. I'm not sure why the bot was making this change if it was not authorized. — Carl (CBM · talk) 19:26, 2 February 2012 (UTC)
It turns that setting no for collapsed in this banner it changes the visual effect. Thanks for letting me know. -- Magioladitis (talk) 19:43, 2 February 2012 (UTC)

Kumioko is doing it again except this time not under his registered spam bot. In the diff please notice that {{bots|deny=Kumi-Taskbot}} is in place to prevent his bot vandalism. Since he has no authorization to vandalize remove collapse parameters with his bot, he's just doing it without the bot. This is the second or third time this has happened after Kumioko said he would stop. This isn't the fault of AWB, this is a planned removal by the AWB operator. Brad (talk) 19:38, 8 February 2012 (UTC)

I am confused last time you complained that he was uncollapsing boxes. This time he collapsed a box. Secondly as an editor he is allowed to make those kinds of changes under his own account. And you are able to WP:BRD them as well. There is no policy that I am aware of stating that once someone sets one of those that no one can change them. -DJSasso (talk) 19:44, 8 February 2012 (UTC)
Actually reading further I do see that the FAQ was mentioned. I was thinking about the banner shell part. But that doesn't change the fact that any editor can do that on their main account. However since he was asked to stop he should stop and discuss. -DJSasso (talk) 19:45, 8 February 2012 (UTC)
If this were an isolated incident I would not be here complaining. Kumioko is using AWB and a bot to make sweeping changes to articles (removing collapsed parameters) without consensus to do so. This is abuse of both AWB and bot privileges; he does not have authorization or any approval to edit in this manner. Brad (talk) 20:10, 8 February 2012 (UTC)
You gotta be kidding me with this Brad, I made the change, I saw I did it and fixed it, then you reverted my edits fixing the problem to a version you did completely wiping out all of the good changes I did. Then you start forum shopping leaving comments here and on the AWB page calling me out of control. You need to get a grip, grow some thicker skin, stop showing such ownership over the article and stop being a jerk. --Kumioko (talk) 20:14, 8 February 2012 (UTC)
Actually now that I look at the page history that is true...you did fix it and he blindly reverted. That is not cool. -DJSasso (talk) 20:53, 8 February 2012 (UTC)

Bot bugs not being attended to and bot can't be shut off by victims.

It looks like this bot is performing quite a lot of unintentional vandalism (See http://en.wikipedia.org/wiki/User_talk:Elvey#Malfunctioning_bot_annoying_uploaders.21 and such.) Shouldn't all bots have emergency shut-off switches?--Elvey (talk) 08:17, 8 February 2012 (UTC)

There's always the block button. →Στc. 08:52, 8 February 2012 (UTC)
Vandalism is a pretty harsh call. http://en.wikipedia.org/w/index.php?title=File:Macy%27s_stupid_coupon.jpg&oldid=427623492 is the version of the image you uploaded that the bot whined about; that version renders with a complaint that WARNING: Fine print does not appear to exist!, so perhaps there was a parsing issue. Josh Parris 09:14, 8 February 2012 (UTC)
I've tried to figure out what User_talk:ImageTaggingBot#Why_are_all_the_images_I_upload_being_wrongly_labeled_as_untagged.3F is about and have failed; there's no user with that name and any editing history on en or commons. Josh Parris 09:20, 8 February 2012 (UTC)
User:Astepintooblivion probably changed their name into User:Obsidian_Soul. -- Luk talk 09:58, 8 February 2012 (UTC)
Ah. In that case, here's why "as such": http://en.wikipedia.org/w/index.php?title=User_talk%3AObsidian_Soul&action=historysubmit&diff=384077370&oldid=384050156 Josh Parris 10:48, 8 February 2012 (UTC)
I noticed recently that there are a lot of Image swith double and triple of some tags, probably partially because the bots adding it multiple times or adding it in addition to others adding it. If the bot is malfunctioning andn can't be stopped with multiple comments going unanwered then IMO the right thing to do is block it until things get sorted out. --Kumioko (talk) 13:53, 8 February 2012 (UTC)
ImageTaggingBot may add two tags to an image in a situation where an image has two problems (most often, "no source" and "no license"), and it's common for a user to later add a third tag for a violation of the non-free content policy. The bot should never add a tag to an image that another user has already tagged. If you know of some examples, please bring them to my attention so I can adjust the bot's settings. --Carnildo (talk) 03:24, 10 February 2012 (UTC)
The reason why ImageTaggingBot and ImageRemovalBot don't have user-accessible shutoff switches is that users in general are very poor at judging when those bots are malfunctioning. In the case that Elvey appears to be complaining about, File:Macy's stupid coupon.jpg, ImageTaggingBot claimed that the file did not have a license tag. Not surprisingly, the bot was correct: the image had an {{Information}} template and a {{Non-free use rationale}} template, but no license tag. --Carnildo (talk) 02:31, 9 February 2012 (UTC)
I wasn't able to determine from the bot's user page or its BRfA what criteria it uses to determine a lack of licensing. It'd be helpful if that was there. Josh Parris 02:48, 9 February 2012 (UTC)
To the bot, "licensed" simply means "the image description page transcludes a license template". The bot has two degrees of "unlicensed" ("no license whatsoever" and "maybe a license, but no template") that it distinguishes between based on various heuristics, but the difference only influences what no-license tag the bot applies and what message it gives the uploader; in both cases, the image is in violation of policy. --Carnildo (talk) 03:19, 10 February 2012 (UTC)
All bots have emergency shut off switches. Honestly, it's the block function. A lot of bots have non-admin stop features, and that's nice and often appropriate, but it not and should not be required. Think if Cluebot NG had a non-admin stop function :D Snowolf How can I help? 02:40, 9 February 2012 (UTC)
It ... uhm ... does -- Cobi(t|c|b) 22:01, 21 February 2012 (UTC)

Bot-operator has retired?

User:Chzz seems to have left, but User:ChzzBot II is still editing (cleaning the sandbox and whatnot). Does anything need to be done in such situations? It Is Me Here t / c 00:43, 19 February 2012 (UTC)

Do the bots need to continue operating? If no, the toolserver will eventually expire forcing the bots to automatically shutdown.—cyberpower (Chat)(WP Edits: 517,570,498) 00:47, 19 February 2012 (UTC)
Let's be frank here, this is Chzz, he'll be back. Chzz has already retired once this year, and he'll probably retire a few more times as the year goes on. The part of him that wants to help new users and the part of him that gets pissed off about the governance of wikipedia-en-help are constantly fighting each other, I wouldn't worry about it unless he's not back by April. Sven Manguard Wha? 01:31, 19 February 2012 (UTC)
Bot task is already duplicated anyway ·Add§hore· Talk To Me! 15:01, 21 February 2012 (UTC)
Yes, but not all like the informing of new contributors adding new, unreferenced WP:AFC submissions. mabdul 15:05, 21 February 2012 (UTC)

Misza13's bots seem to be down

Has anyone been in contact with Misza13, or know anything else about this situation? Recently-active bots seem to be Wikinews Importer Bot (talk · contribs) and the archiving bots MiszaBot I (talk · contribs), MiszaBot II (talk · contribs), and MiszaBot III (talk · contribs). Anomie⚔ 19:12, 19 February 2012 (UTC)

It would be helpful if these bots ran again. My talk page is cluttering.—cyberpower (Chat)(WP Edits: 517,705,980) 19:14, 19 February 2012 (UTC)
I tried to setup ClueBot III, but it ignores {{DNAU}} and doesn't seem to support some of MiszaBot III's features... Is there any chance MiszaBot III will come back? — Dmitrij D. Czarkoff (talk) 03:14, 20 February 2012 (UTC)
Is it possible that the current code used in the bots is the same as archivebot.py? →Στc. 03:19, 20 February 2012 (UTC)
I mentioned to someone yesterday that Mizsa hasn't edited since May 2011 and his toolserver account is probably expired. I think he left a note on Anomies page. --Kumioko (talk) 03:26, 20 February 2012 (UTC)
His account is not expired. See [1]. →Στc. 03:34, 20 February 2012 (UTC)
(edit conflict)Just checked Toolserver. It's still running which means there's got to be another reason why MiszaBot is not. Has anything changed in the code perhaps? I can access Misza13's repository but I can't find the bot code.—cyberpower (Chat)(WP Edits: 517,780,421) 03:36, 20 February 2012 (UTC)
Oh well it was just a guess anyway. I know those things expire every 6 months or so and figured the timing was about right. Its possible he he just logged in and turned them off. We seem to be having a lot of experienced editors and bot operators walking away these days. Maybe hes just another one. --Kumioko (talk) 03:40, 20 February 2012 (UTC)4
Its possible it was an offwiki program running from his PC. Article alerts works like that. Again, he might have turned it off or maybe the computer crashed. --Kumioko (talk) 03:42, 20 February 2012 (UTC)
I don't think you would have a bot running off of the computer. That's a stupid location to store bot code to run. It's possible that maybe the slayerd task (process slayer) in toolserver terminated the bot. MiszaBot III shut down followed by MiszaBot II almost 24 hours later and then MiszaBot I a few hours after that. I checked all the subpages of MiszaBot and they don't seem to be weird although how it's setup. I'm honestly clueless. Misza13 would have to figure this out.—cyberpower (Chat)(WP Edits: 517,782,202) 03:48, 20 February 2012 (UTC)
It's not as bad as you make it out to seem. I've had bots run on my desktop before -- my desktop regularly has 60-100 day uptimes and plenty of resources to run a bot, and is on a UPS that can run it for hours though. However, I typically move them off to a server once they are stable. -- Cobi(t|c|b) 22:07, 21 February 2012 (UTC)
One of the login servers (nightshade) is down since the 17th. user:BotMultichillT is also down. Multichill (talk) 09:40, 20 February 2012 (UTC)
Very true. Nightshade doesn't seem to want to boot. That would explain it.—cyberpower (Chat)(WP Edits: 517,840,726) 12:24, 20 February 2012 (UTC)
Right, the presumption is that Misza (and some other bot ops) put their bots in a crontab on nightshade. With that host currently down/out-of-commission, any bots tied to nightshade's cron are not running. As luck would have it, my crontab is on willow. :-) --MZMcBride (talk) 23:10, 20 February 2012 (UTC)
What features does CB3 not support that MiszaBot III does, other than DNAU? (DNAU is a horrid hack that only works with timestamp-based bots -- CB3 actually checks the history to figure out when it was last modified, and so that hack does not affect CB3). -- Cobi(t|c|b) 04:15, 21 February 2012 (UTC)
On the other hand you can have CB3 look for DNAU stamps to prevent archiving. On the other hand, Misza 13 says that MiszaBot does operate on Nightshade. Misza 13 said s/he will operate them manually on willow for the time being.—cyberpower (Chat)(WP Edits: 518,046,287) 11:50, 21 February 2012 (UTC)

Toolserver

Does anyone know if there is somewhere that you can find a list of all the different tools you can use to get reports like the edit counter or the tool to check number of hits a page receives.Edinburgh Wanderer 22:37, 21 February 2012 (UTC)

From the FAQ: Where can I view all of the available tools and their functionalities?: Sorry, but you can't. The best you can do is look at the list of users, and click each username to see if they have an index.html listing all of their tools (it's recommended that all users do so, though I know I don't. ) Or you can Google what you want with site:toolserver.org. Hope that helps. — madman 23:52, 21 February 2012 (UTC)
Oh, I've forgotten the very obvious point that operators may list their tools on the Toolserver wiki. — madman 16:01, 22 February 2012 (UTC)

New asin-tld parameter

The {{Cite web}} template (and likely others) now has the asin-tld parameter. Any bots that now use id with the {{ASIN}} template should switch to using asin with asin-tld instead. – Allen4names 18:05, 28 February 2012 (UTC)

Looking for Help Gathering Data on Bots

Hello. Some of you might have run into me before...I am doing a research project on bots, bot operators, and technical tools on WP and WM projects. I'm wondering if anyone wants to tackle this problem, which would help me out tremendously. I am looking for stats and data on bots, especially over time. Things like:

  • (#) of bot accounts registered over time (by month would be fine) (on English WP)
  • (#) of bot edits over time (on English WP)
  • (#) of BRFA approved and not approved over time (on English WP only, obviously)
  • same trends for bot use on other language versions (which would be a bonus)

I've found some info on these things spread around WP, but nothing that is both up-to-date and reasonably accurate/reliable. I'm not sure if getting this info involves dealing with a data dump (I suspect it does), or if there are simpler ways to do it. If you're interested in investigating this with me, I'd really appreciate the help. Please let me know here or on my talk page.

And if you're a bot operator or Wikimedia developer (or someone who deals with the technical infrastructure of WP) and you'd like to be interviewed, please see my call here.

Thanks! UOJComm (talk) 23:44, 17 February 2012 (UTC)

I could cook some raw data for you in areas where it's not too much trouble. I've been wanting to do some stats myself. But don't hold me to it and I probably won't be of any help with other languages and stuff. —  HELLKNOWZ  ▎TALK 09:39, 18 February 2012 (UTC)
An example of BRFAs parse: User:H3llBot/Sandbox2. It's buggy (as hell) and I'll make a proper one and group data when I get some time :) Is this something like what you are looking for? —  HELLKNOWZ  ▎TALK 12:14, 18 February 2012 (UTC)
HELLKNOWZ, that looks pretty good...I think I could compile some good stats based on that. A few questions: Is the date column the date that the bot account was created, or the date that the BRFA was closed? (and is there a way to grab a bunch of dates...the date the account was created, the date the BRFA was initiated, the date the BRFA was decided?) Also, what are the final 8 columns? Some of them make sense, but I'm not sure of all the abbreviations, so a key of some sort would help. You say it's buggy, but is that something you can work out if you get more time? I just want to be confident in the data. Thank you so much for your help! UOJComm (talk) 00:45, 19 February 2012 (UTC)
I can answer as to those last eight columns; they're the disposition of the request (approved, speedily approved, denied, speedily denied, withdrawn by operator, expired, open, or revoked). Also, that timestamp looks like the date the BRFA was initiated. — madman 21:23, 19 February 2012 (UTC)
I can probably do this for you fairly easily; being able to run queries on the Toolserver will help. I'll give you the results sometime next week. — madman 15:28, 18 February 2012 (UTC)
Hi Madman...Thank you very much for your help! Since you query the Toolserver, would you be able to get data on other language versions as well? UOJComm (talk) 00:47, 19 February 2012 (UTC)
I thiiink so. I can get information on flagged bots and bot edits, but as far as BRFA information, the process may not exist on other wikis or I'd have to figure out how to parse approval, denial, etc. on each one. — madman 21:21, 19 February 2012 (UTC)
Yes, I don't think many other language WP have such a formalized BRFA process that English does. So just stats on flagged bots and bot edits on other language versions would be quite helpful. Thank you again for your help with this! UOJComm (talk) 22:42, 20 February 2012 (UTC)
Any list built from the BRFA pages will have its limitations -- it won't include bots approved under the old approvals process, and it won't include RamBot, which started running before there was an approvals process at all. --Carnildo (talk) 01:16, 20 February 2012 (UTC)
Right, there are going to be limitations no matter what, but I'm going to start with data on flagged bots and then expand as information is available. — madman 01:17, 20 February 2012 (UTC)
Yes, there certainly seem to be limitations on any data/stats having to do with bots (or with human editors, for that matter). The best that can be done is the best that can be done, and on my end, I can talk about limitations of these stats as limitations of the study, but also as an illuminating quality of Wikipedia...so much data, and it's difficult to be precise in this situation! Any other comments or thoughts on limitations are surely welcomed! Thanks! UOJComm (talk) 22:45, 20 February 2012 (UTC)

If anyone else is interested in this data, it's posted here with a summary of the bot edits data here and the bot registrations data here. I find it interesting, but then I'm the type who loves playing in Excel. :) Cheers, — madman 05:02, 4 March 2012 (UTC)

DASHBot breaks some of the references (4% of my watchlist sofar)

In the past few days, the bot DASHBOT has made four mistakes in pages on my watchlist:

What went wrong was that an archived link was added to the wrong reference. After I noticed the first mistake, I notified the bot owner User:Tim1357, again after the second mistake, and a third time after the following two mistakes. As far as I know, Tim1357 has not taken any action yet. (No judgement on that, we're all volunteers.) Should the bot be given a temporarily break, to find out what goes wrong? It is probably just a detail, but it's breaking some of the references. --EdgeNavidad (Talk · Contribs) 22:33, 2 March 2012 (UTC)

I was literally about to shut it off to give Tim some time to investigate without more errors being introduced, but in the meantime he's shut it off himself, so kudos to him for that. I suspect he'll post here shortly, the bot was in clear error, after all. - Jarry1250 [Deliberation needed] 23:43, 2 March 2012 (UTC)
I am investigating this. Thanks so much for your patience. Tim1357 talk 01:33, 3 March 2012 (UTC)
Gameplan

Jeez. This sucks. I have no idea how I missed this. I checked about 70 edits before I started letting the bot off it's leash. I owe an enormous thank you to Edge for bringing this up. Here's my plan to fix this:

  1. Get a list of every article my bot has touched. Green tickY (There's about 2 thousand)
  2. Scan the articles to find any instances of errors. Green tickY 51 Articles
  3. If there's only a couple, fix by hand. Otherwise, write something to fix them, or revert en-masse. Green tickY
  4. Take a nap.
  5. Take a serious look at what happend, and why. I have a good idea of where the error came from, but I'll do multiple dry runs on revisions of the articles that the bot previously screwed up on.
I'll be posting updates here as I work through the list. Tim1357 talk 20:20, 3 March 2012 (UTC)
Ok, I made 45 edits. Some of the edits I reverted outright, and some of them I salvaged. Tim1357 talk 21:26, 3 March 2012 (UTC)
Kudos for an excellent response. — madman 00:48, 4 March 2012 (UTC)

This seems like strange and preferential treatment

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.
While perhaps a trial should have been requested again for the Addbot 24 BRFA before it was approved, the details of this BRFA are dissimilar to the details of the Cyberbot I BRFA. BAG members are expected to consider all of these details in handling approvals, such as the bot operator's experience, demonstrated understanding of the bot policy, prior requests for approvals, the scope of the task, &c. The question that prompted this discussion has been answered and the discussion itself is growing tiresomely off-topic.

I have to admit I find it a little bit strange that a big issue was made out of this when the operator accidentally ran it without approval. Yet, when this bot ran without approval it was discussed on IRC, which most editors don't use, and then approved and allowed to go on. This doesn't quite seem equal treatment. 71.163.243.232 (talk) 01:20, 4 March 2012 (UTC)

It would be nice if you would sign your comments with your actual username. And if you read the BRFA, the fact that he run without approval was actually not even a serious concern. Snowolf How can I help? 01:25, 4 March 2012 (UTC)
(edit conflict) Interesting definition of "big issue". The fact that the bot ran without approval was only mentioned twice in the BRfA that I can see, once by Snowolf saying "I noticed the bot is currently operating without being approved for trial. Could you explain that, please?" A question which was never answered. And once by me, saying "... running before approval, which also suggests a lack of bot policy knowledge..." as part of a larger list of various problems. Now I don't really see that as it being made a big deal of, but feel free to correct me. - Kingpin13 (talk) 01:29, 4 March 2012 (UTC)
The bot task had already had 3 BRFAs previous to number 24 which has been linked to. (21,22,23). The task has also already had two trial runs from these BRFAs. The BRFAs have also involved much discussion of the task. Once I realised that the bot was operating without a valid approved BRFA I removed the task from cron and the bot did not run again until approval. ·Add§hore· Talk To Me! 01:34, 4 March 2012 (UTC)
Well I don't really have the time or inclination these days to embroil myself in another useless discussion. I just wanted to make a statement that it seems that the good ole boys do what they want when they want and then pick and choose when to enforce the policy, that's all. One case was an open and shut BRFA with almost no trial and the other seemed to have unlimited comments until they finally left the project completely. Yet another user pushed out. We are losing a lot of experienced editors just because we don't enforce policy evenly and fairly. Regardless of whatever that policy might be. As for the IP/Username comment. Either the comment has merit or it doesn't, being made by an IP or a registered user makes no difference. I am sure that you all will just say I am spiteful or butt hurt or whatever. Other than I would have made the statement a little nicer and more tactfully, I would have made a comment about this even if I hadn't been run out. Additionally, a HUGE issue was made out of me using my bot to make a discussion edit just to tell them I couldn't respond. A rule that caused all of my bot activities to be revoked. I would submit that breaking the rule editing without approval is almost as bad. Certainly it was a major issue in the Cyber submission as well. 71.163.243.232 (talk) 02:17, 4 March 2012 (UTC)
Just happened to have run into this. I did answer that question with "The bot can be shut down pending approval. I merely ran it to see if it was functioning. I will shut the bot down now." If you want my attention on this, I suggest you contact my talk page. I'm not watching my watchlist.—cyberpower (Chat)(WP Edits: 520,335,904) 02:43, 4 March 2012 (UTC)
Kumioko, the reason why your bot approvals were revoked, most of all, is that you used your bot account for block evasion. That in itself shows you're not suited to run a bot. But in any case, I thought you said you retired. Doesn't really look like it. Snowolf How can I help? 02:56, 4 March 2012 (UTC)
(edit conflict) @ 71.163.243.232 - Perhaps if you "don't really have the time or inclination these days to embroil myself in another useless discussion" then you shouldn't start one. You started the discussion, why start a 'useless' one? Did you expect folks to just jump in and say "ohh yeah, you're so right."--v/r - TP 02:57, 4 March 2012 (UTC)
Just ignore him. Kumioko feels like the Wikipedia community treated him badly, so he's decided to troll as an IP until someone blocks him, so he can go on about how badly IPs are treated. Better to just let him continue to rant tediously in hopes he'll eventually tire of it. 28bytes (talk) 03:14, 4 March 2012 (UTC)
Your right 28bytes, I do think that I was treated badly and I think the situation was handed poorly. I think it was an overreaction to a small problem that was easily fixed had I been allowed to fix it (the problem never did get fixed by the way). Just because you don't agree, doesn't mean I am trolling.
Reply to Snowolf Your partially right, and I can accept the results and I am basically retired, hence the use of the IP. Think of me as a consultant, I can make a comment and you can choose to ignore it if you don't like it. I still contend the block was inappropriate in the first place and only done to kill the discussions and block me from commenting in the 8 different discussions. Maybe if the Wiki software allowed an admin to grant edit ability to say talk pages by exception, rather than restrict the editor only to their talk page, that might be a good thing too. All this is irrelevant to this discussion and frankly as a Steward I would have expected a better comment from you than to just push it back on me. Also, maybe your right. Maybe I'm not suited to run a bot but that really doesn't have any bearing on this discussion either other than to discredit the statements.
Reply to TP - Actually TP you responded exactly like I expected you too.
I know that none of you want to hear it, your tired of hearing from me and my comments and you don't think or don't care that you did anything wrong. That doesn't mean my comments are invalid nor that they are trolling. You just spent a paragraph each pointing fingers at me and didn't even comment on why one BRFA was rushed through and the other editor was essentially told we don't really need or want your help and both bots had the same issue (yes I know Cyber's bot had a couple of others as well but they were easily fixed and I think they were doing a good job of addressing those as they came up). I don't expect anything to come of this discussion frankly but maybe it will make you think about the situation when it occurs again the future. 71.163.243.232 (talk) 08:44, 4 March 2012 (UTC)
Your reply doesn't address my question. How does the discussion become useless because it goes a direction you don't like and why would you start a discussion knowing it would be useless? It's your approach that is causing the problem. You need to come at this a less confrontational direction and with serious self reflection. I dont even know you and I can pick that up right away. I'm not even part of BAG and you can't lump me in with whomever did anything wrong. But you're not going to get an open ear from me the way you come at folks.--v/r - TP 13:14, 4 March 2012 (UTC)
TP you have a couple good points. Since we never had contact till recently some history on my "attitude". I used to taste my words prior to typing them and frequently word-smithed my posts over before hitting enter. I tried to work with every editor that came crying with a complaint about my edits and catered to almost every whim to avoid confrontation. I did that for years. Recently, I was embroiled in multiple discussions and I lost my composure. That's all it took and now, frankly, I don't care so much if my words seem a bit abrupt and tactless. For me the time for diplomacy is over so now I just speak what I think and if that hurts a few editors feelings or ruffles a few feathers then that's fine. In the end I was painted as an editor who is always battling with other editors and trying to "force my opinion on others. When that's exactly what they were doing to me and I just never made an issue out of it. So now, since being the tactful nice guy didn't seem to matter, I just tell it like it is. If I seem like a jerk or have an attitude its just a conditioned response to constant attacks. 71.163.243.232 (talk) 00:10, 5 March 2012 (UTC)
Frankly, this new way isn't going to work out better than the old way for you. You need to find a third option.--v/r - TP 02:03, 5 March 2012 (UTC)
Your probably right, but it seems to work for quite a few editors so I figure I'll give it a try. The third option is to stop editing completely by the way and I am well on my way there. I edit less and less every day. Pretty soon the user known as Kumioko will just be a memory in the history of Wikipedia and everyone will be a lot happier it seems. At least the servers will be under a lot less stress without the hundreds of thousands of edits I was doing. 71.163.243.232 (talk) 22:07, 5 March 2012 (UTC)
Or, fourth option, just come out and say "Alright, I was mad, but I'm ready to play nice again and I'm sorry" and then probably just avoid BAG since it frustrates you and continue doing the other parts you've enjoyed in the past.--v/r - TP 13:13, 7 March 2012 (UTC)
I already apologized for my actions multiple times and admitted multiple times I let my anger get the better of me. What frustrates me isn't BAG, its people holding me responsible for others actions. Several people screwed up, not just me and I am the only one that got blocked or any other action. Headbomb did eventually leave a message on Brad101's talk page but only because Brad left a message on a BAG page. NO ONE cared what they did or said to me that provoked me and caused me to lose my patience. NO ONE cared that Mark broke the 3RR policy, NO ONE cared that they and others broke several other policies. They only cared that I made a statement that they thought meant I was going to keep reverting. Thats what frustrates me and caused me to lose faith in the system and in the judgement of the community. Including BAG. 71.163.243.232 (talk) 23:46, 7 March 2012 (UTC)
Ok, but my point is that still arguing over it isn't going anyway. Your account isnt blocked anymore (that I can see) and it's time you let the issue go, move on, and put it in the paste so you can return to productive editing.--v/r - TP 00:58, 8 March 2012 (UTC)
This is part of the problem, you keep saying things like no one cared about what others did, which isn't true. Mark was warned as were you. He stopped, you continued. Stop playing the victim. What you did was far worse than what the others did. And you DID keep reverting. It wasn't just a statement. You followed through on the statement. -DJSasso (talk) 01:13, 8 March 2012 (UTC)
Reply to TP - Your point is taken and understood. After this post I won't post here anymore. I am going to let it go because eventhough my account isn't blocked the account Kumioko is dead and its not coming back. That user loved Wikipedia, believed in it and followed its rules. Others did not and they are still here so I will not be and thats too bad because frankly I was a damn good and productive editor.
Reply to Djsasso - I'm not sure what discussion you are referring too because the ones I was involved in didn't say anything like what you are saying. I made the statement after I made a reverstion and that was after Mark already broke the 3RR rule twice, on that one occassion which he has done many times before I just never made an issue out of it. The Only thing I did wrong was to have faith in the system and think it would actually work, when I knew better from seeing it in action before and make an edit through a block using my bot and I still contend that was an invalid block only done to keep me from commenting...Guess what! It worked, after I came back, the discussions died out, I had lost my credibility and everyone moved on. But thats ok because editors like Brad101 and all the comments on his talk page and Mark and his little comments and attitude are what people expect from Wikipedia editors these days. Thats why they don't edit and we are losing editors faster than we can add them. Whats even worse is a lot of editors have seen it, I keep seeing my name all over the place so I know I made a difference. At least I can leave with that. 71.163.243.232 (talk) 01:48, 8 March 2012 (UTC)
Maybe then a WP:CLEANSTART is in your future? If so, then good luck with it.--v/r - TP 02:38, 8 March 2012 (UTC)

Keep on topic please guys. If you want to discuss Kumioko start a new thread somewhere. --Chris 09:32, 4 March 2012 (UTC)

Indeed. If I may try to rescue the thread, the point is more: "Regardless of the treatment of other BRFAs, was the approval of Addbot 24 a good one?" Does anyone actually have any more relevant thoughts on that to contribute? - Jarry1250 [Deliberation needed] 11:10, 4 March 2012 (UTC)
I guess my only concern is that it was closed too quick. BAG really shouldn't get into that habit. The task itself has been done before, it's straight-forward, the botop already filled the same one thrice and was given trial twice. That said, speedy approval should never occur after running without trial, even if there aren't any real problems. —  HELLKNOWZ  ▎TALK 11:44, 4 March 2012 (UTC)
I will happily stop the bot task if there is consensus to do so and will happily have another BRFA opened for this task if need be. Vote..? ·Add§hore· Talk To Me! 13:34, 4 March 2012 (UTC)
There's no need for that. The bot was performing well and up to par before the BRFA, and already had trials, so having yet-another BRFA would be nothing more than an exercise in bureaucracy. And considering that that BRFA was already an exercise in bureaucracy... Of course if the rest of the BAG thinks there's some need for such bureaucy, then I won't stand in the way, but I certainly will wonder about where BAG's collective common sense went. Headbomb {talk / contribs / physics / books} 15:06, 4 March 2012 (UTC)
I agree and think that the original question prompting this discussion has been answered. — madman 17:53, 4 March 2012 (UTC)
Fair enough. 71.163.243.232 (talk) 00:10, 5 March 2012 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Bot owners invited to Berlin hackathon

Hi. I'm helping organize the Berlin hackathon, 1-3 June 2012 in Berlin, Germany. We're going to be discussing and working on the hosted Wikimedia Labs development environment and bots infrastructure, Toolserver, the future of ResourceLoader and Gadgets, MediaWiki's web API, the new Lua templating system, and various upcoming MediaWiki features and changes. We'd love to have power users, bot maintainers and writers, and template makers at these events so we can all learn from each other and chat about what needs doing.

That's one of the upcoming Wikimedia developers' events and I hope some of you can make it.

- Sumana Harihareswara, Wikimedia Foundation's Volunteer Development Coordinator. Please reply on my talk page, here on English Wikipedia or at mediawiki.org. Sumanah (talk) 22:08, 16 February 2012 (UTC)

I will be able to provide some travel subsidies, in case that helps people decide to come! I'll post more information about that as soon as possible, hopefully this week. Sumana Harihareswara, Wikimedia Foundation Volunteer Development Coordinator 02:06, 12 March 2012 (UTC)

TorNodeBot is down and/or needs replacement

TorNodeBot (talk · contribs) worked pretty well, but it seems to have been shut down, leaving TOR holes into Wikipedia.Jasper Deng (talk) 03:37, 14 March 2012 (UTC)

It probably just runs on nightshade. If that is correct, I'll ask Shirik (talk · contribs) to run it on willow, or manually. →Στc. 04:26, 14 March 2012 (UTC)
I restarted my server a while back and totally forgot to put it back up. Sorry about that. I'll restart it. (P.S., next time try just leaving a note on my talk page.) --Shirik (Questions or Comments?) 14:56, 14 March 2012 (UTC)

Can We Get a Ruling

...over here? The bot's been doing the same tasks since day 1, no amendments to the code ... today, it's suddenly blocked. (talk→ BWilkins ←track) 01:27, 12 January 2012 (UTC)

The request at Wikipedia:Bots/Requests for approval/7SeriesBOT says that only deletions that comply with WP:CSD#G7 will be done. Because G7 specifically and explicitly excludes user talk pages, that means that the bot approval must also exclude such pages. It would probably be better to get consensus at WT:CSD that this bot has an exception to the normal G7 rules. — Carl (CBM · talk) 01:35, 12 January 2012 (UTC)
WP:CSD#G7 and WP:CSD#U1 both explicitly state that user talk pages are not to be deleted under those criteria. User pages deleted under WP:DELTALK should certainly not be done by a bot, as they are only done in exceptional circumstances, and bots are no good at detecting something as general and vague as "exceptional circumstances". Seems pretty black and white to me. Just because the bot has been doing it for awhile with no complaints doesn't necessarily mean that it's ok, it just means that no one has noticed until now. —SW— communicate 02:29, 12 January 2012 (UTC)
It seems there is a current discussion about this very topic here. —SW— express 02:32, 12 January 2012 (UTC)
I've re-read both BRfAs and the associated deletion criteria. As Scottywong mentions, there is an "except under very exceptional circumstances" get-out clause, and I'd suggest that only one editor (per Task 1) is an exceptional circumstance. I may be a lack of imagination or experience on my part, but I can't see when a Task 1 deletion would be bad. Josh Parris 03:39, 12 January 2012 (UTC)
...and look at the incidents where the usertalk page has been deleted ... in one case someone warned them, removed the warning, then tagged for G7 ... looks like an obvious case (talk→ BWilkins ←track) 17:04, 12 January 2012 (UTC)
Personally, when I find those, I remove the {{db-g7}} and replace with a welcome template. Reaper Eternal (talk) 17:11, 12 January 2012 (UTC)

Wouldn't the obvious compromise be to only delete user talk pages if the single contributor was the user? I can't imagine there would be any concerns about deletion then. And really, it's not like there's a huge G7 backlog to begin with. CSD G7 seems fairly clear, and consensus on any exception would have to be made at Wikipedia talk:Criteria for speedy deletion, not here. — madman 17:12, 12 January 2012 (UTC)

That could therefore possibly be a U1 ... (talk→ BWilkins ←track) 17:27, 12 January 2012 (UTC)
If someone mistakenly creates a user talk page (e.g. wrong click on Huggle or whatever) and then decides to G7 it, will the user whose talk page briefly existed get a "You have new messages" bar with a red link to their talk page when they log in? That could be a bit confusing for them. 28bytes (talk) 17:24, 12 January 2012 (UTC)
I don't think they will after the page has been deleted. — madman 17:26, 12 January 2012 (UTC)
If that's the case, then I don't see a problem with what the bot's doing. The alternative would be for whoever's requested the G7 to have to wait until it's manually deleted by an admin, and risk giving the innocent user a spurious orange bar pointing them to a scary-looking {{db}} box on their talk page. I think the key concern should be to avoid confusing or annoying a new user. 28bytes (talk) 17:32, 12 January 2012 (UTC)
I understand your concern. But I think that's a fringe case in practice (if users are warning the wrong people that often, they really need to slow down and look at what they're doing; it's not a race). And I agree with the blocking administrator that as it stands, the bot is incorrectly deleting pages per U1 and G7. U1 is absolutely unambiguous, but I do agree that there's a case to be made that G7 should apply to user talk pages with a single contributor. That case should be made at Wikipedia talk:Criteria for speedy deletion. — madman 17:38, 12 January 2012 (UTC)
@Josh Parris: how easy is it to remove the offending code. We want to make sure that it's still deleting compliant User talk:WhoeverWhatever/mysubpage but not User talk:WhoeverWhatever ... is this something I can tweak? (I'm pretty sure the code has not changed since the last one you uploaded). I can run it that way until we clarify things elsewhere (talk→ BWilkins ←track) 16:40, 13 January 2012 (UTC)
Give me a few days to find my feet again. Josh Parris 22:01, 13 January 2012 (UTC)
No problem! Thanks Josh ... good to have you back and around (talk→ BWilkins ←track) 14:02, 14 January 2012 (UTC)
I've finally regained toolserver access. Josh Parris 08:01, 16 March 2012 (UTC)
Are you logged out?—cyberpower ChatPartially online possibly on a phone 10:21, 16 March 2012 (UTC)
Yes. How did that happen? Josh Parris
  • Just a friendly bump ... just in case possibly someone other than Josh can take a peek at the code and suggest changes. It's sad to have ol' 7Series down for so long after running without complaint for so long (talk→ BWilkins ←track) 16:11, 7 February 2012 (UTC)
As a temporary fix, I have removed cmnamespace 3 from the code, and have requested unblock of the bot. This will cut out a substantial part of what it's checking, but I'd rather having it limping than nothing (talk→ BWilkins ←track) 18:38, 11 February 2012 (UTC)

X!'s bots

Anyone with a toolserver account willing to take over some of SoxBot's tasks? I have not contacted X!, but even if he renews his toolserver account it's probably good sense to move the tasks to active users. Code is I all open source at https://code.google.com/p/soxred93tools/source/checkout.

Not sure which tasks are still relevant. Some aren't: I've redirected the admin-highlighter tool to a clone. But others should probably be replaced:

  1. Creating the current events pages, like Portal:Current events/2012 February 9
  2. Maintaining RfX tally and RfX report
  3. Keeping {{badimage}} on all file description pages listed at MediaWiki:Bad image list (BRFA)
  4. Maintain {{Cratstats}} and {{adminstats}} (BRFA, BRFA)

Maybe there were other active tasks that I didn't notice. Neither seems particularly pressing though.

Amalthea 14:32, 9 February 2012 (UTC)

533Mb download, in case anyone wasn't expecting that. --Northernhenge (talk) 22:24, 13 February 2012 (UTC)
And I'm adapting all of it. I am working as hard as I can to get it done quickly.—cyberpower (Talk to Me)(Contributions) 22:46, 13 February 2012 (UTC)
I'll volunteer for at least the first task, and probably the other 3 too if no one else has a strong desire to claim those tasks for their bot. 28bytes (talk)
For what its worth I also notice that the edit counter isn't working anymore either because his Toolserver account expired. --Kumioko (talk) 16:47, 9 February 2012 (UTC)
I'm working on taking over 2, 4, and the edit counter.--v/r - TP 19:03, 9 February 2012 (UTC)
I could help out with the edit counter and SoxBot. I have a toolserver account. (It seems I am not the only one trying to take over.)—cyberpower (Talk to Me)(Contributions) 22:43, 9 February 2012 (UTC)
I would love it if someone would take over the Article Blamer tool from the toolserver account as well. GoingBatty (talk) 00:41, 10 February 2012 (UTC)
Getting taken care of.00:58, 10 February 2012 (UTC)
No you don't... →Στc. 01:05, 10 February 2012 (UTC)
Not by me. User:TParis is taking care of it and has already gotten the edit counter up and running. You can find it here. BTW, I thought I had one. Turns out I only have the Wiki account. Facepalm Facepalm—cyberpower (Talk to Me)(Contributions) 01:48, 10 February 2012 (UTC)

Two more tasks that seem to have been active in 2012:

  2. Remove {{REMOVE THIS TEMPLATE WHEN CLOSING THIS AfD}} from closed AfDs. (BRFA)
  3. Remove {{uncategorized}} from pages that have (non-hidden) categories. (BRFA)

I may look at these, and #3 above, unless someone else wants them. Feel free to steal them from me. Anomie⚔ 21:03, 9 February 2012 (UTC)

Yobot occasionally does #6. -- Magioladitis (talk) 21:33, 9 February 2012 (UTC)

I've put in a BRFA for #1. Looks like other folks have the rest covered. 28bytes (talk) 22:12, 9 February 2012 (UTC)

I am rewritting the code behind the bots. This should take care of most if not all of the bots. I will be placing them under Cyberbot [place roman numeral here]—cyberpower (Talk to Me)(Contributions) 03:43, 10 February 2012 (UTC)
Note that Snotbot already does #5 above, although it only reports pages with this problem to a report page, it doesn't fix the problem automatically (because I believe it was decided that a bot shouldn't do this? I can't remember). As the bot already detects the problem, it could just as easily fix it instead of report it. —SW— chat 08:20, 10 February 2012 (UTC)
Thanks for that. I hope to have them operational and approved within 7 days. Being that these were already approved bots, I don't anticipate needing to take these through a trial.—cyberpower (Talk to Me)(Contributions) 10:46, 10 February 2012 (UTC)
If you rewrite them then I think you do.
Personally, I think it makes much more sense if TParis and Anomie take over those tasks like they offered, they have toolserver accounts and can use the tried code directly without problems.
Also, I saw that you moved some of SoxBots subpages related to WP:CHU clerking: That task doesn't need to be replaced at all, it hasn't been active for quite some time and I believe Chris G is clerking there now.
Amalthea 13:06, 10 February 2012 (UTC)
Amalthea: No toolserver account for me. But only #4 really needs it, IMO, since {{uncategorized}} isn't very widely used. Anomie⚔ 17:16, 10 February 2012 (UTC)
Cyberpower678: It looks like everything is "claimed" already. And Amalthea is correct: rewriting it would probably want a trial to verify that they rewritten code was correct, and Chris G Bot replaced SoxBot for clerking WP:CHU a rather long time ago. Most of SoxBot's historical tasks haven't been done by SoxBot for years now, actually. Anomie⚔ 17:16, 10 February 2012 (UTC)
Scottywong: Please do take #5 then. SoxBot did it once or twice in 2012 already, so if that was decided it must have been fairly recent. Anomie⚔ 17:16, 10 February 2012 (UTC)
Will do. —SW— squeal 00:52, 14 February 2012 (UTC)
Magioladitis: Will Yobot do it regularly enough, or should I have AnomieBOT do it? Anomie⚔ 17:16, 10 February 2012 (UTC)
I've already got this project started. I've already created User:Cyberbot I to replace SoxBot. The rewrite is basically adapting SoxBot's code and replacing anything with Sox to references to my to be bot account. I've already got the majority of the code translated. My goal is to have Cyberbot I operate exactly like SoxBot did at it's last current state. This would require me to transfer the bots subpages over to mine where needed. (All of them).—cyberpower (Talk to Me)(Contributions) 18:52, 10 February 2012 (UTC)

What about the tool that can report on contributions? It is broken now but very useful. Anyone took this one over?--Codrin.B (talk) 00:00, 16 February 2012 (UTC)

TParis should have it. I changed your link. Try it now. I am working on the bots but at the rate I am going, I believe I may get done quicker than he is able to get it done.—cyberpower (Chat)(WP Edits: 516,979,856) 00:04, 16 February 2012 (UTC)

Cyberbot I is almost ready. There appear to be a few bugs that are causing the bot to hang up.—cyberpower (Chat)(WP Edits: 519,827,507) 11:11, 1 March 2012 (UTC)

Please all see the request for approval here ·Add§hore· Talk To Me! 13:59, 3 March 2012 (UTC)
I have withdrawn the request (obviously). Cyberbot I is still here and ready to run if ever needed. See User:Cyberbot I for a list of tasks it's able to do.—cyberpower ChatLimited Access 11:43, 20 March 2012 (UTC)
It appears that task #3 hasn't been taken over. I can have User:Cyberbot I take it over but I need to know if this is true or not.—cyberpower ChatLimited Access 12:38, 21 March 2012 (UTC)

Wikimania 2012

Well this is mostly about two things. First, I made a submission to present something concerning bots at Wikimania 2012. Assuming it gets selected, I invited you all to come to it (you can sign your name at the bottom of the submission if you plan to, although this isn't required to attend).

Second, who here will attend WM 2012? It might be a good idea to meet up and socialize. Headbomb {talk / contribs / physics / books} 18:32, 18 March 2012 (UTC)

If I get the fellowship I am coming and I am certainly following this talk! -- Magioladitis (talk) 18:34, 18 March 2012 (UTC)
I'll be attending and I certainly shall check out the presentation. Cheers! — madman 14:04, 19 March 2012 (UTC)
I'll probably be coming, depending on my college plans. If I do, I'll most definitely attend the talk.~ Matthewrbowker Talk to me 22:42, 19 March 2012 (UTC)
And we can all meet up & work together at the Hackathon as well! Sumana Harihareswara, Wikimedia Foundation Volunteer Development Coordinator 01:52, 20 March 2012 (UTC)
I might be around for the Hackathon, but will try and attend the presentation. -- DQ (ʞlɐʇ) 19:42, 21 March 2012 (UTC)

Hi, I want to inform somebody about this bad bot edit (H:ILLs on a /doc page should be inside the <includeonly>...</includeonly>, not in a <noinclude>...</noinclude>, and bot shouldn't have removed those ILLs already present) - but the bot's talk page redirects to a redlink. My message would be something along the lines of those that I posted to a different bot a few weeks ago, see here. What's the procedure? --Redrose64 (talk) 17:52, 20 March 2012 (UTC)

I've blocked the bot and asked its owner to respond here. MBisanz talk 17:58, 20 March 2012 (UTC)
As far as I can see, this only happened on one page, which was a "/doc" subpage. I'm not sure how pywiki handles these, but it should have? I can't say if that was interwiki.py problem or not. I did warn the operator on the BRFA though. —  HELLKNOWZ  ▎TALK 18:05, 20 March 2012 (UTC)
I'm sorry my robot to make the wrong editor, but I do not know the cause of the fault occurs, I use the command: interwiki.py -warnfile:logs\warning-wikipedia-en.log -lang:en -cleanup -async -autonomous, pywikipedia version is 10001(2012-03-10). Who can help me solve the problem?--YFdyh000 (talk) 18:17, 20 March 2012 (UTC)
I found a method that reproduces the problem, but still do not know its cause.
problem reproduce steps:
In the logs folder to create a test.log, are:
WARNING: wikipedia: [[en:Template:Infobox theatre/doc]] does not link to [[zh:Template:Infobox Theatre]]
Use the command: interwiki.py -warnfile:logs\test.log -lang:en -confirm
I get the wrong results. Can anyone tell me why this is.--YFdyh000 (talk) 18:41, 20 March 2012 (UTC)
It would appear to be because on zh:Template:Infobox Theatre, there's an interlanguage link RIGHT before contributors are told not to put interlanguage links on that page. [[en:Template:Infobox_theatre/doc]]<!-- Add cats and interwikis to the /doc subpage, not here! --> — madman 01:47, 21 March 2012 (UTC)
I do not quite understand what you say. My script then run in automatic mode, is there any way to avoid this problem in order to continue my script?--YFdyh000 (talk) 09:36, 21 March 2012 (UTC)
It's an error on zh:Template:Infobox Theatre and needs to be fixed there. I assume you could better do that than me, though I can try if you can't figure it out. — madman 13:49, 21 March 2012 (UTC)
Oh, I have learned, and has fixed this fault. Now how can I do to lift the banned and continues to run my script? --YFdyh000 (talk) 14:29, 21 March 2012 (UTC)
I note that MBisanz unblocked you. But this is something of a last chance with regard to editing template (talk) space, so please do be very careful. - Jarry1250 [Deliberation needed] 14:43, 22 March 2012 (UTC)

Comments on Wikipedia:Bots/Status

Well after Wikipedia:Bots/Status has been updated the inactive list (which is a subpage) is not to big to be included on the main status page. I just need some opinions of what should be done with this page..? ·Add§hore· Talk To Me! 15:56, 16 March 2012 (UTC)

1026 Inactive Bots, that's a lot. What's the upside in putting them on the main status page? Josh Parris 04:38, 17 March 2012 (UTC)
It's already transcluded on the main status page, the mediawiki software is just stopping it from showing because it's too big. - Happysailor (Talk) 18:19, 17 March 2012 (UTC)
Is there any need to show the inactive bots? Or just have a link to them from the status page? Is this a better idea? ·Add§hore· Talk To Me! 17:25, 18 March 2012 (UTC)
Well, MediaWiki ain't going to change transclusion limits any time soon. So if we want to show all bots on the same page, the entries could be in plain text. The bot can do that; make the list from the main page, which uses templates. I don't see how else we can show all the info on the same page. —  HELLKNOWZ  ▎TALK 12:18, 24 March 2012 (UTC)

Insane RepLag

Anybody know what's going on with the replication database. Toolserver has a replag of almost 48 hours now.—cyberpower ChatOnline 19:50, 21 March 2012 (UTC)

Didn't you already ask this and get an answer at User talk:Jimbo Wales? Dpmuk (talk) 19:58, 21 March 2012 (UTC)
He said he was foing to forward it to the tech people but I want to know what's causing it.—cyberpower ChatOnline 20:05, 21 March 2012 (UTC)
This is a Toolserver issue, nothing to do with the tech staff of the WMF. As to what's causing it, feel free to check on jira, likely recorded there. Snowolf How can I help? 20:07, 21 March 2012 (UTC)
Specifically, feel free to look at http://jira.toolserver.org/browse/MNT-1225, which seems to have underestimated the impact. --R'n'B (call me Russ) 21:07, 21 March 2012 (UTC)
See also WP:VPT#Toolserver replication lag. --Redrose64 (talk) 21:10, 21 March 2012 (UTC)
Thanks guys. That's what I was looking for.—cyberpower ChatOnline 21:14, 21 March 2012 (UTC)
I heard they were adding a rev_sha1 column to the revision table and the database wouldn't be up to date until it was done adding that new column [6]. I've turned off all my database-dependent bots until it fixes itself. Tim1357 talk 21:18, 21 March 2012 (UTC)
def replag_ok(cursor):
    " Returns true if replication lag is under 8 hours "
    cursor.execute("SELECT UNIX_TIMESTAMP() - UNIX_TIMESTAMP(MAX(rc_timestamp)) FROM recentchanges;")
    (replag,) = cursor.fetchall()[0]
    return replag < 8*60*60
Automating it mean you don't need to constantly look after your bots. — Dispenser 23:00, 29 March 2012 (UTC)

Random Tasks getting killed by Cron Daemon on Willow

Is having the same problems I'm having. Regularly running scripts (which having been running without issues) are randomly getting killed by the cron daemon. Lucky for me they start up again. Has anybody else been getting this though?—cyberpower ChatOnline 22:23, 29 March 2012 (UTC)

I believe User:The Earwig has/had the same (random) problem. mabdul 10:56, 12 April 2012 (UTC)
That could be for any number of reasons. We need the specific errors you are getting to help. And a better venue may be the toolserver-l mailing list. My gut instinct is that you had a query to sql that took way too long. -- DQ on the road (ʞlɐʇ) 13:54, 12 April 2012 (UTC)
Or you may have exceeded the memory limit; you should have gotten an e-mail from slayerd, though. (As you'd also get an e-mail from the query killer.) — madman 14:03, 12 April 2012 (UTC)
No the cron daemon only sent an email with the task as its subject and "killed" as its body.—cyberpower ChatLimited Access 14:11, 12 April 2012 (UTC)

RM bot inactive

This bot, which maintains the list of requested moves, suddenly stopped working aroung 13:00, 12 April (UTC). I have emailed the owner, HardBoiledEggs, but he hasn't edited in nearly two months, so I'm asking this group if anybody knows how to kick start the bot. Favonian (talk) 17:03, 13 April 2012 (UTC)

Oh dear, looks like the bot has been blocked. Now what do we do? Favonian (talk) 17:16, 13 April 2012 (UTC)
Go to WP:BOTREQ and request a replacement? --MZMcBride (talk) 17:20, 13 April 2012 (UTC)
(ec) I've left a message for the blocking admin. Perhaps the erroneous notice it was leaving could be ignored for now as the RM backlog is surely growing. ​—DoRD (talk)​ 17:25, 13 April 2012 (UTC)
I changed the non-standard section title == Move? == to the standard == Requested move == at Talk:Loving You (disambiguation), in case that was confusing the bot—seems it did not understand which section had the move discussion. –Wbm1058 (talk) 18:09, 13 April 2012 (UTC)
I have unblocked the bot and will monitor it for any further issues on that page. (update) Well, the bot replaced the notice, but perhaps it should just be left there to prevent unintentional edit warring. ​—DoRD (talk)​ 18:21, 13 April 2012 (UTC)
Certainly seems the lesser of two evils. The bot has resumed its duties. Thanks for your effort! Favonian (talk) 18:36, 13 April 2012 (UTC)
I noticed that the bot changed the title "== Move? ==" its first pass at Wikipedia:Requested moves/Current discussions after it was unblocked. Or maybe it was these moves [7] [8] that effected it somehow? It's made a second pass without further incident. Anyways, thanks for getting it running again! -Wbm1058 (talk) 19:04, 13 April 2012 (UTC)
As the blocking admin, I have no problem with the unblock, although as a developer I can't say that it's definitely the lesser of two evils. It's an observed bug, and it may be a symptom of bigger problems, and the block was placed per the bot page's instructions. I am wildly unfamiliar with the bot segment of Wikipedia, but depending on the effort involved, it might be worth separating the two evils, so that one bot places Talk page notifications and another bot processes the RM Wikipage updates (and so could be blocked independently if needed). Cheers! -- JHunterJ (talk) 19:05, 13 April 2012 (UTC)
Thanks for your understanding and input. I notified the bot's developer to see if he has any insight into this (hopefully isolated) issue. ​—DoRD (talk)​ 19:25, 13 April 2012 (UTC)
The bot's source is available. It looks to me like the bot is assuming that—as documented at Template:Move-multi—the discussion is being hosted on the talk page of the page listed as "current1", and the talk pages for other "currentN" parameters are to be notified. But in this case the discussion is being hosted on "current2"'s talk page. So Talk:Loving You ("current1") is not being notified when it should be and Talk:Loving You (disambiguation) is being notified when it shouldn't be. Anomie⚔ 20:40, 13 April 2012 (UTC)
Looks like you're right: this fix is holding. -- JHunterJ (talk) 14:04, 14 April 2012 (UTC)

WP:HIDEBOT

Bot owners whose bots are not prone to errors, and whose tasks are confined to one specific area may wish to link to this section in their edit summaries / bot profile. I also remind you that the possibility of ignoring your bot is not an excuse to dismiss user concerns of WP:COSMETICBOT, nor should it create a 'Don't like it? Ignore it!' mentality. Headbomb {talk / contribs / physics / books} 14:24, 14 April 2012 (UTC)

Notice: Slight change in AWB in-template handling

This shouldn't affect most of you, but AWB's search/replace logic for "in-template" check has been revised to address this issue in r8062. Headbomb {talk / contribs / physics / books} 19:44, 15 April 2012 (UTC)

IRC bot operators - heads up

https://bugzilla.wikimedia.org/show_bug.cgi?id=30245 has been fixed. This may break bots that consume the mediawiki IRC feeds. The fix deploys in ten days. Josh Parris 11:42, 13 April 2012 (UTC)

This post is baffling. bugzilla:30245 was marked fixed several months ago and there hasn't been any activity on the bug since February 2012. What are you talking about regarding a fix being deployed in ten days? --MZMcBride (talk) 17:25, 13 April 2012 (UTC)
MW 1.20 is due to be deployed soon, so I suppose that's what Josh is referring to. Ucucha (talk) 17:38, 13 April 2012 (UTC)
The fix was reverted in 1.19, as the change in format broke many bots and scripts. Now 1.20 is coming, and those bots and scripts will break again. But this time notice is being given, so this time it won't be reverted. Anomie⚔ 17:56, 13 April 2012 (UTC)
Ah, here we go: https://blog.wikimedia.org/2012/04/12/mediawiki-1-20wmf1-deployment/. Okay, thank you both for clarifying. --MZMcBride (talk) 20:18, 13 April 2012 (UTC)
Related links, mw:MediaWiki 1.20/wmf1/overview#API and bot-facing changes & the deployment timetable at mw:MediaWiki 1.20/Roadmap. Snowolf How can I help? 03:53, 14 April 2012 (UTC)
The change wouldn't have affected enwiki, but in any case [9] looks like it's getting reverted. Snowolf How can I help? 14:21, 20 April 2012 (UTC)

Citationbot

The bot is down due to toolserver account expiration. The maintainer has not edited since 9 April. Can anyone restore the tool? See also User_talk:Citation_bot#403:_User_account_expired --Mblumber (talk) 04:32, 22 April 2012 (UTC)

Playing

So I'm poking around with pywiki, learning the ropes a little. I'd like to start testing out edits but I'd really rather not run the edits on a live wiki. So my next step is to set up a local wiki mirror — but I'm a little clueless at that end — could anyone point me in the direction of some resources for setting oneself up a local mirror in such a way that pywiki can get it's teeth in? Fayedizard (talk) 19:37, 2 April 2012 (UTC)

You could test the bot in your userspace. That won't cause any issues and is much easier than setting up a mirror.—cyberpower ChatOnline 20:34, 2 April 2012 (UTC)
mw:Installation will tell you how to install and configure MediaWiki. If you really want to have a mirror, you could then download and import a database dump; alternatively, you could use Special:Export to export a few articles for testing, and then use Special:Import on your test wiki to load it in. Anomie⚔ 20:46, 2 April 2012 (UTC)
I'm sure you know this but, you would need your own accessible server with access to the internet if you want to test it from toolserver.—cyberpower ChatOnline 22:32, 2 April 2012 (UTC)
There's a test Wikipedia you can use. --MZMcBride (talk) 17:26, 13 April 2012 (UTC)
Although this wiki might be the wiki you are looking for? Rich Farmbrough, 14:10, 6 May 2012 (UTC).(Using some automation)

Helpful Pixi Bot and Citation Bot

I've noticed lately that a number of pages I created, which were on my watchlist, were dropped from my watchlist. One or two articles could be something accidental I did myself. But there seems to have been a lot of them. And, no, I don't have so many on my watchlist that I exceeded any kind of limit. And the only common factor I can find in all of them, is that they were visited by Helpful Pixi Bot or Citation Bot. If there's something odd going on with bots, that should be looked at and fixed. Maile66 (talk) 00:04, 4 May 2012 (UTC)

If the pages are truly not on your watchlist anymore (they do not appear on View and edit watchlist or Edit raw watchlist), the only possibility I can think of is that you may have accidentally edited your raw watchlist and deleted a chunk of it. Bots quite simply cannot affect your watchlist with their edits; no contributors can. It is possible, however, that you have bot edits hidden from your watchlist; see if that preference is enabled here. — madman 00:41, 4 May 2012 (UTC)
If the top level edits were by bots and you have "hide bot edits" activated on your watchlist, you will not see them. Killiondude (talk) 04:37, 6 May 2012 (UTC)

Wikipedia talk:Blocking policy#Unblocking bot accounts

Bot operators, particularly those who are also admins, may be interested in the discussion at Wikipedia talk:Blocking policy#Unblocking bot accounts. Anomie⚔ 01:32, 10 May 2012 (UTC)

Request for help for an AWB bot

I would like to create a bot that executes an AWB 'setting'. I will manually control the setting by changing the 'make a list' to look at different items--not more than AWB can handle at once. The AWB edits will not normally change. The bot would just have to 'push save' to accept the edits and not run faster than allowed. I have no idea how to go about doing this and do not know any of the bot languages. Any help would be much appreciated. Hmains (talk) 15:21, 11 May 2012 (UTC)

I think the first question to ask here would be what types of edits are you trying to perform? Could you give an example? Kumioko (talk) 15:26, 11 May 2012 (UTC)
Of course. I am unlinking years such as 2010 in all non-date related articles. Per WP:YEARLINK and MOS:UNLINKYEARS, I do not touch any year/month/day, etc article by using list filters and edit skips. For recent years, there are about 5,000 articles with links per year; for 100-year ago years, there are about 250 such articles per year. Hmains (talk) 16:02, 11 May 2012 (UTC)
Personally I think that seems like a good idea and I have code that would do a lot of that for you that I would be happy to pass your way. There are quite a few caveats that you'll need to watch for such as XXXX year in (insert your favorite topic here). I don't think this will get approved though as a bot task. The task has come up before and was too contentious so I don't think the BAG folks would approve such a bot. Kumioko (talk) 16:37, 11 May 2012 (UTC)
If you are happy using AWB then this is an easy task to write - and others have the code as Kumioko says. If you want to fully automate it, then you are unlikely to get approval, but you don't know until you try. If an account has a bot flag and is on the AWB list (I think both are required) then a "bot" tab comes up. AWB can handle massive lists, so no worries there. Rich Farmbrough, 18:31, 11 May 2012 (UTC).
Normally I wouldn't give code to just anyone but given this users long history with the project and the number of edits I gave him some code on his talk page that will help. Kumioko (talk) 18:53, 11 May 2012 (UTC)
I thank Kumioko for his code which will replace my long, complicated AWB edit/replace table. But it still requires a 'push save' for every article. I can only try to get a bot approval; as things are, this routine work is taking up all kinds of time. What else can you tell me, Rich or anyone, about such bot AWB logic? Hmains (talk) 22:17, 11 May 2012 (UTC)

Helpful Pixie Bot

It seems like Helpful Pixie Bot is editing at a rate that gives many edits in the same second (e.g. 50 edits from 23:58:53 to 23:59:03 - that's 50 edits in 10 seconds). Could someone from BAG look into that, and give an opinion about whether it falls within bot policy? — Carl (CBM · talk) 00:01, 6 May 2012 (UTC)

(Not a BAG member) The edits themselves seem to be fine and not against policy. However, it may not be following the bot requirements by editing just too rapidly and gobbling server resources. WP:BOTPOL states that "While performance is not generally an issue, bot operators should recognize that a bot making many requests or editing at a high speed has a much greater effect than the average contributor. Operators should be careful not to make unnecessary Web requests, and be conservative in their editing speed. Sysadmins will inform the community if performance issues of any significance do arise, and in such situations, their directives must be followed." On the other hand, WP:BOTPOL also states that "Bots doing non-urgent tasks may edit approximately once every ten seconds, while bots doing more urgent tasks may edit approximately once every five seconds." On the other other hand, this issue seems to have stopped—the bot is editing around once every 5 seconds now. Reaper Eternal (talk) 01:49, 6 May 2012 (UTC)
The bot policy doesn't really cover editing speed anymore; even the non-urgent task vs. urgent task guidelines that are there are kind of deprecated (and we might want to consider removing them). Use of the maxlag parameter allows bots to make requests as fast as they can in off-peak times and when the servers otherwise have a light load (and that seems like an off-peak time to me). Most major frameworks such as AWB and pywikipediabot use the maxlag parameter. So unless there's any evidence to the contrary, I'd assume even at that speed that bots are treating the servers just fine. Cheers, — madman 05:01, 6 May 2012 (UTC)
The devs mind. Back in February one of them was looking for an English Wikipedia admin who could deal with this. I'm not sure what rate is best, though. Part of the reason why editing speed is important, imo, is because sometimes bots make bad edits and if they're going super fast for a period of time it can take some time to fix/check which edits were bad. Depends on circumstances, of course. Killiondude (talk) 06:04, 6 May 2012 (UTC)
I agree with all of the above. Performance considerations seems to imply a rate of 1 edit per second, which is far higher than it used to imply. That said, even one a second means that a malfunctioning bot will spew dozens of edits before it can be blocked. This is the consideration we need to remember. I would suggest even the simplest bot should not exceed 1 edit every three seconds for this reason. - Jarry1250 [Deliberation needed] 14:25, 6 May 2012 (UTC)
Killiondude has hit upon the exact reason why edit rate needs to be limited. Other than that reason, worrying about edit rate should be as unnecessary as worrying about performance (if rates are anywhere within the realm of reasonable). However, as a corollary, it's extremely important that bot operators understand the wikitech robot policy that applies to this wiki in addition to the bot policy, particularly User-Agent requirements. Bots should always identify their operatos; this is helpful both for statistical reasons and so the sysadmins can block a bot that is malfunctioning, since it should have a unique UA string. Perhaps this should have greater prominence (or any prominence) in BOTPOL. — madman 16:47, 6 May 2012 (UTC)
How would bots do that? I've never seen a bot that didn't, either by directly redirecting user talk pages, leaving a note on the user pages, or both, indicate who the operator was. If you mean that we should mandate that bots carry their operator's names in them, I suppose that's exceptable. user:Svenbot would, under this policy, become user:Sven Manguard (bot), or something like that. It starts to become problematic when a bot operator has a long name, or runs multiple bots, but we could figure something out. Sven Manguard Wha? 05:28, 7 May 2012 (UTC)
I think he was referencing the robot policy's "user agent" section. Setting a custom user agent is trivial. →Στc. 05:52, 7 May 2012 (UTC)
Indeed; Sven, that means the User-Agent header used in HTTP requests. — madman 14:10, 7 May 2012 (UTC)
I'm not sure how a Greasemonkey bot would do that, but doubtless there's a way. LWP users are required to set an explicit useragent, as WP refuses the LWP agent (contrary to what Wikitech says). AWB I imagine sets a custom user agent, but not for each user. Again I'm guessing but I expect any bot (or software) that uses the API, regardless of the bot assertion in the calls, will be fairly clearly logged as an API call rather than an index.php call. The wikitech page is not really focussed on bots in the community sense of the word, but crawlers, spiders, mirrors and repudiators. Pixie, pywikiepedia and AWB all respect maxlag. Rich Farmbrough, 12:04, 8 May 2012 (UTC).
In my custom-built bot, I set the useragent string to "ReaperBot/1.0.0 (incompatible-notwebbrowser:robot:exclusion-noncompliant) <C#-wikipedia-bot>". Is that an acceptable useragent string? Reaper Eternal (talk) 13:11, 8 May 2012 (UTC)
I should think so; it identifies at least the bot's username on the project which you are querying. The additional information regarding bot functionality isn't really necessary for sysadmins or contributors (the latter can't see it), but it could be useful for your own statistical purposes. — madman 14:52, 8 May 2012 (UTC)
You are correct that API calls are clearly logged as such (and interesting statistical information is also collected, such as how much each format and module is used). However, I'd disagree with your characterization of the wikitech robot policy; it seems clear to me that it applies to bots interacting with the MediaWiki API in addition to crawlers, mirrors, and the like (obviously there are some more specific injunctions to the latter). — madman 14:58, 8 May 2012 (UTC)
It's pretty much something Tim wrote one day, a long time ago. Editing bots are tied to their user accounts, as Sven says, they can be better controlled through the MediaWiki interface, than digging through Apache logs looking for user-agent strings and extracting email addresses. From 25 May the Terms of Use will be the governing document. Rich Farmbrough, 18:40, 11 May 2012 (UTC).
How can sysadmins determine the identity of and/or stop bots that are performing expensive and/or rapid API queries (not actions or edits)? See the API usage section, remote loaders section. The point is that providing contact information in a User-Agent header is simply a courteous thing to do (and that's a guideline that applies to all HTTP bots, not only those using MediaWiki or its API). — madman 18:47, 11 May 2012 (UTC)
I'd guess that the sysadmins sometimes look at aggregate numbers of queries by user agent (from the apache or squid/varnish logs) in order to find clients that are being problematic. And sometimes they might look at who is making access to specific URIs that are problematic; this would be easiest to do from the apache and/or squid/varnish logs, otherwise they'd have to add extra debug-logging into MediaWiki itself to capture this information. In either case, they don't have the on-wiki account or the login cookies or anything like that. Without identifying information in the user agent, they'd pretty much have to checkuser the IP addresses involved to find out which bot account (if any!) was making those accesses. Anomie⚔ 20:41, 11 May 2012 (UTC)
Excuse me if I am way off here but this sounds like AWB would no longer be allowed as a bot app. The above discussion seems to indicate that one would need to custom build their bot, using the API, in order for this information to be collected. If so this would prevent a lot of the currently approved bots from opperating and ensure that only bona fide programmer types could run bots. That to me is wrong on a lot of levels. Kumioko (talk) 20:48, 11 May 2012 (UTC)
If it turns out to be a problem, I'm sure the AWB devs will do whatever is necessary to fix it. Anomie⚔ 21:06, 11 May 2012 (UTC)
Indeed, it'd be pretty easy for AWB to automatically append the logged-in user's name to the User-Agent (though if I remember correctly, AWB uses the IE ActiveX control, which doesn't allow that level of control). — madman 03:10, 12 May 2012 (UTC)
Precisely what I was getting at madman. Kumioko (talk) 03:17, 12 May 2012 (UTC)

You were insinuating that if the sysadmins insisted, no one with a stock version of AWB would be allowed to edit. My point is the same as Anomie's: That is quite simply not true. — madman 03:22, 12 May 2012 (UTC)

Sorry let me clarify. I don't think that no one would be able to edit with AWB. What I do think is that its going to make it far more difficult to use AWB for bots. Will it still be possible, probably, but much more difficult. I am not that great of a programmer frankly so it may not be hard at all. But, if AWB is designed how I think it is I think it would require some significant engineering to satisfy the new requirement. I am also not that familiar with this new requirement so its possible that the sysadmin folks would be happy with a generic text for the User-agent to allow it to work, I don't know. Kumioko (talk) 03:28, 12 May 2012 (UTC)
I think you're missing my and Anomie's point: That if the sysadmins insisted that AWB instances' User-Agent headers be unique, the AWB developers could make this change and it would be completely transparent to the end user. It would require no effort on the part of an editor or bot developer to comply. Also, this is not a new requirement, but a guideline that's been in place for quite some time. — madman 04:11, 12 May 2012 (UTC)
Ok I'll drop it but I still think that there is more to this discussion than just a simple header edit and would require effort on someones part. That's all I am saying. It just seemed like this discussion was a sneaky way of cutting the legs out from under AWB because of Rich's mischief. I know I am a Pariah here these days but I also know that there are a growing number of people who do not like AWB and would prefer that bots be left to programmers. Not saying you are one of them or that this discussion is about that but it would sure be a shame if one day someone started enforcing a policy like the one above and all the AWB bots (and potentially AWB in general) suddenly stopped working because they didn't understand or didn't care how AWB was designed. Kumioko (talk) 04:22, 12 May 2012 (UTC)

300EPM is way to damn fast for any bot doing any non-urgent tasks. It's doubly-unacceptable when the bot breaks articles ([10]), and that the bot-op is aware of it. Headbomb {talk / contribs / physics / books} 16:47, 12 May 2012 (UTC)

Where did Helpful Pixie Bot get to 300 EPM? --Dirk Beetstra T C 08:06, 20 May 2012 (UTC)
Here, for instance, 270 edits at 23:58. SpinningSpark 10:34, 20 May 2012 (UTC)
Wow. OK, I agree that the bot policy should be adapted to have a maximum speed for bots to edit - whatever the method. Even if the tasks is performed 99.99% of the edits correct, there never will be a need for such a high speed. What would be a reasonable maximum speed, even for a bot that would be perfect? --Dirk Beetstra T C 11:37, 20 May 2012 (UTC)
I think that 1 per second is a good max speed and that would top out at about 60 per minute. Kumioko (talk) 12:52, 20 May 2012 (UTC)
A couple things should also be noted here. Technically AWB will limit the number of edits being done if the server tells it too (thats a very untechnical way of saying it) based on server load. AWB limits the number of edits done per minute so the reason the count is so high is because Rich probably has multiple versions of AWB open at the same time, possibly doing different types of edits. Kumioko (talk) 13:17, 20 May 2012 (UTC)

ClueBot III ongoing problems.

I've been pointing out bad behavior with ClueBot III (talk · contribs) for quite a while now. Posts on the talk page outlining the troubles have for the most part been ignored. The problems still continue. The most recent notice I made is here. Please note that each thread I've started has a link back to the previous thread. So the link given above will lead to them. At this point I didn't know where else to go for help so I'm posting here. Brad (talk) 20:37, 18 May 2012 (UTC)

Could you give a few examples and links of this bad behaviour? The best way to deal with it is to email the bot's operator directly and ask them to check the code for any bugs or anything, and work it out with them. Rcsprinter (yak) 19:35, 21 May 2012 (UTC)

ClueBot

I don't know how responsive the operators of the ClueBots are, so I am posting this notification here in case someone else wants to look into this. See User talk:ClueBot Commons/Archives/2012/May#Minor glitch in bot edit summary. -- Toshio Yamaguchi (tlk−ctb) 10:45, 22 May 2012 (UTC)

If I understand this problem correctly, then I know what the real problem. The way it seems, nothing can connect to the database containing the reverts ClueBot NG made. It uses these edits and those that were false positives to obtain better results in future reversion. Each reversion it makes, it creates an ID that links to the database. Without access to the database, no link. This may explain why I appear to be seeing more false positive than normal. The solution, reestablish the link to the SQL database and ClueBot NG should function normally. I believe the database is stored on the Wikimedia Foundation Labs.—cyberpower ChatOnline 19:57, 22 May 2012 (UTC)
ClueBot's Web applications show the following error message; someone should contact Ryan to see if this can be fixed on his end or if an issue with ClueBot's client configuration needs to be fixed. — madman 20:00, 22 May 2012 (UTC)
I just shot Cobi an eMail.—cyberpower ChatOnline 20:02, 22 May 2012 (UTC)

RM bot stupid

Editor has been provided the answer, they just don't agree with it (talk→ BWilkins ←track) 00:00, 23 May 2012 (UTC)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Hi. RM bot did A, but it was wrong, so I tried to push the Emergency-button -- not existing, and ended up at the talk page of (me no choice) HardBoiledEggs, who did not react. (Somehow a stranger solved it?) -DePiep (talk) 23:24, 22 May 2012 (UTC)

Um, so I don't fully understand the above, nor do I understand the post you left on HardBoiledEggs ... you should always properly link to the bot and diffs (talk→ BWilkins ←track) 23:30, 22 May 2012 (UTC)
The bot homepage should always have a emergency-button first, not? -DePiep (talk) 23:31, 22 May 2012 (UTC)
Not necessarily. You go to WP:ANI, you provide diff's of the problem, and admin takes care of it ... after all, stopping the bot usually involves blocking it (talk→ BWilkins ←track) 23:32, 22 May 2012 (UTC)
Nonsense, BWilkins (remember this is bot owners' page).
A bot should have an emergency button. Go ANI when a bot is going astray? What would an ANI admin do? Yes: push the bot-emergency-button. What else? Talk to an unrelated absent user page? -DePiep (talk) 23:41, 22 May 2012 (UTC)

Formal compaint: bot RM bot has no emergency button. After this, the Talk page links to a personal talkpage (i.e. the bot operator may be off line). -DePiep (talk) 23:45, 22 May 2012 (UTC)

Are you freaking kidding me? You've been told the answer: go ahead and try to stop my bot using the emergency button. Doesn't work for you because you're not an admin? Now click on its talkpage ... oh wait, where does it lead to; my talkpage? You have the answer (talk→ BWilkins ←track) 23:51, 22 May 2012 (UTC)
I tried to stop User:RM bot. Where is the red button? (Must say, you are overly rude now.) -DePiep (talk) 23:58, 22 May 2012 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

CHECKWIKI errors needs to be fixed

Yobot is still blocked even after code changes in AWB that prohibit changes is the rendered html output is the same. (The discussion is in User_talk:Yobot#Blocked). Some editors believe that Yobot should not keep fixing some of the errors described in WP:CHECKWIKI (BAG approved this task) even if these are being fixed by other editors/bots.

Some whitespace issues occurred lately were bot owner's mistake.

CHECKWIKI has over 100 participants in en.wiki and it's a well-established task in many other wikis. AWB fixes many of the errors reported bu the project. More details in User:Magioladitis/AWB and CHECKWIKI.

The admin who blocked Yobot seems to disagree with the CHECKWIKI project as a major task of editors/bots. I have contacted them in User_talk:Materialscientist#Yobot.

I would like also to stress that WP:COSMETICBOT is not well written since they are many general fixes of AWB that are worth to save a page. There are many bots around that do part of AWB's general fixes.

I think the bot must be unblocked by another admin and the process to continue. -- Magioladitis (talk) 10:15, 1 May 2012 (UTC)

I commented on the original thread that was linked just above. I think BAG needs to revisit the BRFA for this task, to clarify exactly which CHECKWIKI tasks are approved. That would make it clear what the bot is authorized to do, which would help convince people not to block it. — Carl (CBM · talk) 11:12, 1 May 2012 (UTC)
The confusion is simply that it's the "minor fixes" that should be skipped, not "general fixes". Rich Farmbrough, 17:00, 1 May 2012 (UTC).
While most of AWB's "major" genfixes are good enough to warrant an edit, there are several AWB's non-minor fixes that should also be skipped, such as template redirection, or the removal of the |format= field with null value when the url is an HTML page, or renaming |date= to |year= in citation templates, etc.... Headbomb {talk / contribs / physics / books} 17:09, 1 May 2012 (UTC)
Two particualr items that I think BAG needs to comment on are moving all interwikis to the bottom, and replacing e.g. <i></i> with wikicode, when nothing else is done at the same time. The full list of CHECKWIKI reports is at [11]. — Carl (CBM · talk) 17:14, 1 May 2012 (UTC)
The issue with italics is different, this relates to the nature of the rendered page, and a separate switch. If the Wiki markup for italics renders differently from the XHTML markup then counting it as different may be a good thing. Rich Farmbrough, 06:36, 2 May 2012 (UTC).
Some evidence that this is actually beneficial is needed before we send off bots to mindlessly purge the wiki of html tags. Even if the page does actually render differently (and I'm not sure that it does) what is it that makes the bot's choice of style superior to the original editor's choice? Another example of this kind of behaviour is replacing <references> tags with the templated {{reflist}}. There is no good reason to be doing this kind of thing unless a style guideline actually demands that it is done a particular way. SpinningSpark 10:52, 8 May 2012 (UTC)
These things are generally guidance. For myself I have a horror of making demands, I believe that editors should edit, everything being equal following MoS, grammar rules, spelling correctly. Things are however not equal, and i would much rather someone write about neo-palladium architecture with HTML anachronisms, spelling errors, spoonerisms and all, than not write about it, because some jumped up rules lawyer keeps templating his ass. (And this is the story we hear in the blogosphere all the time where we have anything firmer than guidelines.) Let the gnomes and bots clear all this away so that with our editing hats on we needn't worry about it. Rich Farmbrough, 12:38, 8 May 2012 (UTC).
Of course spelling errors and spoonerisms should be corrected, and in any case this is a strawman since yobot does not carry out these functions. That is entirely different from what you describe as "HTML anachronisms" and should not be lumped together. Are html italics against guidelines? Personally, I find it extremely irritating when an article I have written or worked on is gone through by someone systematically changing the reference formatting, or spelling system, or whatever with no substantive contribution of their own. If I then add something to the article, I am faced with the choice of conforming to somebody elses imposed format, reverting and risk an edit war (and edit warring with bots is a particularly unrewarding experience), or get slapped down at GA or FA review for inconsistency. For a human editor to do something like this because they are working on the article and it fits better with how they can work is acceptable. For a bot process it is not. SpinningSpark 13:34, 8 May 2012 (UTC)
Renaming date to year should probably be removed altogether. Rich Farmbrough, 06:36, 2 May 2012 (UTC).
I don't think renameing date to year if only the year is present is a bad thing nor does it violate the rules since it does in fact change the rendering of the page. I also think that Most of the Checkwiki errors are perfectly fine except maybe one or 2 but even those are not really worth a major rewrite of the code if that would be required. The problem with leaving blank or invalid fields is that people tend to copy the template to other articles which just compunds the problem. Or they try and fill it in and wonder why it doesn't work. As for the template redirects, I agree that should be marked as minor edits although there are exceptions. For example, Replacing Portal box with Portal since Portal is a protected template and Portal box is not. There are several other examples of this as well. I also don't think BAG should have to individually approve every Checkwiki error, thats just a waste of time. The Checkwiki errors have been deemed by the community as problems that need to be corrected, that means it meets consensus. There is no reason to over politicize this and turn it into another Minor edit witch hunt. Kumioko (talk) 13:30, 2 May 2012 (UTC)
@Headbomb: I also think redirect bypassing should be optional and turned off in bot mode. -- Magioladitis (talk) 17:47, 2 May 2012 (UTC)
Bypassing redirects and such is fine in bot mode, as long as it's not the only thing being done (WP:NOTBROKEN). Headbomb {talk / contribs / physics / books} 18:01, 2 May 2012 (UTC)
Bypassing redirects is absolutely not fine for a bot to do automatically. It is useful if the bot can establish that the redirect is an alt name, misspelling, synonym etc. However many redirects are from titles which should have articles. When the article is created the bot will then have created a redirect to the wrong place. SpinningSpark 10:39, 8 May 2012 (UTC)
On second thoughts, this is never a suitable bot task, see WP:NOTBROKEN. At best it will introduce unwanted code clutter into the article, and in the case of misspellings, the spelling should be corrected, not bypassed. SpinningSpark 11:00, 8 May 2012 (UTC)
Sorry I should have been more clear in my statement above. I also think it should probably be turned off in bot move however as long as a more significant edit is being done I would be fine with it. Kumioko (talk) 17:59, 2 May 2012 (UTC)
Agreed. But this is not the main issue discussed here. CHECKWIKI error fixes is a well populated project and I don't understand why my bot (see: bot flag) can't do it while many editors (see: no bot flag i.e. more visible in watchlists, less easy to check correctness if it's done by random editors) participate in the project without restrictions. -- Magioladitis (talk) 19:32, 2 May 2012 (UTC)
Check wiki fixes are, AFAIK, fine for bots to do. I can't say that ALL checkwiki fixes should be fixed, as there might be some corner cases, but by far and large are not problematic. And AFAIK, the only problem Yobot had was it made a series of useless edits (like moving stub templates before/after interwiki links, or on changing some whitespace). Headbomb {talk / contribs / physics / books} 19:38, 2 May 2012 (UTC)
But it seems this is only one of the reasons Yobot keep being blocked. I am trying to reduce these things but I keep getting complains even if the edit is an error fix. -- Magioladitis (talk) 19:43, 2 May 2012 (UTC)
Yes there seems to be an exceptional contempt for AWB use and Bots by some users these days. Kumioko (talk) 19:54, 2 May 2012 (UTC)
The thing is, with a reasonably productive bot, especially a wide ranging one, you can expect a constant stream of visitors commenting, of whom a certain percentage will be complaining. Most of them will be satisfied with a polite explanation of what the bot is doing, most of the rest with a real attempt to fix any problems. There are always, though, what Iain M Banks calls the awkwardistas. Little can be done about these characters, they will latch on to the tiniest thing and keep coming back, day after day, month after month and even year after year. The community at large will not see the behaviour for what it is, just a long series of complaints that reflects badly on the botop. For that reason I have instituted a bug tracking system in the hope that it will enable comprehensive data on bugs and alleged bugs to be used objectively, as well as to handle bugs better. I believe that the data will allow me to comprehensively dismiss some of the base calumnies spread by the more unsavoury elements of our community, and, of course, other botops are welcome to borrow the idea, and templates. (Bot suuport for numbering etc might even be forthcoming.) Rich Farmbrough, 02:00, 4 May 2012 (UTC).
There is a complaint that not all CHECKWIKI fixes are "worthy", unfortunately no discussion seems possible, unless some sensible negotiation takes place. Mag offered error 61 as "worthy", which would probably be widely accepted, but the only response is a discussion about various projects rating systems.
If people want to resolve this issue they need to be prepared to negotiate. I understand, of course, that some people will not want to resolve it, and for others the price of negotiating outweighs the benefit of resolution. Nontheless as a functioning community that is what we need to do. Therefore I intend to make a working proposal, the idea is that this is a best endeavours agreement by Mag to address the issues at hand.

Yobot will make edits under the auspices of the CHECKWIKI bot request only where the edit makes a visible correction to the page.

This seems to be the touchstone for most, and correspond to Mag's aims. Comments, suggestions? Rich Farmbrough, 12:25, 8 May 2012 (UTC).
I still think it would be better to get approval for individual CHECKWIKI tasks, so that it is clear what the bot is authorized to do. Once there is approval, nobody can complain at the bot operator. — Carl (CBM · talk) 12:58, 8 May 2012 (UTC)
Agreed, it is the lack of precision in what Yobot has been approved for that has led to the block in the first place. SpinningSpark 13:40, 8 May 2012 (UTC)
I still think that forcing an individual BRFA for each individual checkwiki task is a pointless waste of time for the operator, for BAG and for the community as a whole. If BAG wants to look at the list and determine that some should not be done alone then that seems reasonable but to require individual submissions just because 2 out of a million users think they need to make a point about the merits of minor edits is petty and pointless. Kumioko (talk) 13:53, 8 May 2012 (UTC)
That is a strawman argument, no one has suggested a BRFA for every task, only that the intended tasks should be listed at the BRFA. Not everyone is familiar with the inner workings of AWB and cannot judge what is being approved with a carte blanche for all AWB functions. Counting the number of opposers (with some ad hominem comments thrown in as well) is hardly a convincing argument on a site that is supposed to work by consensus and arguments from policy, not voting. The illogicality of that is only highlighted by the failure to count the number of supporters. I could equally well state that this is supported by only n editors (I haven't bothered to actually count them). I would have expected a higher standard of logic in debate amongst those who write code, but apparently not. SpinningSpark 14:55, 8 May 2012 (UTC)
Your kidding right? CBM said, and I quote "it would be better to get approval for individual CHECKWIKI tasks". Now I admit I am not an English professor but it seems fairly clear what that statament means. Especially in context to CBM past contempt towards edits he feels are not needed. A feeling that not all of us possess BTW. Kumioko (talk) 15:30, 8 May 2012 (UTC)
If that's what he meant then I agree that is not very reasonable, but perhaps he can answer for himself. SpinningSpark 17:26, 8 May 2012 (UTC)
The existing BRFA here is completely vague about what actual tasks are being performed. Yobot has had chronic problems with making edits with a "checkwiki" edit summary that don't seem to be related to any checkwiki task, but since the BRFA is so vague it's impossible to tell whether the edits are authorized. A more specific BRFA would say exactly what CHECKWIKI tasks would be performed by the bot. — Carl (CBM · talk) 19:05, 24 May 2012 (UTC)

<Indent>Magioladitis: not inviting me and saying "The admin who blocked Yobot seems to disagree" .. and thus "the bot must be unblocked by another admin and the process to continue" is just rude and is not how WP works. Rich: the idea that any edit is good if it improves, no matter the magnitude of the improvement would be fine in ideal world, but such edits do cover vandalism, and just rogue edits which need checking. Just to give an easy example: if an article is missing integral parts (reflist, categories, etc) it is usually more appropriate to revert last edit(s), at least partially, than try to fix the problem with a bot. If redirect has been recently changed, this might be an editing conflict (recreation of non-notable article or vandalism), i.e., many bot fixes, including checkwiki, may require editor's attendance and individual evaluation. Rich: "Yobot will make edits under the auspices of the CHECKWIKI bot request only where the edit makes a visible correction to the page" fails to specify "visible correction" - even adding/removing a single space may be a useless or valid edit. Materialscientist (talk) 01:16, 9 May 2012 (UTC)

I apologise Materialscientist. I had no intention to exclude you from the discussion. I assumed that you followed the discussion in bot's talk page and you were aware of this discussion. -- Magioladitis (talk) 09:25, 9 May 2012 (UTC)

And yet once more the discussion remains unfinished?... -- Magioladitis (talk) 20:32, 19 May 2012 (UTC)

After the Arbcom decision against Rich some people are probably scared to be next so doing any type of edit that does edits that even one editor perceives as minor are going to be hard to do. Its really sad that Wikipedia has devolved to this point, where doing edits is discouraged. Kumioko (talk) 13:13, 20 May 2012 (UTC)

@Carl (CBM): Was the answer in italics satisfying for you or do you still think it should not be done? -- Magioladitis (talk) 16:09, 24 May 2012 (UTC)

I have to be honest here Magio. I realize that CBM has a particular view on doing certain types of edits but I don't think you should have to get his personal approval for the task. If the BAG folks could comment about the task and are ok with it, which they seemed to be, then you should not have to sit idle. Of course thats my view of things but hopefully others will comment as well. Kumioko (talk) 16:21, 24 May 2012 (UTC)
We are in the weird situation to discuss if a task that is done by many users and it's part of a wikiproject can be done by my bot instead of being done randomly and with higher risk of human err. :( -- Magioladitis (talk) 16:31, 24 May 2012 (UTC)
I must have missed it, but I don't see an answer in italics. One issue with the CHECKWIKI fixes in general is that it isn't clear whether all of them are in line with community norms. Many previous checkwiki tasks have been "deactived" for enwiki (see the list at [12]) - examples include #18 "capitalize the first letter of every category link" and #11 "HTML named entities". In general, just because something is listed at CHECKWIKI doesn't mean there is community consensus to make the change here (particularly because CHECKWIKI covers projects from several languages). So individual users ahve to be very careful about blindly changing things that are listed at CHECKWIKI, and bots need to have some sort of consensus to point to before they do it. — Carl (CBM · talk) 18:59, 24 May 2012 (UTC)
Yobot only does errors active in enwiki. Moreover, disactivated tasks don't grab entries from toolserver (i.e. 0 items even if the list of Yobot get outdated for some reason). -- Magioladitis (talk) 19:26, 24 May 2012 (UTC)
Sure; I was just saying that not everything in CHECKWIKI is necessarily vetted here for enwiki consensus before being added there. — Carl (CBM · talk) 19:55, 24 May 2012 (UTC)
And Frescobot checks CHECKWIKI id 64 (Link equal to linktext). There are probably more. Yobot's advandange is that fixes multiple errors reducing the number of times a page is visited. -- Magioladitis (talk) 07:42, 25 May 2012 (UTC)
That "advantage" is the reason the bot was blocked, though. Over a long period of time Yobot has never been successful at the goal of doing lots of approved things in one edit without also doing unapproved things. — Carl (CBM · talk) 11:52, 25 May 2012 (UTC)
I don't think this was the reason. I think the reason is that there was not enough discussion and some admins are not informed of what edits are done by bots and what edits should be done. yobot gwent under the spotlight due to the many edits it does while similar edits are done by other bots and editors with unknown error rate. Yobot is highly active and is natural people to expect more of it. -- Magioladitis (talk) 09:36, 26 May 2012 (UTC)
Yes it was the reason. The blocking admin has clearly stated "repeated violaion of WP:COSMETICBOT despite numerous notes on the talk page". Several other admins have subsequently supported the block. To suggest that all these admins are "not informed" or that there was "not enough discussion" is to entirely fail to address the substantive issues and goes a long way to explain why Yobot is still blocked now. The blocking admin explicitly proposed a clear way forward on Yobot's talk page. Magioladitis has shown no inclination to take that advice either at Yobot's page or here. SpinningSpark 11:25, 26 May 2012 (UTC)
Personally, I am not much interested in the inner workings of Yobot, AWB, or the bot approval process. I am only interested in what the end result is going to be. It is entirely arse about face to start the discussion from the point of view of what is, or is not, in AWBs CHECKWIKI. The discussion should begin with what we want the bot to do and not do. The result of that can be used to determine whether or not AWB is a suitable tool. Arguments I have heard from Magioladitis and others in the past that edits are done because they are included in AWB are simply not good enough to justify an action. SpinningSpark 11:38, 26 May 2012 (UTC)
The discussion shows to me that COSMETICBOT needs to change. It's not good enough and prohibits editings in the right direction. Yes, I still believe that blocking Yobot for edits done by others shows that someone is not informed of what is going on. Unless, you are willing to go and block everyone participating in CHECKWIKI or everyone who saves without providing a nice edit summary. Some edits that annoyed a small fractions of admins are instructed by the Manual of Style and this has to be discussed. There is still a mystery which CHECKWIKI ids should (not) be done and the reason is that the blocking admins do not seem to have a common opinion on that. The exact problem of Yobot's edits is still unclear and it's only based on a very small fraction of edits, a problem which is under improvement. COSMETICBOT doesn't really help because it is generic and contradicts common practice. -- Magioladitis (talk) 13:44, 26 May 2012 (UTC)
I also agree with Magio, we shouldn't be bickering about making edits to articles. A lot of these edits would be required for an article to be promoted past GA and some even for B-Class. Additionally, like Magio said these are being done already by other bots and other people. It really is a minority of people who don't want to do edits here and I can't seem to understand why there is such a major problem with doing edits. Kumioko (talk) 14:42, 26 May 2012 (UTC)
The argument that these edits are needed to pass some kind of quality rating really does not hold much water. The vast majority of articles edited by Yobot are not on their way to GA. Yobot will not bring them any closer to GA (although it might, with a dozen or so other incremental edits, succeed in turning a poor hoax into a difficult to detect one). To get to GA requires the attention of a human editor who will be quite capable of running AWB over the article if necessary. Many of the CHECKWIKI tasks will make no difference whatsoever to GA pass or fail, and those that do will be quickly fixed on review. Bot input to these processes is trivially small.
I am surprised that you don't understand why editors object to cosmetic changes. The answer is very straight forward. Bot generated cosmetic changes fill up the watchlist - a problem not caused by human editors making the same edits, but on a smaller scale. The watchlist is the main working tool of every editor and a heavy run of bot edits scrolls important information off the bottom, or even worse, hides other edits behind the bot's. We can, of course, cope with this, but it makes processing the watchlist more difficult and time consuming. Citation needed on your "minority" claim - which you have repeatedly made without justification. To see the importance of the watchlist to the editor community at large one has only to witness the furore caused by the recent addition of bolding for unvisisted changes. I happened to like this change but it was very rapidly suppressed due the strength of objections. SpinningSpark 16:35, 26 May 2012 (UTC)
Actually you are wrong on several points you are trying to make. First every edit brings articles a little closer to GA regardless of how trivial. We're not talking about removing blank spaces here. I don't get the impression that Magio has a problem with making some changes to the code but he is trying to get folks to discuss what edits are appropriate with the knowledge that other editors and other bots along with an Entire WikiProject are doing these edits. I also think the argument that these edits make a potential hoax articles hard to get is a pretty weak counter argument. Its almost as weak as the argument saying I don't like bot edits on my watchlist which ironically you use in the next paragraph. I would much rather a bot does these so I can set my watchlist to ignore bot edits than to have them individually done. The citation on the minority claim is that its the same 5 or 6 editors saying it. The rest of the community either doesn't really care or is like me and would rather these minor edits be done by bot. Back to the point. If the bot was breaking articles or screwing up formatting that would be different but again I would much rather a bot does these tiny little things than to have to waste my time fiddling with it when I'm building up an article. Kumioko (talk) 17:07, 26 May 2012 (UTC)
You say I am wrong but offer no evidence that I am wrong. "Every edit improves articles" does not equal "every edit brings articles a little closer to GA". Find me a single GAN in the archives that failed due to an issue a bot could have fixed. Find me a couple where there was even a comment by the reviewer on an issue a bot could have fixed. The fact is that the vast majority (perhaps all) of Yobot's tasks would not cause an article to fail GA if they were not done because they are simply not requirements of the Good article criteria.
he is trying to get folks to discuss what edits are appropriate. During the course of this discussion I have had cause to object to at least three proposed tasks. Other editors have also objected to various tasks. On the other hand, we have heard nothing from Magio on the proposed tasks for Yobot. It is more than likely that there will be others that are troublesome, but I just don't know at this stage.
The comment about hoax articles was an aside, not intended as a principle argument. That's why it was in brackets. My apologies if that was not understood and I withdraw it as an argument. Nevertheless, it is a fact that a number of long lasting hoaxes on Wikipedia have probably had their life extended by being cleaned up by bots and humans. In my opinion, both humans and bot owners should think carefully before cleaning up an article that is tagged as a hoax, or otherwise suspect. It may just be giving it more credance, and hoaxes are a way more serious issue to us than minor formatting issues. There are other tags as well that might be cause not to clean up an article, especially ones that have notability or verifiability issues.
I could set my watchlist to ignore bots. Bots could stop making crazy edits I need to revert. Neither of those things is likely to happen soon.
The citation on the minority claim is that its the same 5 or 6 editors saying it. This comment which has resurfaced a number of times in this debate is getting a little tired and I am beginning to become rather irritated by the implication. In what way does that indicate they are a minority? A similar claim could be made regarding those on the other side. The idea that our views don't count because we have said it before is insulting, wrong-headed, and failing to actually address the real issues. What is more, it is simply untrue, at least for some of the editors here. I have only recently started watchlisting this page. Materialscientist clearly doesn't watchlist it - he was unaware of this thread altogether until I informed him. You don't seem to understand what a citation is, so let me explain it to you - that is anecdote, not evidence.
The rest of the community either doesn't really care or is like me and would rather these minor edits be done by bot. No evidence again, you really have no idea what the community thinks. The best evidence is what the community have created in policy, and they have created WP:COSMETICBOT. Other than that, go run an RfC to find out. SpinningSpark 19:52, 26 May 2012 (UTC)
You should hear yourself. You want me to provide evidence when you are giving the same sort of arguments to counter my arguments. No I cannot provide links to an Article GA or better that failed due to bot edits but if you show me a GA or better article I can show you an article that could have gotten there quicker of the person wouldn't have had to perform every single edit manually. WP is a very fickle environment and people generally speak up about things they care about so when its the same 5 or 6 editors with personal grievances to a certain type of edit that are always the ones complaining, that is proof that the majority of the community either agrees with the edit or doesn't care about it. If they cared they would be lining up to complain and they aren't. Its the same half dozen editors. Some of which I might add rarely find the time to edit, they just complain on talk pages and fumble around in places like ANI. Reading through all the comments you have left on this talk page the one thing I see abundantly clearly is you saying IDONTLIKEIT. You and a couple of editors not liking it is not a reason not to do the edit. Kumioko (talk) 21:52, 26 May 2012 (UTC)
I also think its pretty silly that so many people are quoting COSMETICBOT when there is really only one sentance of ten words that apply. That statement is written in such a vague fashion that it leaves it open to interpretation what a minor edit is and even refers to the AWB general edits and saying many apply. What does that mean? Which ones apply? When do we enforce it? I also think its pretty stupid to throw it in peoples face like their doing something wrong when it says at the end of it that its generally ok to do these edits if a more significant one is done. Again in extremely vague language. Kumioko (talk) 22:03, 26 May 2012 (UTC)
Ok, I'm done discussing this with you. This is degenerating into insults. Let me know when you have something rational to say. SpinningSpark 23:50, 26 May 2012 (UTC)
Its obvious we are going to have to agree to disagree but no one is insulting anyone here. Your just getting mad because I have been around long enough to know better than to believe the comments you are trying to sell represent the majority of the community. They do not, they do not even represent the majority of the administrators. Kumioko (talk) 00:00, 27 May 2012 (UTC)

Rich Farmbrough's bots

As you may already know, ArbCom has indefinitely restricted Rich Farmbrough from using any automation whatsoever on Wikipedia, which means his bots have been blocked. I have compiled a list of tasks currently listed as being performed on an ongoing basis by his bots; if any of these tasks would be missed, another bot operator should file a BRFA to take them on. If any are missing, feel free to add them to the end of the list.

  1. Create missing monthly clean-up categories (link)
  2. Update oldest backlog list at Category:Wikipedia backlog (link)
  3. Update recent chages oages for projects (Wikipedia:Bots/Requests for approval/Femto Bot 6)
  4. Correct "External link" to "External links" (link)
  5. Correct "Reference" to "References" (link)
  6. Various ISBN fixes:
    • Correct "ISBN:" to "ISBN" to allow wikimagic to prevail (link)
    • Correct ISBN hyphenation (link)
    • tag invalid ISBNs (link)
  7. Other minor improvements to layout of ISBN to allow wikimagic to prevail (link)
    • Insert correct hyphenation of ISBNs, label articles with invalid ISBNs (too short, too long, lang code, pub code undefined
    • Check check digit, categorise articles with invalid ISBNs.
  8. Replace tags {{Disambig}} and {{Cleanup}} with {{Disambig-cleanup}} (link)
  9. Add references section and appropriate mark up (link)
  10. Unlink days of week and month names
  11. Remove ":" from redirects (link)
  12. Remove unnecessary capitals from section headers ("many hundreds" of common errors)
  13. Hyphenate adjectival uses of nn mile (link)
  14. Fix up the name of navboxen to be the actual name of the template (link)
  15. Clean up certain members of YEAR in categories (link)
  16. Link "Expand language" tags to their correctly interwikied counterpart (link)

Do note that some of the above may no longer be necessary, may not have consensus anymore, or should be done in a different manner. Anomie⚔ 19:55, 15 May 2012 (UTC)

I've already written code for #1 in the list above, BRFA will be filed shortly. Anomie⚔ 19:55, 15 May 2012 (UTC)
I could try to take some of them on.—cyberpower ChatOnline 20:52, 15 May 2012 (UTC)
Yep I gotta say this is so much better for the community and for the project than letting Rich continue to edit [in my best sarcastic tone]. Kumioko (talk) 01:16, 16 May 2012 (UTC)
I have offered to Rich to let me resurrect Helpful Pixie Bot into one of my own bots so that he doesn't feel that his scripts and his hard work is lost. He is strongly considering it right now.—cyberpower ChatOnline 01:24, 16 May 2012 (UTC)
Kumioko, I know you are bitter about Rich's Arb case, but this is not the place for it. Please take it elsewhere. Anomie⚔ 01:59, 16 May 2012 (UTC)
Re cyberpower: that would work, but at the same time you would need to turn off the various non-consensus changes that HPB was making. One of the main factors that led to the sanctions was Rich's implementation of non-approved cosmetic changes in addition to the approved tasks. — Carl (CBM · talk) 02:08, 16 May 2012 (UTC)
Yeah I think its pretty stupid alright. But on a more serious note I noticed that Headbomb Denied Rich's pending bot tasks. Shouldn't someone also go and revoke the other ones that were previously approved? I realize it wasn't exactly revoked but someone should probably mark them in some way. Kumioko (talk) 02:10, 16 May 2012 (UTC)
Nobody usually goes through and changes old task requests when an operator stops using them. I don't see how it would help to edit them all, Rich already knows about the case. — Carl (CBM · talk) 02:13, 16 May 2012 (UTC)
The difference here is that his bots were blocked and the tasks, if retained at all will be under the pervue of another operator. I also think that it would probably be better if these tasks went back through the BRFA process. As CBM stated there are a lot of things in there that some editors have a problem with and this would make it easier to keep the drama level down.Kumioko (talk) 02:17, 16 May 2012 (UTC)
There is precedent: when Lightmouse was returning from being banned, we formally revoked his old approvals to make it clear that those approvals were no longer valid. Anomie⚔ 03:02, 16 May 2012 (UTC)
You're right, IMO. I was planning to propose that in (probably) a few days, as I was hoping to get this squared away first. Anomie⚔ 02:54, 16 May 2012 (UTC)
Occasionally I have a good idea. :-) Kumioko (talk) 03:09, 16 May 2012 (UTC)
I would go through code and shut down any cosmetic changes it makes if I resurrected Helpful Pixie Bot along with the ISBN task since it appears to not work properly.—cyberpower ChatOffline 09:57, 16 May 2012 (UTC)

Please be careful with tasks 6 and 7. Helpful Pixie Bot loved to break perfectly valid citations that it didn't recognize, and would do so repeatedly, even after you reverted it. Also, adding dashes into ISBNs where they weren't before is a waste of edits, it's purely an aesthetic change, and one that many people that use ISBNs don't like, as some library search engines get all confused by dashes. I normally let other people's bots go about their business, but if an ISBN bot breaks my citations again, I'm going to get pissed, again. Sven Manguard Wha? 03:31, 16 May 2012 (UTC)

I wonder if there is a relatively easy way to check whether these two tasks are worth doing...that is, if there is a significant amount of improperly formatted ISBN numbers on articles. I think Sven's point is well taken, and requesting approval for any of these tasks could very well open a can of worms due to the controversial nature of the bot that performed them. Therefore it seems that any task that is put up for re-approval should be something that is proven to be worth doing in the first place. Robert Skyhawk (T C) 03:55, 16 May 2012 (UTC)
Since I've wandered up to this page somehow, I should say what was probably said elsewhere - watchlists of most regular editors contain several hundred or more pages containing ISBNs; those ISBNs are usually copy/pasted without dashes, and last week the Helpful Pixie Bot has filled up our watchlists with itself, adding hyphens. Fixing ISBNs is great, but mere addition of dashes to ISBNs (purely cosmetic change) should not be a separate bot task. Materialscientist (talk) 04:01, 16 May 2012 (UTC)
I agree, and along with the question I raised previously wonder which "errors" in ISBN formatting actually have an effect on the technical functioning of the encyclopedia. We certainly don't need to have bots running around reformatting numbers whose sole purpose is to be copy/pasted into Amazon or library catalogs by humans. Robert Skyhawk (T C) 04:08, 16 May 2012 (UTC)
I dislike having dashes inserted into ISBNs because it forces future editors of the article to continue inserting dashes if ref style consistency is to be maintained. Many popular databases do not display the ISBN with dashes, including the three most used (at least by me) Gbooks, WorldCat, and Amazon, so they cannot be simply pasted in any longer. There was a huge sigh of relief from the community when the rampaging Pixie was finally stamped on. Please do not recreate it. SpinningSpark 09:58, 16 May 2012 (UTC)
Not only that, some databases will not accept a query with dashes in the ISBN so they have to be taken out again by any user who actually wants to make use of the information in Wikipedia articles. SpinningSpark 10:23, 16 May 2012 (UTC)
Lets not be overly dramatic. There were a lot of us that thought that the majority of the edits the bot was doing was fine. Kumioko (talk) 11:10, 16 May 2012 (UTC)
I don't see the point of that comment. This sub-thread is discussing only dashes in ISBNs (which is the task Pixie was doing when it came to everyone's attention). Are you disputing the facts: that some databases do not use/will nto accept dashes? SpinningSpark 13:28, 16 May 2012 (UTC)
No I am not debating that some databases don't like it I am debating that "There was a huge sigh of relief from the community when the rampaging Pixie was finally stamped on". I believe some sighed in relief, many of us sighed in disappointment. Kumioko (talk) 15:49, 16 May 2012 (UTC)
I would be interested to know which significant databases contradict the ISBN standard. I already asked for this information. Rich Farmbrough, 16:47, 16 May 2012 (UTC).
Rich, at the time I wrote that I was thinking of Amazon, but looking at it again in more detail, that is not so, or at least only partially true. It is more to do with the way I was entering the data. I will put the full details on your talk page if you like. SpinningSpark 21:25, 16 May 2012 (UTC)

I could do many of these by using AWB's general fixes! But some people complain that this is against WP:COSMETICBOT. When Rich tried to fix the wording there, he has been reverted. -- Magioladitis (talk) 09:00, 16 May 2012 (UTC)

Let's wait for now until I get a reply from Rich.—cyberpower ChatOffline 09:57, 16 May 2012 (UTC)
I think AWB already does some of them but in general I would say that AWB should probably include 4, 5, 8, 9, 10, 11, 12, 13 andn 16 from the list above. Also parts of 6 and 7 I think. I think most of these are non contentious. Kumioko (talk) 11:16, 16 May 2012 (UTC)
Maybe non-contentious, but many of them are merely cosmetic. SpinningSpark 13:06, 16 May 2012 (UTC)
Thats fair and I could agree that they may not need to be done as individual edits which makes it all the better to add them to the general fixes section of AWB. That way when folks are making other changes these will get done then. I suspect if its desired Magio could add them as minor edits so that if the skip if minor box was checked AWB would simply skip them. Although I would argue that several of these would be required of the article was submitted to anything GA or higher so that being the case, IMO, they should be allowed. Kumioko (talk) 13:20, 16 May 2012 (UTC)
I also seem to remember that 4 and 5 are already in the gen fixes. Kumioko (talk) 13:22, 16 May 2012 (UTC)
Yes, but not all projects were set up with Femto Bots to monitor recent changes. The Texas project has been in existence for years, but it never was part of the Femto Bot run. Nor is the Women's History project set up for it. I don't know who sets up the Watch List pages for projects, but these two projects have nothing like that. And it was Rich who switched the Hawaii watch list over from the WatchList bot to Femtobot6 in 2010. Maybe this is like TedderBot, where there's a page somewhere that projects have to be signed onto before the bot runs for them. I find no instructions anywhere on how to do this, so it lies with Rich (or his successor). Maile66 (talk) 00:29, 22 May 2012 (UTC)
Oh, I see at Wikipedia:WikiProject United States/Recent changes, that it was indeed Rich Farmbrough who set up the WPUS recent change page in 2010. And at the bottom of the main page of Wikipedia:WikiProject_Council, it specifically says Femto Bot is by request for the individual projects.Maile66 (talk) 01:17, 22 May 2012 (UTC)
I wasn't commenting on your post here or at WP:BOTREQ, just on Rcsprinter123's comment that Femto Bot tasks needed doing too in case someone missed that they were already on the list. Anomie⚔ 02:04, 22 May 2012 (UTC)
OK. Then we just wait for all the task assignments to work out. Maile66 (talk) 11:14, 22 May 2012 (UTC)

In reference to task #5 above, I have been attempting to identify which pages contain a section called "Reference" instead of one called "References," in order to see if the number of pages that have this error is significant. But this search term: "==Reference==" (with the quotation marks) gives myriad false positives using Wikipedia's search engine (both via Special:Search and AWB), suggesting that it is ignoring the equal signs. Obviously a dump scan using AWB would allow me to accurately find sections like this, but is there any better way using an online engine? Robert Skyhawk (T C) 21:24, 21 May 2012 (UTC)

Since there didn't seem to be a better way, I went ahead and downloaded a dump to scan for such mis-formatted sections. AWB found 1291 pages that contain either "External Link" or "Reference" sections. I am planning an approval request to fix these with AWB. BRFA filed. Robert Skyhawk (T C) 21:55, 24 May 2012 (UTC)

WRT task 16, please be careful when doing this. When checking the articles created by Jaguar (currently at WP:AN, I noticed that HPB had added a translation link to Vita, Spain. However, that article didn't have any interwiki links, and the translation link added by HPB was apparently a random, alphabetically close article, Viriville [13]. Looking further, the same happened e.g. at Zamčanje, Zasad and Zeoke (Lazarevac), which all three linked to the Serbian equivalent of Zaisenhausen (again, alphabetically a close article)[14][15][16]. Another wrong link appeared at Xiva[17]. I don't know how many of these incorrect links were added, quite a few apparently, but they need to be found and corrected, and anyone taking over the task should try to avoid making the same error. Fram (talk) 12:32, 23 May 2012 (UTC)

I'll take on Task 8 and maybe a few others.—cyberpower ChatOnline 21:05, 24 May 2012 (UTC)

BRFA filed here (part of #6)  Hazard-SJ  ✈  02:04, 29 May 2012 (UTC)

DPL bot still working?

Does anyone know if DPL bot is still adding dablink notifications to editors' talk pages? I added a link to an page I know was a dab page, yet DPL bot didn't ping me. I checked the bot's contributions, and sure enough, the bot hasn't been giving notices to editors. I added a message on the talk pages of DPL bot and JaGa (the bot's operator) but no one responded. Is the bot broken or something? David1217 23:43, 1 June 2012 (UTC)

Most likely there was a blip in whatever script that runs on and it stopped, and JaGa hasn't been online to restart it. However approved bots are not required to run, so maybe it dropped that task. Rcsprinter (warn) 15:54, 3 June 2012 (UTC)

Delinking dates

A proposal has been offered to create a bot to delink dates. Please comment at WT:MOSNUM#Delinking dates. Jc3s5h (talk) 12:35, 21 June 2012 (UTC)

Bots delinking Romanian interwikis

I've been noticing a number of different bots removing interwiki links to Romanian Wikipedia articles over the past couple of months, and it's not clear to me why they're doing that, since the links appear to be valid iw links to the appropriate pages on ro-wiki. Can someone with better knowledge of interwiki bots explain why the bots are doing that? (Examples: from April, from today). 28bytes (talk) 12:36, 22 June 2012 (UTC)

I believe those two examples you gave are as a result of the interwiki links pointing to disambiguation pages.  Hazard-SJ  ✈  01:04, 23 June 2012 (UTC)
They're supposed to, aren't they? Plenty of interwiki links on Baker County, for example, pointing to disambiguation pages in other languages. What's different about the Romanian disambiguation pages? 28bytes (talk) 02:20, 23 June 2012 (UTC)
Unless ro:MediaWiki:Disambiguationspage is incomplete, the affected templates are retrieved from the working copy of the wikipedia family file and we need a bug tracker to update it. I've done it in pyrev:10415, see also [18].  @xqt 06:18, 23 June 2012 (UTC)
Thanks! 28bytes (talk) 16:23, 23 June 2012 (UTC)

The Un-authorized use of a Bot in the Latvian Wikipedia by User in the Chinese Wikipedia, being also an authoized Bot-owner in the English Wikipedia, or, an (a possibly) un-authorized Global-Bot.

[19]

[20]

[Copied verbatim from the Chinese, in English.]

"Hi, I just noticed your bot is running in the Latvian Wikipedia without a required bot flag. According to the policy a request must be made for every bot before actually running them. Thank you in advance for your cooperation and best regards," -- zh:User:Edgars2007 (User:Edgars2007; lv:User:Edgars2007)（zh:User_talk:Edgars2007 (User talk:Edgars2007; lv:User_talk:Edgars2007)） 2012年6月25日 (一) 07:17 (UTC)

— KC9TV 22:13, 25 June 2012 (UTC)

Is there a bot being run on the English Wikipedia? Is it unauthorized? Every language project has their own rules, their own admins, etc. (talk→ BWilkins ←track) 22:54, 25 June 2012 (UTC)
Yes, it is; even on the title, I said so, and I was merely giving a "heads-up", so to speak, especially to himself. This is not a vindictive report (and I am too sickly to do any such thing at the moment, anyway). — KC9TV 22:58, 25 June 2012 (UTC)
If his English bot is approved and not causing problems, then this report is simply being a "tattle-tale" ... it's moot for this project (talk→ BWilkins ←track) 23:17, 25 June 2012 (UTC)
  • The point is, the English language is not his first language, and he is probably not very good at it either, and he might in fact had mistaken the "local" English approval for a "global" one, — a global approval; and these, at [21] and at [22], appear to suggest that it is so. Perhaps you, Sir, care to brush up your Chinese by informing him of this (which is anyway perhaps best done by an administrator)?
  • However, is this really a matter for Meta instead? Well, I don't know, and I am sorry if that is the correct avenue instead. — KC9TV 01:19, 26 June 2012 (UTC)
If there's no issue with his English bot, and the problems span multiple Wikipedia's, then YES, meta is obviously the location - as you already answered. ANI certainly is not the right location. ✉→◌BWilkins◌←✎ 12:19, 26 June 2012 (UTC)
Answered in ANI.Justincheng12345 (talk) (urgent news here) 15:28, 26 June 2012 (UTC)

Marking inactive bots so

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.
Closed per request at Wikipedia:Administrators' noticeboard/Requests for closure. I have read Wikipedia:Administrators' noticeboard/IncidentArchive754#Block Review, User talk:Rcsprinter123/Archives/10#May 2012, and this discussion.

The consensus is that Rcsprinter123 (talk · contribs) is permitted to mark bot user pages in Wikipedia:Bots/Status/inactive bots inactive with several caveats.

1. Rcsprinter123 should verify whether the bots are truly active by checking the contributions page and the log page. Some bots' revisions do not show up in the contributions page but in the logs page. Failure to do so may lead to editing restrictions or blocks. Wikipedia:Bot policy#Bot-like editing states (my bolding):

Human editors are expected to pay attention to the edits they make, and ensure that they don't sacrifice quality in the pursuit of speed or quantity. For the purpose of dispute resolution, it is irrelevant whether high-speed or large-scale edits that involve errors an attentive human would not make are actually being performed by a bot, by a human assisted by a script, or even by a human without any programmatic assistance. No matter the method, the disruptive editing must stop or the user may end up blocked.

The community rejected Rcsprinter123's position that:

As for whether or not the bot in question is active, that shouldn't be my problem because the people updating the inactive list should have weeded out the editing ones. It is hardly my fault that 7SeriesBOT was on that list, because someone else must have made the mistake of not seeing they do deleting only instead. I had, and have put my trust in that list to tell me which pages to do, and if it is checked and updated often enough, I don't see any problem with what I am doing. Rcsprinter (converse) 19:54, 11 June 2012 (UTC)

This is reflective of a poor attitude incompatible with doing this semi-automated task. Participants have found that Wikipedia:Bots/Status/inactive bots has contained errors in the past. As 28bytes (talk · contribs) notes, "Propagating wrong information across the 'pedia is indeed the problem of whoever does it, and if you're not willing to take responsibility for the edits, you shouldn't be making them."

2. To prevent friction, if an inactive bot's operator has edited within the past three months, Rcsprinter123 should ask them if their bot is inactive. He should give the operator one week to reply, after which, if there is no response, he may tag the bot as inactive. If the operator's response is to disagree with the changing, he is advised to "just pop their bot into an 'ignore' file and put a copy of the list of bots you're ignoring into a prominent place related to your monitoring activities" (from Tony Sidaway (talk · contribs)). If he would like to tag the bot as inactive over the bot operator's objections (this is inadvisable), he must gain consensus first at a community venue like Wikipedia:Bot owners' noticeboard. Cunard (talk) 23:28, 14 July 2012 (UTC)

I am requesting permission to continue marking every bot userpage in Wikipedia:Bots/Status/inactive bots inactive by adding the |status=inactive parameter to the {{bot}} template found on most userpages. I believe this is a simple task for which I would not need permission, as the list is regularly updated and bots who are editing taken off and added to the active list. I am here however in the light of the recent thread on my talkpage in which I was blocked by Bwilkins for doing it 'without consensus'. I am confident I can get the green light without too much further ado, because if it's on the inactive list, you might as well be able to see it right away when you visit the userpage, plus it adds it to the inactive category. Thank you. Rcsprinter (message) 12:51, 10 June 2012 (UTC)

Support - I agree it should be done but it doesn't surprise me that someone stopped it. It seems these days almost anything is a contentious edit to someone no matter how needed or benign. Additionally the argument that the bot owner should be consulted is IMO silly because in many cases the bot is inactive because the operator is also and in all honesty the owner should have done it and we shouldn't have to but thats not what happenes. Kumioko (talk) 13:45, 10 June 2012 (UTC)
What a messed-up situation: an unapproved (semi?)-automated task, and an admin who is clearly emotionally invested in the subject issues a long block and then harangues all who disagree. What exactly is the procedure you're seeking approval for here, i.e. what checks do you do between reading the bot on Wikipedia:Bots/Status/inactive bots and marking the bot inactive? Anomie⚔ 17:32, 10 June 2012 (UTC)
I hope you're not referring to me as "emotionally involved" :-)
If you'll notice, my primary points in marking possibly inactive bots were:
not all bots edit pages, so marking them as inactive when they have not made edits is improper
any task that marks bots inactive should first check if the bot owner is inactive - this makes the marking of bots as inactive to be more appropriate/easy
if the bot owner is still active, ask if their bot is inactive
This isn't rocket science (talk→ BWilkins ←track) 17:49, 10 June 2012 (UTC)
I agree with Bwilkins. In the case of 7SeriesBOT (talk · contribs), it doesn't edit, but speedy deletes G7-ed pages. See the log. Thus marking the bot as inactive is inappropriate. Armbrust, B.Ed. WrestleMania XXVIII The Undertaker 20–0 18:07, 10 June 2012 (UTC)
Agreed but the vast majority of those from the inactive bots list are indeed active and if one or 2 accidentally get marked someone who knows is going to notice and revert. Its easily solved but its better than leaving a large number of inactive bots looking to the general public as though they are still editing. IMO, we need to take this one step further and, after they are marked inactive for a month or so, giving ample time for review. They be removed from the AWB bots list of applicable. IMO the block was unnecessary and the task itself is needed. Kumioko (talk) 18:31, 10 June 2012 (UTC)
I never said the marking of 7SeriesBOT inactive was correct, or that this automated/semi-automated task had approval. Anomie⚔ 18:41, 10 June 2012 (UTC)
I was referring to you, actually. It appears to me as an uninvolved observer that you took great offense that your bot was marked "inactive", and the demands that Rcsprinter123 must agree to undo all the edits if unblocked point all the more to emotional rather than logical decision making there. But beyond hoping you'll be less quick with the block button in the future when it's your own bot involved, I see no reason to dwell on the matter. I'll also point out that the number of bots that are active without making edits is very small, and the number that make logged actions without also editing is even smaller (what is it, yours, a few other deletion bots, and two proxy-blocking bots?). I do think that "check for logged actions" and "ask the operator if they appear to be active" are certainly good suggestions. Anomie⚔ 18:41, 10 June 2012 (UTC)
I didn't think I required approval for a wholly human-controlled task, after all AWB is there just to assist us, I could have added all the tags manually had I cared to. As for whether or not the bot in question is active, that shouldn't be my problem because the people updating the inactive list should have weeded out the editing ones. It is hardly my fault that 7SeriesBOT was on that list, because someone else must have made the mistake of not seeing they do deleting only instead. I had, and have put my trust in that list to tell me which pages to do, and if it is checked and updated often enough, I don't see any problem with what I am doing. Can I see some more supports and opposes please? Trying to get a consensus, not an argument here. Rcsprinter (converse) 19:54, 11 June 2012 (UTC)
Blindly trusting a list that was recently updated by a bot is not the wisest course of action. If you want to do this in the future, I'd suggest checking the contribs and log activity of the bot for activity, and if the listed operator is still active asking them about the bot's status wouldn't hurt either. Anomie⚔ 14:50, 12 June 2012 (UTC)
Would have to agree with this since I know the list of inactive bots has been wrong in the past. -DJSasso (talk) 23:37, 17 June 2012 (UTC)
Clearly you don't need the words "oppose" or "support" to read that you don't have consensus. The majority of discussions on Wikipedia don't use such words overtly (talk→ BWilkins ←track) 20:16, 11 June 2012 (UTC)
With respect Bwilkins, one oppose also does not infer that he lacks consensus. Kumioko (talk) 20:38, 11 June 2012 (UTC)
And with all due respect, others have noted their desired changes to his MO for this task - which is, indeed, all I have suggested as well (talk→ BWilkins ←track) 13:14, 12 June 2012 (UTC)
  • A bit off-topic, but a simple question no one has asked yet: Why the hell is it so important that inactive bots are marked as such? If a bot goes inactive (and it was doing something halfway important), then generally a thread is started here so that the tasks can be taken over by someone else. Apart from that, I don't see any use in spending your time modifying templates so that a category that no one ever looks at is accurate and up to date; especially when it generates drama like this. Is there nothing else more important that needs to be done around here? Maybe I'm missing something. If it were me, I'd be looking for something more productive to do than tagging inactive bots or blocking people who are tagging inactive bots. -Scottywong| gossip _ 15:47, 12 June 2012 (UTC)
  • It's an odd one, this. I think it's useful to mark as inactive those bots that don't appear to be actively editing the wiki, so I support the activity. But I would hate to see this marking become something that might hamper the activity of useful bots that have to cease activity for long periods because the user cannot attend to them. So please be careful and responsive about your actions. If somebody objects just pop their bot into an "ignore" file and put a copy of the list of bots you're ignoring into a prominent place related to your monitoring activities. --TS 00:07, 18 June 2012 (UTC)
  • Not what I'd call a high-value activity, but seems harmless enough. I can't imagine an incorrect tagging would break anything. However, this bit: "As for whether or not the bot in question is active, that shouldn't be my problem because the people updating the inactive list should have weeded out the editing ones. It is hardly my fault that 7SeriesBOT was on that list, because someone else must have made the mistake of not seeing they do deleting only instead." is completely the opposite of what I want to see from any bot operator. Propagating wrong information across the 'pedia is indeed the problem of whoever does it, and if you're not willing to take responsibility for the edits, you shouldn't be making them. Change that attitude and take full responsibility for your edits and I'll have no problem with this task. 28bytes (talk) 12:47, 22 June 2012 (UTC)
  • Support - why not? mabdul 18:10, 27 June 2012 (UTC)
The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.

Request: "Bad link" advice for interwiki bots

Some time ago I came across a spurious interlanguage link that connected hantavirus to the musician mg:Hanta from Madagascar. I removed the link, only to find it soon re-added by a bot. I eventually realized that I needed to remove all of the incorrect links from the articles in all the other-language Wikipedias. I did this -- quite a daunting task, given the difficulty of verifying the topic and editing an article in an unfamiliar language -- and it worked for about eight months. After that, a bot added it again, for reasons that are not clear to me. Looking at the talk pages for bots that handle interwiki links, I see that I'm not the only one who has had this frustrating experience; and I see others being advised to remove the link from all the Wikipedias. This is bad advice: first, because fixing an error in an article in one language shouldn't require editing in a list of other, unfamiliar languages; and second, because it doesn't work permanently or reliably. I only now learned that this can be solved simply and easily by commenting out the interwiki link. I'd like to request that bots that add interlanguage links make this experience easier by including information in the edit summary and the bot user page that will give clear and correct advice on how to fix incorrectly added links. Thanks. --Amble (talk) 16:09, 27 June 2012 (UTC)

The reason it happened looking at it is that a number of bots used the -hint:en parameter in pywikipedia assuming that the article named Hanta on en would be the same one as on mg. But it wasn't, it was a redirect. This is technically human error. Not actually bot error persay. The problem with the commenting out method you have now used is that it will forever prevent proper interwiki links from being created unless done manually. -DJSasso (talk) 16:15, 27 June 2012 (UTC)
Why will that prevent proper interwiki links? It should only affect mg:Hanta. If it has something to do with the comment appearing in the middle of the list of interwikis, perhaps mg:Hanta should be commented out and moved out of the list. Ryan Vesey Review me! 16:23, 27 June 2012 (UTC)
I take his comment to mean that bots will not create an interwiki link from en:hantavirus to a hypothetical Malagasy-language article on the same topic. Interlanguage links to/from other articles should be fine. --Amble (talk) 16:26, 27 June 2012 (UTC)
(ec)It will also prevent any future interwikis for a mg:hantavirus article that is created cause bots will skip interwikiing it because it already has a link en and will consider it a double interwiki situation which bots will skip. That in turn will also mean bots will skip any future changes to the Hantavirus article on all wikis because of the double link from mg. So it will affect up to two articles on every langauge that has an article on each of the topics. The easier fix would be to (however badly it may turn out) is create a stub on mg for hantavirus and link it to the en article...this will prevent the other from being added as an interwiki when a bot operator makes a mistake and says that hanta should be linked to hantavirus. -DJSasso (talk) 16:30, 27 June 2012 (UTC)
Thanks. That's not much of a problem, compared to the larger problem of incorrect links being persistently recreated. If some day someone does (hypothetically) write an article on hantavirus on the Malagasy-language Wikipedia, I'm sure it won't take long for the correct interwiki link to be created. If you believe that's a significant problem, could you suggest an alternative way of cleaning up the incorrect links that is both reasonable and effective? Whether the root cause involves human error or not, I'm asking for help in avoiding unnecessary frustration and futility in trying to clean up after these bot edits. --Amble (talk) 16:26, 27 June 2012 (UTC)
Usually these situations are cleaned up quickly and easily by removing the offending link from all wikis. But you have already said that is something you don't like. And I do understand it doesn't prevent another bot operator from accidentally thinking the Hanta article on both wikis is the same. But there isn't much we can do to stop such human error. -DJSasso (talk) 16:32, 27 June 2012 (UTC)
I disagree that repeatedly editing a number of articles in unfamiliar languages is quick or easy. And human error is at most a small part of the problem here. Any such human error is rapidly compounded by bot edits. The idea of asking users to create stubs in unfamiliar languages, as the easiest way of fixing a small error in their primarly language Wikipedia, is not sensible. Commenting out the interwiki links is better: if bots have a proven history of making a mess of the interlanguage links in a given article, then it's a good thing to stop them from editing those links. It's perfectly find if manual intervention is required to create a hypothetically correct link at some future date. --Amble (talk) 16:37, 27 June 2012 (UTC)
The example of hanta is extremely rare. It is not often that an article on one wiki is a redirect on another with the same spelling. So the solution to the hanta case is not the solution I would suggest for all articles, just the solution for this one. As for the others, it most certainly is easy to just go and remove interwikis from langauges you don't edit in a matter of seconds. I managed to fix the hanta problem in less than a couple of minutes because all you are doing is deleting the mg link. The fact that the other languages are not my languages doesn't matter because I don't actually have to read the article. I just have to be able to read the bad interwiki. As for the interface itself, I just change my preferences to show me english for all the interfaces, you do that once and you never have to do it again. It is completely unacceptable to stop all bot activity on multiple articles when the fix is a very simple one. That is like trying to kill a fly with a cannon. -DJSasso (talk) 16:41, 27 June 2012 (UTC)
Where is this preference that shows English for all the interfaces, that you only need to set once? --Redrose64 (talk) 16:57, 27 June 2012 (UTC)
I didn't say it was a preference for all interfaces. I said I change my preferences (note the s) to show me English for all interfaces. I only have to do it once as in once for each language, meaning I don't have to change it every time I go back to that wiki. So the initial time I make a fix there will be a slight increase in time to make my interface English, but any future visits it will still be English. -DJSasso (talk) 17:10, 27 June 2012 (UTC)
You managed to "fix" the hanta problem in the same way I did, in 2009. Unfortunately, that's not adequate for several reasons. First: as you yourself said, you haven't read the articles. If you can't read Cyrillic, you have no idea whether the Ukrainian-language article was about the virus, the musician, or neither. Second: even if that is the correct solution, most users confronted with this situation will waste time repeatedly reverting the bots before they find that out, because the bots give no guidance on how to revert their incorrect edits. Third: it's a fool's errand, because one bot or another will add it back sooner or later. But really, all I'm asking is that bots making these sorts of edits should link users to sensible guidance about how to repair them -- preferably in the edit summary and on the bot user page. Some bots, such as User:ClueBot NG, do already include such a helpful note in their edit summaries. --Amble (talk) 17:18, 27 June 2012 (UTC)
Actually it is easy to tell if it is the same article or not even if you don't speak the language, you go look at the categories its in and the interwikis on those categories. Again takes all of a couple of seconds. The guidance is on the interwiki page it tells you how to fix incorrect interwikis I believe. However, I believe it gives you the same advice you reject here. That being said incorrect interwikis are not the end of the world. If someone doesn't know how to fix them or doesn't feel comfortable going to those wikis to fix them, someone who does will. Looking at that page it only happened a couple of times a couple years apart. That isn't really a big problem. -DJSasso (talk) 17:29, 27 June 2012 (UTC)
It only happened a couple of times only because the rest of the time, nobody was putting in the effort to prevent bots from re-adding the incorrect links. I didn't reject the advice: I tried it, and found it was not effective. More broadly, I'm a bit troubled by the implication that wasted editor time, unnecessary frustration, and spurious interwiki links are minor concerns, while the important thing is to avoid getting in the way of (hypothetically useful) bot edits. This seems to me to be precisely backwards. --Amble (talk) 17:46, 27 June 2012 (UTC)
An additional note: apparently, when I cleaned up the same problem in 2009, I did include mg:Hanta in an HTML comment, successfully protecting the article from having the spurious interwiki link reinserted. However, this was broken by the following edit a few months later: [23]. That edit was a user on AWB, which set off the avalanche of bot edits. Another user removed the interlanguage link (yet again) on Hanta, but did not know to clean up all the other languages, so a bot incorrectly restored it (yet again) shortly thereafter. The point remains that the correct response to this is not obvious, and users have unnecessarily gone through a lot of frustration and wasted time as a result. --Amble (talk) 18:38, 27 June 2012 (UTC)
A bot readded mg:hanta. There has to be a better solution to this. Ryan Vesey Review me! 13:52, 28 June 2012 (UTC)
Wow, it happened even sooner than I expected. I guess I was right to put the supposedly quicky and easy "fix" in quotation marks. What happened is that a user from English Wikipedia, User:LeadSongDog, re-added the link at mg:Hanta. To me, this proves the points that a technical measure is needed to stop the bots from re-adding the incorrect interlanguage links, and that blind editing of articles in a language you don't understand is the last thing we should be encouraging people to do. --Amble (talk) 14:42, 28 June 2012 (UTC)
Again this is a human problem, not a bot problem as evidenced by the fact it was a human who caused the problem. We have to assume that people are doing the right thing. -DJSasso (talk) 15:00, 28 June 2012 (UTC)
It is entirely a bot problem. There are situations where bots are unable to realize that a human has made an error and where human error is likely. The only solution to this is to add a hidden comment around the interwiki link. Ryan Vesey Review me! 15:03, 28 June 2012 (UTC)
Which then affects multiple articles on multiple wikis. That isn't an acceptable solution. The solution is simple if you aren't willing to fix the incorrect wikis you ask the bot op to. That is part of the required conditions to run a bot like this. To fix their mistakes. Either way in this particular case it won't happen again because someone has re-created a dab page at Hanta which someone erroneously changed to a redirect. IW bots only link dab pages to dab pages so it won't happen again, and explains why it hadn't happened on that page in years. -DJSasso (talk) 15:05, 28 June 2012 (UTC)
You're again saying that it hadn't happened on that page in years. This is entirely wrong. The reality is as I've already pointed out: the reason why it hadn't happened in years is simply that the incorrect links have stood for years, except for brief periods when someone removed them. Even when all the incorrect links on all languages were removed, it took months (in 2009) or one day (just now) for them all to reappear. --Amble (talk) 15:17, 28 June 2012 (UTC)
Yes, it only took a day this time because someone trying to help you messed up. Either way this is the solution until the centralization of interwikis in one location is completed. -DJSasso (talk) 15:33, 28 June 2012 (UTC)
I was responding to your specific claim "it hadn't happened on that page in years", which (as I said) was quite incorrect. In response to which you tried to change the subject. Also, I'm still troubled by the implication that the overriding concern is to avoid tying the bots' hands in any way, and needless editor frustration and wasted time are an acceptable trade-off. --Amble (talk) 15:41, 28 June 2012 (UTC)
We save more editors time by being able to use bots than we lose due to them. So yes it is an acceptable trade off that once in awhile we have to fix a problem like this. The alternative would be having to do it all manually wasting more editors time and causing more frustration. -DJSasso (talk)
That's a bit of an exaggeration: commenting out one interwiki link in one article is note remotely the same as "having to do it all manually". And you're still not acknowledging that your "it hadn't happened on that page in years" is incorrect. Why? --Amble (talk) 15:51, 28 June 2012 (UTC)

What is the rationale for having multiple bots wheelwarring, doing the same wrong edits that were previously reverted? Shouldn't there be a partitioning of effort so that only one (or at most a few) bots will be in play? Shouldn't there be a check for prior similar edits before committing? Shouldn't these bots be exclusion compliant? These are pretty basic measures. Is interwikilinking the wild west of botdom? LeadSongDog come howl! 18:26, 28 June 2012 (UTC)

Some of these points are certainly worth considering. Per policy some the bots running interlanguage maintaining scripts at toolserver are denied since april. As we are expecting m:Wikidata I no longer feel caused to spend a lot of time into that script, maybe except of some bug fixing if necessary.  @xqt 06:32, 8 July 2012 (UTC)
Some procedural notes (I'm not responding to this discussion in depth as I'm on vacation): Interwiki bots are exclusion-compliant, and interwiki bots have been consolidated on the Toolserver due to concerns about memory usage, not the bots' behavior. Cheers, — madman 13:47, 9 July 2012 (UTC)

No bots are handling {{Cite jstor}}

I used {{Cite jstor}} on Vincent Price (educator); however, no bot is handling it. I assume Citation bot should be the one that did it. I even clicked the "jump the queue" button, but could find no way to have the bot take action. Ryan Vesey Review me! 17:12, 10 June 2012 (UTC)

Could you specify how bots are supposed to handle this template, and why they should do this instead of relying on editors to do it themselves? It isn't fair to expect volunteer programmers to monitor all templates and volunteer to support them. --TS 00:10, 18 June 2012 (UTC)
To be fair to Ryan the instructions at Template:Cite jstor do say that "a bot will automatically generate a full reference". There is a small amount of discussion on the templates talk page as to what the problem may be. However, there also seems to be no documentation as to what bot carries out this task and this seems an obvious place to ask for confirmation of what bot does do it. Dpmuk (talk) 00:37, 18 June 2012 (UTC)
Citation bot is supposed to handle it, thus it is an issue with that particular bot. Materialscientist (talk) 00:41, 18 June 2012 (UTC)
Sorry, I missed all of this. User:Smith609 isn't actively editing. Should {{Cite jstor}} be deprecated or should a notice be left? What's the solution. Ryan Vesey Review me! 02:26, 23 June 2012 (UTC)
Citation bot *is* still operating, and its documentation seems to suggest that it needs to be pointed at pages for translation. (Which strikes me as a bit suboptimal, cite jstor could easily put pages into a category, and a bot could traverse that category, but nevermind that.) --j⚛e deckertalk 21:30, 16 July 2012 (UTC)
Interesting: I pointed it at Vincent Price (educator), it clearly found the right page, but didn't believe any changes were warranted. --j⚛e deckertalk 21:49, 16 July 2012 (UTC)
Tried following the instructions at User:Citation bot/use to look at pages "on demand", but it seems the bot is turned off. GoingBatty (talk) 00:30, 17 July 2012 (UTC)
If you use {{Cite doi}} instead with the jstor doi, it still does not work. The bot is having trouble. There is a category: http://en.wikipedia.org/wiki/Category:Pages_with_incomplete_JSTOR_references

Along with the comment above I have a question. What does having a seperate template for Jstor offer us? Why not just use Cite book, web or something else? Kumioko (talk) 00:11, 22 July 2012 (UTC)

All it is is a wrapper on Template:Cite doi/10.2307 blah blah blah. AManWithNoPlan (talk) 18:07, 22 July 2012 (UTC)

Unapproved bot?

User:1exec1 appears to be operating an unapproved bot. The user's contribution history shows several edits per minute with the edit summary "date formats per WP:MOSNUM. The user's edit to Julian calendar showed errors such as treating 1.14.13 as a date even though it is actually a scholarly method of indicating which section of an [had been completed, in 45 BC ancient Latin text] was being cited. Jc3s5h (talk) 18:03, 16 July 2012 (UTC)

Nope, that's not a bot, I'm assisted by a script and fully responsible for my actions. Are there any problems with that I review a number of edits at a time and then quickly save them to the wiki using a sequence of Ctrl-Tab + Enter? 1exec1 (talk) 18:13, 16 July 2012 (UTC)
The edits display errors that would not be made by a thinking human, and some of them are to articles that a human would know to take extreme care with when doing anything with dates, such as Anno Domini and Julian calendar. I believe these edits fall under the Bot-like editing section of the bot policy. Jc3s5h (talk) 18:19, 16 July 2012 (UTC)
Just wanted to note that 1exec1 has responded courteously to complaints about this problem, and doubts about his humanity are perhaps not on point. The problem is indeed acute with articles on classical antiquity, since one conventional way to cite passages from ancient texts is the numerical form XX.XX.XX (= Book.Chapter.Section), as for instance "Aulus Gellius, Attic Nights 12.2.13" (Book 12, Chapter 2, section/sentence 13). (For those who don't deal with dusty old stuff, this form of citation allows for precise location of the passage regardless of which printed or online edition might be used.) Since I rarely see forms like 07.23.13 used as a date anywhere except on movie trailers, I wonder whether automating this supposed correction is more trouble than it's worth. Cynwolfe (talk) 18:43, 16 July 2012 (UTC)
The problematic part of the script is already disabled. I just wasn't aware that this citation format is widely used and will be a problem. 1exec1 (talk) 18:47, 16 July 2012 (UTC)
Preventing future such edits is a start. Will you be re-checking your previous tagging to correct the problematic edits that were created by means of the faulty script? --EncycloPetey (talk) 19:55, 16 July 2012 (UTC)
Apparently this is a sequence of semi-automated edits that should have had a BFRA done? (✉→BWilkins←✎) 21:06, 16 July 2012 (UTC)
Why? There are, for example, many users of User:Ohconfucius date formatting script, which use it without needing an approval. Since my script is essentially a cousin of the Ohconfucius's one, why the distinction? 1exec1 (talk) 21:52, 16 July 2012 (UTC)
Read WP:MEATBOT. You may think your approach is different from what is described there, but I believe it does apply to your edits. Jc3s5h (talk) 00:03, 17 July 2012 (UTC)
I've read that section several times. I do still genuinely not understand what part of the policy I violate. The section says that:
1) the use of a bot in not an excuse for decreased quality of edits. I have never made that assertion.
2) quick editing is not, by itself, disruptive. That's what I'm arguing about. I fix any issues to my previous edits once I become aware of them. Thus my edits as a whole don't have poor quality even if some of them introduce mistakes.
Could you clarify what exact part of MEATBOT do I violate.
1exec1 (talk) 11:02, 17 July 2012 (UTC)
The claim is that your edits did in fact have a reduced quality, specifically that you have made errors that a normally attentive editor would not have made had they been actually checking the edits they were making. This is the exact thing that WP:MEATBOT is describing: even though you were manually hitting the "Save page" button, you were making edits that were automatically prepared without properly checking them for correctness. It is good that you have gone back and fixed the errors you made; now, going forward, you need to be more careful to review the script-suggested edits to be sure they actually make sense. Anomie⚔ 13:23, 17 July 2012 (UTC)

I become aware of the edits when they affect an article on my watchlist. Granted, many of the articles on my watchlist are time or calendar related. But the vast majority of edits that I saw on my watchlist were wrong. A human who was seeking to follow DATERET would consider the following factors:

  • Is this a time or date related article, where a variety of time and date formats are likely to be intentionally presented?
  • Is it really a date, or just formatted like a date?
  • Is it in a direct quote or title?
  • Excluding dates in the previous 2 bullets, is date usage consistent within each of these three categories (inconsistency between, but not within, categories is OK):
  1. Article body
  2. Citations except access date
  3. Access dates within the citation section
  • If inconsistency is found, is there an indication on the current contents of the talk page about what format to use?
  • Failing that, find the first occurrence of a full date in each of the article body, citation access date, and other citation date. Follow the formats first established. Or, ask on the talk page to establish a new format.
  • If the first version of the article with dates has mixed usage, the date format is up for grabs.

Considering the number of errors I am seeing and the rapid rate of editing, I do not believe DATERET is being considered. Jc3s5h (talk) 13:33, 17 July 2012 (UTC)

User:Ohconfucius has retired rather than address concerns of violations of DATERET at Wikipedia:Arbitration/Requests/Clarification and Amendment#Clarification request: Date delinking. I think citing his scripts as evidence of unproblematic editing is counterproductive. Jclemens (talk) 05:56, 17 July 2012 (UTC)
For the record, I also don't support the actions of Ohconfucius that violate DATERET. I don't intend to violate this policy myself. Thus I can't think of a reason why the retirement of Ohconfucius is relevant here. He abused his scripts. However, there have been a lot of users that have used the scripts without any problems. They didn't need Bot approval. 1exec1 (talk) 11:02, 17 July 2012 (UTC)
A bot approval in this case would be mainly so you could make these edits with the 'bot' flag, so the watchlist/recent changes option to hide bot edits would hide them. Anomie⚔ 13:23, 17 July 2012 (UTC)
Sure. It's very easy to check and fix such issues, since the pattern of the problem is known. The work can be semi-automated too.. 1exec1 (talk) 21:52, 16 July 2012 (UTC)
I think a mountain is being made out of a mole hill here and at ANI for the same user and problem. The user made a mistake with an edit due to a rather rare style of dating and it was fixed, the individual was notified and the problem was solved. I do not think that a bot is needed nor would it be particularly helpful in this case. Kumioko (talk) 13:31, 17 July 2012 (UTC)
I don't see anything yet in this discussion that indicates that all the problematic edits have been checked and corrected. My direct request for an answer to this received no reply. Until the problematic edits that were made all have been corrected, the problem is not solved. --EncycloPetey (talk) 21:03, 17 July 2012 (UTC)
I believe I've fixed all mistakes I've made. 1exec1 (talk) 10:27, 18 July 2012 (UTC)
  • Comment. There were a large number of 1exec1 date edits to the pi article. An edit included changing a fraction-followed-by-modal-verb "22/7 may" to "22/7 May". Glrx (talk) 18:10, 18 July 2012 (UTC)
Jclemens says: "User:Ohconfucius has retired rather than address concerns of violations of DATERET at Wikipedia:Arbitration/Requests/Clarification and Amendment#Clarification request: Date delinking. I know you've been running that argument at arbcom and that most of your colleagues don't agree with your assumption of bad faith: do you have the slightest evidence that OC retired to avoid addressing DATERET concerns?

Now, I'm surprised that Jc3 is conducting this rather aggressive campaign against anyone who harmonises the the date formatting in article ref lists—a task at which most editors, not to mention readers, wouldn't even raise their eyebrows. One senior Wikipedian said to me a while ago, perplexed, why on earth would be not harmonise them within the list, and why on earth wouldn't we harmonise them with the prevailing date format in the article. He just couldn't believe the fuss.

Sooner or later we're going to have to run an RfC, probably via WT:MOSNUM, but properly advertised this time, to gain consensus for a sane approach to the messiness that will naturally evolve with the "anyone can edit" ethic. Fixing the dates using automation (and this hardly counts as a bot or as BAG-approval stuff) should be no big deal; indeed we should be thanking editors like 1exec1 for doing what Sue Gardner refers to as "the scut work". BTW, Gimmetoo has virtually run Ohconfucius—a highly valued editor—off the project with continual threats using what many people regard as the dysfunctional MOSNUM guidance on this matter; this is counterproductive. My advice is to let these editors get on with the job that the rest of the participants here (including me) don't seem to want to have to do, and to think seriously about the 50/50 split at the last RfC at MOSNUM on DATERET, which with greater advertising might well show consensus for updating the guidance. Tony (talk) 06:40, 22 July 2012 (UTC)

At the time I started this thread, the edits I saw looked like a bot or a script with only the most cursory inspection of the results before saving. It has since become apparent that 1exec1 would like to follow DATERET and is willing to go back and fix mistakes, which is a different and better approach from what I have seen when I encountered batches of edits like these in the past from other editors. Certainly our method of choosing article style, whether for date format, variety of English, which unit of measure to list first, or whether to denote eras with AD/BC or CE/BCE is unfriendly to automation, because first it requires a subjective decision about what the article is about and whether that subject matter drives certain style choices, such as Oxford English for John R. R. Tolkien. For style choices that are not determined by article subject matter, ideally one would first search the article talk for any consensus that might have been found in the past; this isn't practical for talk pages that contain a large amount of text. Next one would find the first version non-stub version of the article that reflects a style choice. This isn't as hard as searching a large talk page, but it isn't easy either.
Certainly it would be nice to agree on a more efficient method to document style choices, but no such agreement has occurred. There are some templates that might be more-or-less recognized that partially document some choices, but these templates lack the expressiveness to document all valid choices. I'm convinced that a paper-over-the-problem approach that allows automation without a thorough consideration of all issues will lead to campaigns (by editors other than 1exec1) to change massive numbers of articles to their preferred style and execute a Wikipedia:Fait accompli. Jc3s5h (talk) 13:29, 22 July 2012 (UTC)

Code review...

...this may be the wrong place, but I'm looking for a code review on something I knocked together for this. It was genuinely knocked together in a naive way, and it would be great to get some experienced editors to point out any more elegant solutions to the problem, or good practice suggestions. The code is in python using the standard API. 

import wikipedia 

def getNominatorx(article):
          article=article.replace("Wikipedia:Featured_article_candidates/","").replace('/archive1',).replace('/archive2',)
     page = wikipedia.Page(site, article)
     try:
         original_text = page.get()
         temp=original_text.split(' ');
         print article+" length "+str(len(temp))
         return
         temp= re.search(r'small>\'\'(.*)</small', original_text, flags=0)
         if temp:
             print  temp.group(1)
     except:
       return
     
def getAllForAMonth(pageIn):
   print pageIn
   page = wikipedia.Page(site, 'Wikipedia:Featured_article_candidates/Featured_log/'+pageIn)
   if page:
       original_text = page.get()
       temp=original_text.split('\n')
       temp.pop(0)#removes the opening stuff
       temp.pop(0)#removes the opening stuff
       for line in temp:
           article=line.replace('{',).replace('}',).replace(' ','_')
           getNominatorx(article) 
       print 'end' + pageIn 
       return
   print 'page not found'    
site = wikipedia.getSite('en', 'wikipedia')
for year in range(2011,2013):
   print year
   for month in ['January','February','March','April','May','June','July','August', 'September', 'October', 'November', 'December']:
       getAllForAMonth(month+'_'+str(year))

Any suggestions? Fayedizard (talk) 23:38, 24 July 2012 (UTC)

The following will run much faster, and is much more expressive of what you want.
select
      page_title,
      page_len
from
      categorylinks
      join page
            on cl_from = page_id
where
      cl_to = 'Featured_articles'
      and page_namespace = 0
order by
      page_len asc;

3600 rows in set (0.02 sec)

-- Cobi(t|c|b) 03:26, 25 July 2012 (UTC)
oooh - yes SQL would be a lovely way of doing that - but to ask a fairly naive question (google didn't help me much) what's my 'hello world' for using SQL within pywiki? Or have I got the wrong end of the stick?Fayedizard (talk) 13:29, 27 July 2012 (UTC)
from __future__ import unicode_literals
import oursql

connection = oursql.connect(read_default_file="/home/sigma/.my.cnf", db="enwiki_p", host="enwiki-p.rrdb.toolserver.org")
cursor = connection.cursor(oursql.DictCursor)
cursor.execute("SELECT * FROM watchlist WHERE wl_title='Coal_ball' AND wl_namespace=0")
for stuff in cursor:
    print stuff

→Σσς. 18:44, 27 July 2012 (UTC)

Superstar! Thank you :) Fayedizard (talk) 08:15, 28 July 2012 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Bots/Noticeboard/Archive_7&oldid=1185028574"