Wikipedia:Bots/Requests for approval

If you want to run a bot on the English wikipedia, you must first get it approved. To do so, add a request below.

Admin: When bot is approved or dissapproved, you can use {{subst:debate top}} and {{subst:debate bottom}} to encapsulate finished discussions.

Current policy on running bots

Before running a bot, you must get approval on this page. State below precisely what the bot will do. Observe discussion and await authorisation from someone in the approvals group. Approval may take up to one week. After approval you may run the bot for a short period so it can be monitored. After the trial, the bot will be evaluated from someone in the approvals group and you may ask that the user be marked as a bot at m:requests for bot status.

Again, please DO NOT start running your bot without following these instructions.

I	Creating a user page for the bot describing its functions. please state the following: Describe the bot's purpose, language it uses, what program(s) it uses (pywikipedia framework, etc). Describe whether it is manually assisted or automatically scheduled to run. The period, if any, we should expect it to run. Describe who the maintainer is. Add the bot's user page to Category:Wikipedia bots (By adding {{bot|your user name}} to the bots user page)
II	Listing your bot here. please state the following: Whether the bot is manually assisted (run by a human) or automatic scheduled to run The period, if any, we should expect it to run What language or program it is running The purpose of your bot Why do you need it? Is it important enough for Wikipedia to allow your bot?
III	Waiting for approval. You must wait for approval from someone in the approvals group.

1. When naming your bot, please make sure that it does not look exactly like your username, and that a person can immediately determine that it is a bot. One common way is to name the bot as your username + "bot".
2. Sysops should block bots, without hesitation, if they are unapproved, doing something the operator didn't say they would do, messing up articles or editing too rapidly.
3. The operator should be at, or logged into, the machine the bot is running on to terminate it if necessary during the debugging phase, or the bot is liable to be blocked without notice.
4. If you are planning to use a "spider", recursive wget, or similar software to get a local copy of wikipedia, please download the database dumps instead.
5. Dynamic loading of Wikipedia pages may also be unacceptable. Please see Wikipedia:Mirrors and forks.
6. Operators should separate their edits from their bot. This means that you should not be logged in as your bot replying to people. Questions or concerns can be addressed at bot's talk page or the operator's talk page, but the one who is responsible for replying is the operator not the bot.

The burden of proof is on the bot-maker to demonstrate the following:

1. The bot is harmless
2. The bot is useful
3. The bot is not a server hog
4. The bot has been approved

Note that according to Wikipedia:Categorization of people certain types of person categories should not be filled/emptied using a bot. Before adding sensitive categories to articles by bot, the input should be manually checked article by article, rather than uploaded from an existing list in Wikipedia.

In the assistance to prove the bot is harmless and useful, a trial period may be asked to demonstrate the bot. Complaints made about the bot during the trial period requires the bot to be immediately stopped, and the issue should be resolved below the application. If the trial period passes with no problems, then a bot flag may be requested at m:requests for bot status.

Organisation

Active trial runs

Please document active trial runs at Wikipedia:Bots/Trial runs. When doing so, include

• Username of both the bot and the operator (denote as needed)
• The purpose of the bot
• The date and time the trial run was approved
• The name of the person who approved the trial run

The original discussion should remain on this page, but can be moved to the #Trial run holding pen.

Bot flag approvals

When a user has completed a trial run (if needed), and their bot flag has been approved, please

1. Archive the discussion at Wikipedia:Bots/Requests for approvals/Archive
2. Add an item to Wikipedia:Bots/Approval log

Current requests for approvals

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

Closed with no consensus. Martin 09:04, 11 April 2006 (UTC)

Bobblebot

I have just been trying out Martin's AutoWikiBrowser. I would like to register User:Bobblebot as a bot account to use it. The task is to reduce linking of solitary months, solitary years, etc in accordance with the manual of style.

If another bot is already doing this task, please let me know. It is a huge slow task (for me anyway) and I would rather do something else. Bobblewik 18:06, 15 December 2005 (UTC)

Um, reduce linking? I don't quite understand. --AllyUnion (talk) 11:32, 16 December 2005 (UTC)

Yes. For example Economy of Algeria has many solitary year links, including 14 to 2004. The policy and popular misunderstandings are explained at:

Wikipedia:Make_only_links_relevant_to_the_context#What_should_not_be_linked

Wikipedia:Manual_of_Style_(dates_and_numbers)#Date_formatting

For example:

• Text with links to solitary years: The short-lived ABC Cable News began in 1995; unable to compete with CNN, it shut down in 1997. Undaunted, in 2004 ABC launched a news channel called ABC News Now.
• Text without links to solitary years: The short-lived ABC Cable News began in 1995; unable to compete with CNN, it shut down in 1997. Undaunted, in 2004 ABC launched a news channel called ABC News Now.

See this diff working towards that: [1]. Bobblewik 13:32, 16 December 2005 (UTC)

A bot can't distinguish between when it should or shouldnt be linked, it needs a human to watch over it. Plus it's pretty damn easy to do with the AWB! Martin 14:14, 16 December 2005 (UTC)

There is rarely, if ever, a valid reason to link to a year article when other date elements are not present, so removing such links (with proper edit summeries) is IMO one of the safest possible editing tasks for an automated or semi-automated process. Detecting the presence of other adjacent date elemetes (and they must be adjacent dor the link to function as a date preference mechanism, normally the only valid reason for the link) is pretty purely mechanical. DES ^(talk) 22:30, 16 December 2005 (UTC)

If the dynamic date formatting feature can recognize correctly-linked dates, there's no reason a bot can't. These links are pointless, look silly and should be removed by a bot. — Omegatron 22:41, 16 December 2005 (UTC)

Oppose. I don't think bots should do the task suggested by Bobblewik nor I'm a particuarly convinced that Bobblewick bothers thinking about in which cases such links are definitely worth it, e.g. at Talk:Luxembourg_(city). -- User:Docu

Support. There is almost never contextual reason to link a solitary year, if these can be detected by a bot they can be removed by one. Hence, the bot makes perfect sense and should be run. Neonumbers 09:46, 17 December 2005 (UTC)

• I support the use of bots to remove useless links for dates. There should also be an effort to get people to stop making links for every date they see. Is there a policy on this or just this style guideline: "simple months, years, decades and centuries should only be linked if there's a strong reason for doing so"? --JWSchmidt 17:07, 17 December 2005 (UTC)
• i support the removal of date links. Does anyone use them? They are a distraction and i would be happy to see them go. David D. (Talk) 17:09, 17 December 2005 (UTC)

• The MoS says generally do not use them, not never use them. Martin 19:13, 17 December 2005 (UTC)

• I definitely support. The more linkcruft we get out of Wikipedia the better. --Cyde Weys _talk^contribs 19:19, 17 December 2005 (UTC)

• Strong support—Trivial chronological links, including such moronic items as 2005, 20th century and 1980s, degrade WP. The issues, in my view, are (1) the dilution of high-value links, (2) the reduction in ease of reading (any reading psychologist will tell you that), and (3) the messy appearance, particularly where linking is otherwise relatively dense. Please eradicate this scourge, far and wide. Most people see the light when you refer them to Wikipedia:Make only links relevant to the context, Wikipedia:Manual of Style (links)#Internal links and Wikipedia:Manual of Style (dates and numbers)#Date formatting. The post-bot relinking of one or two dates for which there might be a strong reason to link is a small price to pay. Tony 23:30, 17 December 2005 (UTC)
  • Good synopsis Tony. i agree that if the bot makes a few mistakes they can be relinked by hand. However, i have rarely seen a date link that is useful. David D. (Talk) 23:33, 17 December 2005 (UTC)
• Strong support as well for the above reasons. 99% of the time, the linked years are unnecessary. If the bot is done correctly, it should be fine. Gflores ^Talk 23:56, 17 December 2005 (UTC)
• I'm not sure this is a good idea, since, as Martin points out, it says to generally not use them. In fact, in the section on dates of birth and death (which immediately follows), it does use bare years when giving the format for when only the year is known. So removing the links in all cases seems like a bad idea. --Mairi 00:14, 18 December 2005 (UTC)
• Support. To Mairi: it should be possible to identify birth and death years from [[Category:xxxx births]] and [[Category:xxxx deaths]] and being in the first sentence.Susvolans ⇔ 17:05, 18 December 2005 (UTC)
• Support Date links have been rendered useless by overlinking. --Wetman 10:06, 19 December 2005 (UTC)
• Very very strong support This is a great idea! Not to mention saving so much of my time as I explain (once again) only necessary appropriate links, paste link to MoS/Date formatting, change tons of leeetle links....did I say this is a great idea? If there is a date which needs a link, fine we put it in, no big. The ratio must be a thousand to one. Did I say how great I think this is? Allow me to say it again: This is a great idea, wonderful bot, thank you! KillerChihuahua^?!? 10:24, 19 December 2005 (UTC)
• I Support the use of bots to remove useless links. Keep up the good work!! Armindo 11:40, 19 December 2005 (UTC)
• Comment. Bobblewik has been adding a link to this conversation to edit summaries. Example.--Commander Keane 11:52, 19 December 2005 (UTC)
• Support. I'm not a big fan of using bots for anything, but this seems to be working very well. Kafziel 12:21, 19 December 2005 (UTC)
• Support - I have to admit I had it in mind that I had to link the year on its own (as it was a separate link anyway from the day month combination). Only when the bot chanced across Royal Society did I find I was in error! Scottkeir 12:24, 19 December 2005 (UTC)
• Support – The improvement from removing these overwhelmingly useless links is worth the inconvenience of reverting the small number of mistakes it will make. – Smyth\^talk 12:28, 19 December 2005 (UTC)
• Question. Will the bot cover infoboxes? I understand that certain infoboxes include incomplete dates. ╫ 25 ^ring-a-ding 12:33, 19 December 2005 (UTC) ╫
• Oppose. This is annoying enough. If Bobblewik cares that much, he can do it manually. Ambi 12:36, 19 December 2005 (UTC)
• Support. Who the hell linked all these dates in the first place? - Randwicked 12:44, 19 December 2005 (UTC)
• Oppose Needs to be done by hand; some links to years should be kept. For early years, what links here function is extremely useful in expanding articles. Warofdreams talk 12:48, 19 December 2005 (UTC)
• Comment. I confess myself guilty of overlinking until recently. I'd vote support provided the solitary birth and death years linked at biographical articles stay like that. I'd like to see someone change solitary year links to more contextually relevant ones (for example, 2005 to 2005 in film in movie/actor articles) rather than simply removing all links; that's what the policy aims at. --Pablo D. Flores (Talk) 12:56, 19 December 2005 (UTC)
• Though I support the discussion of the issue in the Manual of Style, I'm not sure what "harm" comes from linking dates or how it "degrades" Wikipedia. Sometimes it's sort of interesting to be one click away from what else happened in 1755, even if the relevance to the article is not immediately apparent. It's not a big plus, but neither is it a big minus. --Dystopos 16:31, 19 December 2005 (UTC)
• Strong support Having a bot for this kind of cleaning up is a great idea. There are so many over-linked years, especially in lists like discographies or filmographies... --Fritz Saalfeld (Talk) 16:47, 19 December 2005 (UTC)
• Limited Support I agree, lots of date links are really of little value. But it seems to me there is a value in having the ability to see what else happened on that date (or in that year) in history. This is more interesting for the more historical (older) years than more recent ones (say pre-1950) and for events of at least some significance - Marshman 20:28, 19 December 2005 (UTC)
• Oppose. This needs to be done by hand. Removing linked dates makes it difficult to organise articles per "x in music". I fail to see the long-term use for this bot. As an alternate idea, what about setting up a bot to detect articles that have an overage of linked dates, and collecting them for user review, in some sort of cleanup/wikiproject? -Tim Rhymeless (Er...let's shimmy) 04:46, 20 December 2005 (UTC)

(moved from User talk:Bluemoose/AutoWikiBrowser by Bobblewik - please reformat comment): This bot is the bane of most history articles! J. D. Redding 20:46, 19 December 2005 (UTC) [... beginning to understand why most people are uneducated about history ..] (Ps., please don't let this loose on a timelime or similiar article... )

• Is there a reason why it's being done while we are discussing if it should be done and in which cases it might be done? -- User:Docu
  • One reason (as eluded to by J. D. Redding above) to link to years is that is provides landmarks in a history article. For example, if you are reading History of Australia but are interested in events of the 1850s, the blue linked dates are easy to see (Note: it is not my personal belief that date linking should be used, so so get cranky at me about it).--Commander Keane 01:20, 20 December 2005 (UTC)

Aha, link=highlight.

My own theory is that this is a huge unintended consequence of using square brackets to perform two entirely different functions: links, and date preferences. If date preferences could be done a different way, date preferences would not feed the link all dates myth. Bobblewik 02:21, 20 December 2005 (UTC)

• Support. Removing all those spurious year links would be good. It would be good to have a syntax the bot will not remove a link for -- does the piped syntax serve? -R. S. Shaw 05:59, 20 December 2005 (UTC)

• • You said reduce, not eliminate. Some dates must be linked. You should set the bot to eliminate repeated linked years or months across the text, not exterminate them. So, the idea is good, but not so good as I thought in the beginning. Armindo 21:31, 20 December 2005 (UTC)

• Oppose. As a bot, false-positive matches are very likely. For example, Mii-dera links to 1117; and the 1117 page has a factoid about Mii-dera. Links such as the 1117 link from Mii-dera should not be deleted, because it can provide historical context to the event concerning the temple. Neither User:Bobblewik or the author of User:Bluemoose/AutoWikiBrowser seems concerned about this problem, although these types of links can be detected via the Whatlinkshere page. For example, on Special:Whatlinkshere/Mii-dera, 1117 is listed, so it seems clear that there is a reason to keep the 1117 link on the Mii-dera page. Since most year articles mention events of that year, the Mii-dera example is far from the only case where this could happen. Neier 12:41, 23 December 2005 (UTC)

• You clearly havent read the discussion, "the author" (myself) has already objected to a bot being used for this for roughly the reasons you point out. Martin 12:45, 23 December 2005 (UTC)

• Sorry. I was going from the response on your talk page about how much bandwidth it would take to implement the check into the bot. I didn't notice that the comments above also came from you or that Martin is the sig for Bluemoose. So, apologies for thinking you didn't care about the problem. Neier 13:52, 23 December 2005 (UTC)

• No problem ;) Martin 14:03, 23 December 2005 (UTC)

• Support. I support this effort to reduce overlinking. SlimVirgin ^(talk) 22:08, 23 December 2005 (UTC)

• Strong oppose: No link is ever "useless." That is determined by the eye of the beholder. --AllyUnion (talk) 12:27, 25 December 2005 (UTC)

• This is not only not what a bot should be used for, it's not what you should spend your time doing on Wikipedia either. All these edits are incredibly annoying, they often remove useful wikilinks in appropriate areas, and they make the headings more confusing by deleting the spaces in between the equals signs and the words. I strongly oppose any automation of these tasks which more often than not make it worse. Talrias (t | e | c) 02:57, 27 December 2005 (UTC)

It might be useful in this debate to distinguish between comments about the merits of:

• Manual of Style guidance
• a bot proposal to edit articles consistently with the Manual of Style.

For example, Talrias raised a perfectly valid question suggesting that he prefers spaces in headings. I don't care either way and had not thought about it. Now that I check the Manual of Style, it does not use spaces. So the criticism of making edits in accordance with the Manual of Style does not seem fair. If the Manual of Style needs changing, then we can change it easily enough. There are plenty of changes I would like in it. Please can we could separate our judgement about the merits of a bot proposal from our disagreements with guidance in the Manual of Style. Just a thought. Bobblewik 18:15, 27 December 2005 (UTC)

Fine, I oppose the automation of tasks like this because bots get it wrong, making the article worse. I've seen cases where dates have been unwikilinked from in image captions. That's not useful and no human would delink them as "excessive". I find the intended target of this proposed bot highly unnecessary and "fixing" a nonexistent problem. Talrias (t | e | c) 18:47, 27 December 2005 (UTC)

I think you are reinforcing my thoughts that this is an issue for the Manual of Style. Dates in captions are not mentioned in the Manual. If you think they must be linked, then it is easy enough to add that to the Manual. As far as 'no human would delink them' is concerned, I have delinked plenty and I am a human (despite occasional accusations to the contrary). We should be debating these points of style in talk:Manual of Style, not here. I am not targetting you, but a lot of the discussion on this page has been people disagreeing with the content of the current Manual of Style. A bot proposal should be criticised on its merits, not on the basis that the Manual of Style that it follows is inadequate. Bobblewik 19:26, 27 December 2005 (UTC)

No, you are not reading what I'm saying. I'm saying removing excess linkage (per the manual of style) is good, but the way your script/bot has been doing it in the past is bad. Therefore, I oppose the use of a bot flag for these "cleanup" edits. I don't think this is something which should be automated. I don't think I can make it any clearer than this. Talrias (t | e | c) 22:49, 27 December 2005 (UTC)

What is the bot doing that a human being who was following the MoS would not do, apart from being faster? SlimVirgin ^(talk) 03:57, 28 December 2005 (UTC)

Getting it wrong would be a fair summary. Talrias (t | e | c) 04:00, 28 December 2005 (UTC)

Yes, of course, but what would "getting it wrong" amount to in this case? SlimVirgin ^(talk) 07:35, 28 December 2005 (UTC)

I've already explained this above, to you personally, and on a number of other pages. I've provided links to those pages above and in other places. I find your consistent badgering of me annoying. Please stop it. Talrias (t | e | c) 16:04, 28 December 2005 (UTC)

• Oppose. An automated procedure makes these reversions easy to make, but it makes more mistakes with wikilinks than a proper manual review. Policy says that: "...should not be linked ... Months, years, decades or centuries, unless they will clearly help the reader to understand the topic." -- and a quick glance in an editor (AutoWikiBrowser) can not be sufficient at considering that. Also I consider the Manuals recommendation (Manual of Style...Date formatting) not be 100% a law, as also a (recent?) note there states it to be controversial. feydey 02:39, 29 December 2005 (UTC)

• Support. While there may be a very few limited instances where a year link may be helpful for readers to understand the context of an article, the vast majority of such links are just clutter. older≠wiser 21:01, 29 December 2005 (UTC)

A further example of how this bot screws up is this edit to List of Prime Ministers of the United Kingdom, which removes unlinks the dates the particular Prime Minister in question is in office for. Talrias (t | e | c) 21:33, 29 December 2005 (UTC)

No dates were removed in that edit as far as i can see. >ots of dates were unlinked. I would have made the exact same changes manually, had I come upon that article. DES ^(talk) 21:38, 29 December 2005 (UTC)

How did he screw it up? it was much better without the over linking. Martin 21:39, 29 December 2005 (UTC)

Sorry, I didn't proofread. I meant of course unlinked. The links are of course useful so you can then go and find out what else happened in that year in question, in the same manner you would click on the link to the Prime Minister's article page to find out more about the Prime Minister. It is clearly better to have more links in an article than less, because it gives people ease in browsing to related articles. Removing them is an aesthetic difference and this should not be forced on other people - consider changing your own style preferences rather than forcing them on other people. Talrias (t | e | c) 21:43, 29 December 2005 (UTC)

1. Strongly oppose; while there may be some basis to removing date links, it's still being debated. In addition, I don't believe that running it as a bot, even a manual bot, allows the user to know what pages they're editing, and whether the date links on that page might actually be useful. Ral315 (talk) 21:47, 29 December 2005 (UTC)

• Oppose. A bot getting it wrong IMO does more harm than a link of possibly questionable merit. --IByte 22:24, 29 December 2005 (UTC)

• Oppose. I agree that there are too many year links in some places, but I don't think that using a bot to remove them is a good idea. It's something that needs more judgement than any automated system can give -- sannse (talk) 22:46, 29 December 2005 (UTC)
• I think not. Many historical subjects will have years in which actual things happened but not "other date elements". This should never be done unconditionally, especially for the tiny amount of "value" it provides. Demi ^T/_C 22:47, 29 December 2005 (UTC)

• Oppose Per Sannse for the most part. There doesn't seem to be any real reason to go through with this from what i'm hearing. karmafist 22:53, 29 December 2005 (UTC)

• Strong oppose. Style issues should be handled on a discretionary article-by-article basis, as a matter of editor judgement, rather than something cranked out by an automatic bot. More importantly, the Manual of Style is not absolute gospel, and personally I think date links are useful for cross-indexing of information by chronology. If you look at it from the perspective that date links do no harm, and clearly show some merit for finding information about dates specified in the article, I cannot see merit in a bot for the purpose. --NicholasTurnbull | (talk) 00:27, 30 December 2005 (UTC)

• Oppose. There is a contention here that date link clutter is harmful. If so, it is a very marginal harm. On the other hand they do provide an article with context, even if that is a very marginal benefit. On the whole, this seems like a wash to me and so whether or not to keep them should be left to the discretion of the article's contributors. I am opposed to anyone, bot or not, mounting a campaign to systematically remove such links. Dragons flight 00:42, 30 December 2005 (UTC)

• Qualified support. I think that the semiautomatic use of AutoWikiBrowser for reducing the date linking is fine. In the example of 14 instances of links to "2004" in a single article, delinking 13 of them with the assistance of a semiautomatic bot (requiring human intervention) seems to be a good thing. However, instead of eliminating the first occurence of the link to a year, I have often found it more useful to change the link to something more relevelant to the article such as [[2004 in music|2004]], [[2004 in sports|2004]] or [[2004 in books|2004]]. A bot could not do that (at least not without some level of AI) and I would rather leave the link for a human to fix in the future if the person manning the AutoWikiBrowser doesn't take the time to do so. -- DS1953 ^talk 01:07, 30 December 2005 (UTC)

• Oppose. A solution in search of a problem. If it is a problem in an article, just do it; if it's too much trouble, it wasn't much of a problem, was it? --Calton | Talk 03:09, 30 December 2005 (UTC)

• Oppose using a bot to do such. And please don't interpret my vote as either support or oppose of unwikifying/wikifying dates; this vote solely opposes the use of a bot to do such, but does not imply endorsement (or disendorsement) of doing so manually. (In other words, don't read into my vote too much! :-) ) Thanks! Flcelloguy (A note?) 03:15, 30 December 2005 (UTC)
• Oppose, I agree with the idea to reduce overlinking, but I think that this is a task that needs to be done manually in order to ensure that relevant links are not removed (see also my comment at Wikipedia talk:Manual of Style (dates and numbers)#Edits_which_just_unwikify_stuff. JYolkowski // talk 03:21, 30 December 2005 (UTC)
• Oppose — Dan | talk 17:11, 31 December 2005 (UTC)
• Oppose. Removing multiple date links is a good thing, but in general I agree with the opinions expressed above that year links provide a valuable tool for investigating context in articles on historical matters. When making such edits, it is very necessary to see the layout of the article as a whole, and I suspect that a bot-assisted mechanism would not provide the detail. User:Noisy | Talk 20:41, 31 December 2005 (UTC)

Can we close this disucssion now? It seems clear that there is no consensus for User:Bobblewik to run a bot. The votes that continue to pile on aren't of any beneift and are disruptive to this page.--Commander Keane 22:49, 31 December 2005 (UTC)

That might be a good idea: this is a deeply flawed poll. Take Note: "AutoWikiBrowser is not a bot: any edits made using this software are the responsibility of the editor using it." Kim Bruning 16:49, 3 March 2006 (UTC)

Also take note: There is a button (for bots) in the autowikibrowser that will make the bot go automatically. ILovEPlankton 15:16, 4 April 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

William Allen Simpson running occasionally as Botryoidal

Reading the instructions, it appears that it should be possible to run the standard py bot under our own name. I've been checking stuff by hand (orphaning templates) every day for the past several weeks that would be helped by a bit of automation. It's not currently a lot of edits per day, and it will be fairly slow as I'll be handling it late nights over dial-up. Any objections (or advice)?

--William Allen Simpson 20:46, 19 January 2006 (UTC)

You need to outline exactly what work you are doing. If this has anything to do with disambiguation templates (or {{2LC}}, {{3LC}}, {{4LC}} etc) then I ask you to discuss any changes at the relevent project talk pages.--Commander Keane 03:06, 20 January 2006 (UTC)

Of course, this is exactly the template changes that I've already posted at the relevant talk pages. I'm getting tired of doing them entirely by hand. I'm seeing a lot of repetition, and apparently simple tasks (like mere substitution) are easy to do with standard bot utilities. I won't write any additional code, and will run the utilities "as is" from the repository. Shouldn't affect the performance of this site.

--William Allen Simpson 03:36, 20 January 2006 (UTC)

You might consider using the Auto Wiki Browser for these tasks. It's as near to automated as you can get without people complaining that you use bots. (people with editcountitis, at least) — ₀₉₁₈^BRIAN • 2006-01-20 03:38

Cannot, as I'm a longtime Unix(1977)/BSD(1983)/Mac(1984)/MacOSX kinda guy. But reading more of the instructions convinced me that it would be prudent to run a separate user. So, I just checked many variants of my name, and almost everything has already been taken by usernames with no edits! (FYI: Botch, Bottom, or Bottomless are still available.) Anyway, I'll try out Botryoidal later.

--William Allen Simpson 05:01, 20 January 2006 (UTC)

Happy Happy Joy Joy! Successfully editted a single page. Will try more later.

--William Allen Simpson 06:11, 20 January 2006 (UTC)

This bot doesn't have approval. Stop using it and outline the activities explicitly.--Commander Keane 08:08, 20 January 2006 (UTC)

Gentlefolk, earlier today (middle of night local time) Commander Keene blocked both this (my) User and my new bot User:Botryoidal. At the time, I gave up and went to bed. However, I just figured out that I was unblocked:

• 2006-01-20 08:50:36 Commander Keane unblocked User:Botryoidal (collateral damage from blocking of Botryoidal)
• ...
• 2006-01-20 08:49:56 Commander Keane unblocked #84338 (collateral damage from blocking of Botryoidal)
• ...
• 2006-01-20 08:28:34 Commander Keane blocked "User:Botryoidal" with an expiry time of 24 hours (Unauthorised bot)

I followed each and every step listed for starting to use the bot. The bot was run manually, and run throttled. Indeed, I was manually running in alphabetical batches (20-30 or so edits at a time), and had just started 'E' about four (4) minutes before!

The stated rules for administrator block require that

1. "... they are unapproved, doing something the operator didn't say they would do, messing up articles or editing too rapidly."

• Certainly the bot wasn't doing anything that I didn't say it would do (it was only doing exactly one edit, and that was what I stated, orphaning a template that I'd listed at WP:TFD) several days ago.

• Certainly the bot wasn't messing up articles. I tested the first edits one file at a time by hand, and I checked each and every batch of edits on my screen before running the next batch. Heck, I'm generally considered a fairly careful and cautious "safe pair of hands"!

• Certainly the bot wasn't editing too rapidly, Special:Contributions/Botryoidal shows that the edits were throttled to 30 seconds (as required), and run in the slack time (as required).

The stated rules for starting the bot say that:

"2. New bots should run without a bot flag so people can check what it's doing.

"3. Until new bots are accepted as ok they should wait 30-60 seconds between edits."

Now, how exactly are perfectly performing bots supposed to qualify during their "initial one-week probation" demonstrating they are run responsibly, when an administrator blocks them without any valid reason?

--William Allen Simpson 15:06, 20 January 2006 (UTC)

Botryoidal has not been approved for the one week trial. It was blocked because the operator has not outlined exactly what the bot is doing and why. The operator still has not outlined that. There has already been a complaint about the types of edits that the bot is doing. The edits need to be discussed before the bot makes any more. If the bot edits without approval it will be blocked.--Commander Keane 16:26, 20 January 2006 (UTC)

So, you admit you know exactly what the bot is doing! (I didn't think I could possibly have been more clear — in fact, considerably more clear than any other bot request on this list.)

Anybody who looks at that reference will note that it is actually posted several days before the bot existed. So, you personally object to the edits I've been carefully and considerately doing for weeks by hand. Well, I don't think this is the place to re-argue a two week straw poll, that was started because of the flagrant template redirecting and category closing surreptitiously done on New Years Eve by the person you cite (Tedernst), and fairly quickly reverted.

--William Allen Simpson 18:36, 20 January 2006 (UTC)

No one has yet given you approval to run your bot, and therefore it was blocked a day after when no comment was made on whether it was approved or not. I still don't understand what you intend to do with your bot account, since I really don't see your proposal here. Just a lot of complaining. --AllyUnion (talk) 03:35, 24 January 2006 (UTC)

AllyUnion, please go back to the first paragraph. Orphaning templates. In particular, those I've currently got listed in TfD, or those of mine currently in TfD Holding Cell. Pretty straightforward work I've been doing by hand for weeks.

--William Allen Simpson 10:15, 25 January 2006 (UTC)

Now that the templates (proposed, straw polled, survived TfD, and after related CfD of Feb 20) are clearly approved, please set the Bot flag.

--William Allen Simpson 11:15, 4 March 2006 (UTC)

I strongly object to the operation of this bot.--Commander Keane 11:44, 4 March 2006 (UTC)

As do I. -- Netoholic @ 05:01, 6 March 2006 (UTC)

Object. This user overrides CFD decisions on a whim (see Wikipedia:Categories for deletion/Log/2006 March 16#Category:Human name disambiguation), so letting him loose with a bot could have untold consequences. Noisy | Talk 09:21, 9 April 2006 (UTC)

Calm down, people; let's get this straightened out without a pile on

The bot flag will not be set until approval for it is given; for this, the bot must pass a week's trial run without raising complaints. At no point did you receive approval for this, either. It may seem thoroughly unwiki, but you have to appreciate that we need to be stringent about what's doing what automated; a bot running out of control could do quite a bit of damage before we noticed and responded.

You appear to want to run a bot under the username Botryoidal on a periodic basis, for the purposes of orphaning templates where there is consensus at Wikipedia:Templates for deletion for them to be removed, correct? Rob Church 20:16, 28 March 2006 (UTC)

User:ABot update

I would like to use my bot to change instances of {{NowCommons|...}}/{{nowcommons|...}}/{{NC|...}}, where appropriate (that is, if the image name on the commons is exactly the same as the name here. The code is built on top of the pywikipediabot framework, and is availible upon request. – ABCD^e✉ 04:36, 4 February 2006 (UTC)

Support I don't see anything long with that. Jtkiefer^{T | C | @} ---- 22:47, 16 March 2006 (UTC)

Sorry, change instances of those templates to what? This request isn't as clear as I'd like. Rob Church 20:20, 28 March 2006 (UTC)

To {{NCT}}, probably... Fetofs ^Hello! 23:20, 1 April 2006 (UTC)

Hm, not really good enough. We need to know what the bot is doing, when. Rob Church (talk) 19:39, 8 April 2006 (UTC)

User:Xbot RFP

I have created Xbot for fuffilling the request of this user. I would like permition to run. The bot would only crawl through the military history category. TIA! - Xxpor 18:58, 12 February 2006 (UTC)

The bot wikilinks instances of Osprey Publishing? How does it decide what articles to check, and does it wikilink all instances in a page (not good) or does it do the first one? More information is needed before I can approve a trial run. Rob Church 01:06, 30 March 2006 (UTC)

Accessing the servers for stats

I am not really asking about a bot posting to pages itself (yet) but i run some scripts accessing all Special:Statistics?action=raw pages from Wikimedia servers to automatically create wikisyntax for pages with long statistics tables like 1,2,3, 4 and 5. I just provide the wiki syntax here though and copy paste manually. My question is now: What time intervals are ok when accessing all those stats pages? What timing can i set my cronjob to without being seen as an annoyance? (when running the scripts to update my local database) and should i think about also posting the result automatically or rather not and just offer people to copy and paste. Mutante23 20:39, 12 February 2006 (UTC)

This isn't the place to ask these sorts of questions; ask the wikitech-l mailing list, or in the #wikimedia-tech IRC channel. Rob Church 01:08, 30 March 2006 (UTC)

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

Tawkerbot

Tawkerbot is a python wikipediabot from http://meta.wikimedia.org/wiki/Using_the_python_wikipediabot. It will be running subst: on a list of userpages that I have created and checked that have templates without the subst tag.

I'm taking lists of user pages that have not used subst: on common warning templates and adding the subst in. I will be using a text file of configs.

= Config params given

As per the instructions, Tawkerbot will be run with the commands

python replace.py -file:articles_list.txt "{{test2}}" "{{subst:test2}}

with the articles_list.txt being a file created by myself and the template being the template that needs to be edited.

Sorry that I started running the bot, I misunderstood the bot procedures, thankfully I had 20 people on the IRC room trying to tell me I hard errored.

Update: Tawkerbot presently has an option for auto reverting the squidward mass vandal, I know it is a little step up from the reasons I applied for the flag however, I'm not expecting any complaints (nobody on IRC expressed any concerns besides the innocent page revert which I explained) so it is enabled in case squidward returns. Tawkerbot will only auto revert if one follows the exact conditions of the attack, it should be virtually impossible for an innocent edit to be caught by this process

Tawker 02:35, 13 February 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

LDBot

LDBot is a self named bot (Lightdarkness) which is to be created to do a few tasks. It is based off PHP and the Pywikipedia framework. The purpose of the bot would be to maintain Wikipedia:List_of_non-admins_with_high_edit_counts. The bot would check the edit counts via the toolserver of each user once a week (Spaced out in equal incriments, as to not cause stress on the server) and once every other day, update the counts at the previously mentioned page. The bot is being designed to make the stat tracking of that page automated, so it doesn't require user upkeep.

The Bot will not strain wikipedia servers at all. It will make no more than one request to the toolserver (or a wikimedia server) per hour, with updating to take place once every other day. If this behavior were to ever change, I would first bring it up here, or to AllyUnion to be sure that the bot is still within operating limits.

I've recieved the support of an active admin on that page, and another admin, which is why I've started development on this bot. Development will occur in my Usernamespace, and willnot edit the main page until the bot is approved, and testing is complete. --lightdarkness ^(talk) 03:09, 17 February 2006 (UTC)

Hvae you asked the people who run the tool servers if checking all edit counts once a week would stress their servers? Martin 09:35, 17 February 2006 (UTC)

I haven't specificly asked, I just came to the conclusion that it was common sense because I know several users who check their edit count using Interiots tool several times PER hour. I'm also very certain that Interiot uses the SQL function COUNT() which reduces the amount of processing power, but I will ask Interiot and the other toolserver admins if indeed it will be no trouble. --lightdarkness ^(talk) 12:05, 17 February 2006 (UTC)

A question, currently some users bold their user names, for Users in bold have shown an interest in becoming admins.. What do you think about this, will you add this feature.--Ugur Basak 13:13, 17 February 2006 (UTC)

Of course! Syntax for editing the users interest is as follows: "LDBot mod Lightdarkness <bold>" for making a user bold, if a user is not interested, you'd use the following: "LDBot mod Ugur_Basak (Not Interested) <strikethrough>" Where a comment is given within the parenthises, and the format type is given in < >. --lightdarkness ^(talk) 14:06, 17 February 2006 (UTC)

This sort of thing would be much more efficient to implement on the toolserver itself (eg. a slight modification of this SQL query), and it would probably much less work to just pester someone who has a toolserver account than to write the code for the bot. Even better, wikisign.org has a database dump that's currently 10 days old, and you can run that query there. All that aside, a hit or three an hour isn't going to be noticable, but there may be better ways to do it. --Interiot 17:18, 17 February 2006 (UTC)

The thing about it being on the toolserver though, is that users can't add their own comments on if they want to be an admin, no interest, links to previous nominations, ect. You do bring up a good point though, and I'll ask you about it on your talk page. --lightdarkness ^(talk) 21:36, 17 February 2006 (UTC)

A single-user count is available here, but the tool specifically says "Please note, that mass automated querying of the edit counter (for example, to generate lists of user sorted by edit count) is not allowed.". eg. it's better to pester someone with a toolserver account to get the raw data that you want, with a single query, since that will only take 3 - 5 minutes to generate. --Interiot 22:01, 17 February 2006 (UTC)

But the thing is, the stats will be constantly updated over a period of time, I wouldn't think that one query per hour would be trouble, but I'm very open to ideas as far as what type of things toolserver access would attribute to this project. --lightdarkness ^(talk) 22:46, 17 February 2006 (UTC)

Isn't it easier for the report to be generated on the toolserver, with LDBot copying the results once per day and overwriting the local Wiki page? Or am I misunderstanding something? Titoxd^{(?!? - help us)} 01:28, 18 February 2006 (UTC)

That would pretty much do the same thing, I'm just using my personal server to do all of the data processing, but with toolserver access I could just get the counts and update there, but I'm not sure the toolserver has access to a mysql database to write to (which I'm now using on mine for storage), but I'll look into it. --lightdarkness ^(talk) 03:12, 18 February 2006 (UTC)

The database in the toolserver does not have write access to the Wikimedia databases, so I don't think that is possible. However, the same effect can be done by making the raw query in the toolserver (which is much less expensive than querying Kate's Tool or doing anything like that), and then have LDBot read the file and process it slightly for posting in the local Wikipedia page. Titoxd^{(?!? - help us)} 05:50, 19 February 2006 (UTC)

I knew that the toolserver can't write to the wikipedia database, but the way I'm storing all the edit counts is on my own database, once I query Interiots tool for them. This way, I can space them out rather than gathering all the data at runtime. Even if it were on the toolserver, being run all at once would probably take quite a bit of processing power, and I wouldn't want to disrupt the toolserver. --lightdarkness ^(talk) 05:55, 19 February 2006 (UTC)

After re-reading all of the comments here, I thought I would just sum everything up. Interiot stated that 2-3 hits per hour wouldn't affect the toolserver, but suggests that there could be a better way of going about it. However, the current layout of WP:NA has a lot of information, regarding notes about users seeking adminship, links to previous noms, ect, that the toolserver cannot gather by doing a simple query. Which brings me to my conclusion. The concensus on the talk page of WP:NA suggests everyone is for the bot implimentation. Due to the limitations of the toolserver for this layout, I'll just be querying the toolserver once every hour for an updated editcount of the user whose editcount was updated the longest ago. Every night, my bot will grab commands from a specialized page in LDBot's usernamespace for commands to add, modify, delete users from the master list, and then LDBot will blank the page, signaling that the commands have been processed. Then, if any updates are required, LDBot will update the pages of the appropriate sections. I hope this wasn't too long, but this is the best way I can see doing it while automating the process and keeping the output the same. --lightdarkness ^(talk) 21:48, 19 February 2006 (UTC)

It would be simple enough to generate a report on the toolserver in a machine-readable format to provide this data; this could be kept up to date via a cron job; Interiot hinted above, I think, that he'd be willing to help with this. The thought process I'm in here is that this method is going to produce outdated information.

Nevertheless, approved for a trial run. Rob Church 01:12, 30 March 2006 (UTC)

Thanks Rob, I appreciate it. The bot is going to start it's trial run Friday evening. I'd also like to request one more thing from here. It's been noted that AFDBot hasn't been operational since mid-March, and AllyUnion hasn't been around to fix it. I've offered to write another feature for LDBot to update the AFD pages as appropriate. I've put a note about it on WT:AFD, and am waiting to see if there aren't any objections. If after 5 - 7 days there aren't any objections on that page, would anyone have a problem with LDBot taking over this task? At least until AllyUnion gets back? --lightdarkness ^(talk) 13:33, 6 April 2006 (UTC)

No objections, it would be great if you could take over that task. Martin 14:01, 6 April 2006 (UTC)

• Please note that the WP:NA functions of the bot have been discontinued, as the toolserver isn't getting updates from en_wiki. Have applied for a bot flag for the purpose of AFD. --lightdarkness ^(talk) 03:18, 13 April 2006 (UTC)

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

JoeBot

Hello everybody! I would like your permission to run my little JoeBot. It replaces common misspelled words, like 'begining' (beginning). I sort of started doing it on my own account and then realized that if I wanted to continue doing it I would need a bot account because of the sheer massive amount of these misspellings. So here I am! Wikipedia is getting so large that little mistakes fall through the cracks, and I'd like JoeBot to be able to be there to catch em'.

The bot will be manually run by yours truly, using the mighty fine AutoWikiBrowser by Martin. I'll be using it with the JoeBot account for a trial week to earn my spurs. It will only check the english wikipedia. JoeSmack ^Talk 18:39, 19 February 2006 (UTC)

And yes, I am truly aware of regional spellings and pay special attention not to replace these. (example: 'color' and 'colour') JoeSmack ^Talk

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

Bot flag for SmackBot

Can I have a bot-flag for User:SmackBot to allow me to edit very quickly without cluttering "recent changes". Previously (for example when I put GFDL templates on 3000+ map images) marking the changes as minor has been the thing to do, but times have changed I guess. Plus if anyone can recommend a simple bot apart from pywikipediabot for me to look at I would be grateful. Rich Farmbrough. 13:04, 21 February 2006 (UTC)

What will SmackBot do with its very quick editing powers (why do you need it)? Is it important enough for Wikipedia to allow your bot? How long will it run for? Is it using pywikipediabot or some other? Will it be manually assisted or automatic/scheduled?

Another simple bot apart from pywikipediabot is AutoWikiBrowser. I like it a lot. JoeSmack ^Talk 14:38, 21 February 2006 (UTC)

Well I was pluralising the "External links" header in articles with more than one external link. I have paused (or slowed to a crawl), because a user found it inconvenient when doing recent change patrol. There are currently about 7-8000 articles still to do. I would only use SmackBot account for manual edits, until further request (I.E. I am not requesting to run a bot at the moment). I would also restrict it to very simple edits, because they are the only ones that can be performed at high speed, the slightest complexity means careful checking is required, therefore recent changes is not cluttered, therefore no bot flag is needed and I would use my main account. I have used AWB it is very good, but in CSharp . Rich Farmbrough. 16:43, 21 February 2006 (UTC)

I am using AWB now, just to update. Rich Farmbrough

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

Since the above discussion about SmackBot has been concluded, I created a new section below --Francis Schonken 10:47, 26 March 2006 (UTC)

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

Approved, Martin 09:06, 11 April 2006 (UTC)

Bot flag for DFBot

DFBot has been maintaining the RFA summary and the AFD summary for a number of weeks now, but I have been too lazy to come back here and ask for a bot flag. Are there any objections to getting one? Dragons flight 17:07, 21 February 2006 (UTC)

I've seen no complaints. Bot flag approved. Rob Church 01:12, 30 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

No consensus. Martin 09:14, 11 April 2006 (UTC)

Bobblebot

I would like to use a bot to delink date links.

The task is to reduce linking of solitary months, solitary years, etc in accordance with the Manual of style. It will not delink date links used for date preference. Each edit will be checked visually.

For example Economy of Algeria has many solitary year links, including 14 to 2004. The policy and popular misunderstandings are explained at:

• Wikipedia:Manual_of_Style_(dates_and_numbers)#Date_formatting
• Wikipedia:Manual_of_Style_(links)#Dates_and_numbers
• Wikipedia:Make_only_links_relevant_to_the_context#What_should_not_be_linked

For example:

• The short-lived ABC Cable News began in 1995; unable to compete with CNN, it shut down in 1997. Undaunted, in 2004 ABC launched a news channel called ABC News Now.

will become:

• The short-lived ABC Cable News began in 1995; unable to compete with CNN, it shut down in 1997. Undaunted, in 2004 ABC launched a news channel called ABC News Now.

See this diff working towards that: [2].

Several editors were using Martin's excellent AutoWikiBrowser to do this task. That was a good solution that just got better and better. The shared experience of many users produced a 'whole is greater than the sum' process of logic improvements. I would be more than happy to share or give away the date delinking task to AutoWikiBrowser if it is permitted to do it.

A previous application on 15 December did not succeed. User:Sam Korn suggested that I should reapply and make clear that I look at each edit. So this is my re-application. Thanks. bobblewik 19:48, 25 February 2006 (UTC)

• Oppose, Bobblewik has said in the past that he checks each edit yet his talk page has plenty of examples of where he has clearly not taken enough care to check that his bot script has not done something silly. I don't believe this will be any different. I am also opposed in principle to using bots to "improve" articles. Articles are organic things and rigid scripts rarely help. Talrias (t | e | c) 19:54, 25 February 2006 (UTC)

  By the way, what has changed since the last application? Talrias (t | e | c) 20:01, 25 February 2006 (UTC)

  Basically, it was my perception (and Bobblewik has taken my advice, so please shoot me first :-) that most complaints were more due to concerns that the edits would not be monitored. Is there any reason why Bobblewik should not be allowed another chance? Sam Korn ^(smoddy) 20:44, 25 February 2006 (UTC)

• I have no problem remove 13 links to 2004 in the first example but do not agree that the links are irrelevant in the second example. Worse, I have seen no effort on your part to determine which links are relevant - only mass removal. I understand you do not like these links. I also understand that some other people do. I don't see why a bot is required in this controversy. Rmhermen 19:57, 25 February 2006 (UTC)
• So long as Bobblewik undertakes that he will look very carefully at complaints about mistakes and will examine edits carefully, I see no reason why this should not happen. Support Sam Korn ^(smoddy) 20:44, 25 February 2006 (UTC)
• I support a bot that makes these types of changes, as i did in Decemeber. i hate seeing over linked articles with respect to dates. David D. (Talk) 20:54, 25 February 2006 (UTC)
• Support. I can't see what the damage could be. No information could be lost. As it stands no one could go to a date to find information anyway. Caution should always be exercised in using bots. --Wetman 20:58, 25 February 2006 (UTC)
• Strong support. There are so many articles which overlink dates, I often fix these manually and comment about it in peer reviews, etc. it would be great to have a bot get rid of these overlinkings. --Fritz Saalfeld (Talk) 21:06, 25 February 2006 (UTC)
• Support, provided that the changes only delink dates and don't otherwise affect their formatting. (I'm particularly concerned about [[1941]]-[[1944|44]] becoming 1941-44, and not 1941-1944.) —Kirill Lokshin 21:09, 25 February 2006 (UTC)

• Support/Oppose. Bobblewik's edits are manually checked, therefore they are at most robot assisted. Since we are all fallible, there is no reason these edits should not show on recent changes so a bot-flag should not be granted. Bobblewick is restricting his changes to 2 per minute - slower than some of his critics. If giving him bot-status will allow him to continue editing, then I support his request. If on the other hand, he should be permitted to edit at two changes per minute (or a faster rate) in any case, then I would oppose the bot status, and say let him get on with the editing. If someone has a problem with any of his edits, they can discuss them or revert them individually, and follow the normal dispute resolution procedure if that doesn't work. Rich Farmbrough 21:17 25 February 2006 (UTC).
• Support, let the bobblebot role - get rid of needless date links. Vsmith 21:22, 25 February 2006 (UTC)
• Support. Even if it delinked every year in Wikipedia it would probably be doing more good than harm, considering how bad date overlinking is. Kaldari 21:35, 25 February 2006 (UTC)
• Support. My only concern is that there seems to be an army of "robots" who are just going to mindlessly continue to turn every year and date into a link. It seems like there needs to be a way to get the people who make all these useless links to stop doing so. --JWSchmidt 21:39, 25 February 2006 (UTC)

• Support. I do want to see those year links reduced. -- Donald Albury (Dalbury)^(Talk) 21:54, 25 February 2006 (UTC)
• Support. As long as we are agreed that overlinking of dates is a Bad Thing, i.e. that the MOS isn't in dispute, it makes sense to grant Bot status to any program that will bring the Wikipedia closer to the MOS. Consider the possible good vs. the worst possible harm, as Kaldari did, above. It's hard to imagine a scenario in which net harm would be done.—GraemeMcRae^talk 22:05, 25 February 2006 (UTC)
• Support - All Bobblewik is trying to do is conform to policy, which is a noble goal. We should be thankful we have someone who is willing to do this nitty gritty work; I certainly don't enjoy going through hundreds of pages and doing it manually. Even though he is going AWB-assisted, it is manual in the end, so kudos to him. As an answer to someone who said he shouldn't get the bot flag because these are manual: he's had people repeatedly blocking him because he's "editing too fast" for a non-bot. If he doesn't get this bot flag he's just going to face the same kind of misunderstanding and repeated disruptions of his work. --Cyde Weys 22:16, 25 February 2006 (UTC)
• Support, I've made my views on this clear before. Stephen Turner (Talk) 22:25, 25 February 2006 (UTC)
• Strong Support. As others have said, conforming to the Wikipedia policies is a good thing, and it helps us get this sort of maintenance/MoS work done quicker. — Wackymacs 22:37, 25 February 2006 (UTC)
• Support. – Quadell ^{(talk) (bounties)} 22:48, 25 February 2006 (UTC)
• Support. Unnecessary links are a nuisance and bobblewik wants to do something about them. He has taken enough flak for his efforts and deserves the support of the community. Stroika 22:57, 25 February 2006 (UTC)
• Support --Duk 23:08, 25 February 2006 (UTC)
• Oppose - Works on the nerves of many people (also this whole "crusade" idea, so comparable to the "conversion to BC/BCE" crusade, to the "conversion of references" crusade, etc - the proponents of these crusades all advocated application of guidelines as a justification of their crusade, all these were stopped by ArbCom leaving them on probation without permission for further conversion ***neither by bot nor manually***). This bot proposal would mean Bobblewik is "approved" to come in on any article (say, for instance, Furniture music), and decide which are the significant dates, and which aren't. Sorry, Bobblewik *may* be aware of which the significant dates are for several ranges of articles, but (s)he should not get a ticket to be allowed to meddle with any article (s)he pleases. Further, I take it as a principle that anyone who goes around inviting only parties who previously expressed sympathy to take part in a vote, should better not be granted anything. --Francis Schonken 23:47, 25 February 2006 (UTC)

• What are you talking about, only parties who previously expressed sympathy, Bobblewik contacted some of his harshest critics, including admins who have blocked him before. Go take another look. --Duk 00:16, 26 February 2006 (UTC)

• Comment. Bobblewik has just spammed over 50 user talk pages (with the edit summary "Date links") about this bot account application. I'm assuming Bobblewik was using AWB, because a rate exceeding one edit per 10 seconds was reached.--Commander Keane 23:26, 25 February 2006 (UTC)

  Bobblewiki is not presently enabled to use AWB, plus all edits made with it reference the AWB page in the summary. Martin 23:32, 25 February 2006 (UTC)

  I've done that rate with page moves and talk page notices. Look at successful admin candidates' edit rates immediately after their promotion. 10/minute is not difficult if all the edits are identical, and if you have a reasonable connection. Sam Korn ^(smoddy) 23:36, 25 February 2006 (UTC)

  OK, let's stop this vote, and start an RfC on Bobblewik, what were we waiting for all the time? --Francis Schonken 23:47, 25 February 2006 (UTC)

  An RFC on what, formatting pages to conform with existing policy? I don't think that'd get very far. --Cyde Weys 06:03, 26 February 2006 (UTC)

  OK, my apologies to Bobblewik for the AWB assumption. I didn't realise that edits could be done that fast manually, and I'm glad the the AWB coutermeasures are still enabled.--Commander Keane 23:51, 25 February 2006 (UTC)

  Comment. Spamming or not, can users kindly refrain from having revert wars over this message on my talk page? The orange new message box keeps flashing up on my screen and it's really annoying to find there's no actual new messages. - Randwicked Alex B 06:34, 26 February 2006 (UTC)

• Oppose per Wikipedia:Administrators' noticeboard/IncidentArchive73#Bobblewik and other experiences with user. --M@thwiz2020 01:53, 26 February 2006 (UTC)
• Strong support: trivial date links are a scourge on WP that have somehow crept in on the back of the mechanism for formatting full dates. I can't imagine why anyone would object. I guess the only circumstance in which a year date might be useful is for ancient dates/centuries, but even so, that's not a good enough reason to stop the bot. Tony 04:43, 26 February 2006 (UTC)
• Support. Fully automated is ok too - no text is ever lost. BTW, I think it is outrageous that an admin would block Bobblewik for 2 weeks(!) for making good, valid edits (to date links), following the Manual of Style, without revert warring, as was described on the Administrators' noticeboard link that Mathwiz2020 pointed out above. -R. S. Shaw 04:58, 26 February 2006 (UTC)
• Support, please get rid of the "sea of useless blue" as commented above. But the bot should have had a better name. Tempshill 06:01, 26 February 2006 (UTC)
• Oppose. The edits have no consensus, thus a bot is not appropriate. Ambi 06:22, 26 February 2006 (UTC)

  Are you sure there is no consensus? David D. (Talk) 00:52, 27 February 2006 (UTC)

• Oppose, because I'm not so sure that all the changes being made are positive. Specifically, I'm thinking of this diff; although I agree that some of the delinkings were good style, some of them removed links to years in which a number of events pertinent to the article occurred, and I can easily imagine a reader clicking on the links for useful context. This suggests to me that the delinking should be done more slowly and thoughtfully, not quickly and automatically. Robth^Talk 06:31, 26 February 2006 (UTC)

  Although I support this bot, Robth's example is a good one. It might be advisable for Bobblewik to concentrate on more recent dates, post-1800 or post-1900, say. Such date links are much less likely to be useful, and earlier ones may be useful. Of course, any cut-off point is arbitrary, and may lead to strange results if applied strictly, so I would make this a recommendation rather than a condition. Stephen Turner (Talk) 09:15, 26 February 2006 (UTC)

  Ditto per Stephen. good point (and example). --Quiddity 00:13, 11 March 2006 (UTC)

• Strongly Support any and every means to get rid of unnecessary date links and nearly all of them are unnecessary. This is just a method of implementing the current guidelines, as published. We should all be implementing the guidelines until and unless an agreed upon change is made to them. Not noticing changes made in the past or not agreeing with changes made in the past is no reason to stop implementation of the guidelines. Hmains 06:38, 26 February 2006 (UTC)
• Absolute Support. Bobblewik has proven an excellent and dedicated user who will be able to handle the task of getting rid of unnecessary links after identifying them. Antonio Perrito Martin 09:34, 26 February 206 (UTC)
• Support - removing the "sea of blue" is certainly helpful, and, as far as I am aware, in line with the MoS and consensus more generally. Why should every year in every article be linked? It helps no one. Yes, he makes mistakes; but, in my experience, his improving edits massively outnumber any unhelpful ones. Surely it is better that the worthless links are stripped out, and then more targetted ones added back?-- ALoan (Talk) 09:57, 26 February 2006 (UTC)
• Strong support. Almost all such links are pointless and link-cluttering. I have faith in Bobblewik to identify the few exceptions. He is only working for a better Wikipedia. Neonumbers 10:09, 26 February 2006 (UTC)
• Strong support The overlinking is pointless, and that it goes uncorrected perpetuates it, I know I started linking dates for no reason other than I saw it was being done, and assumed it was wiki policy, it was only in seeing bobble's edits that I then bothered to look at the actual MoS to see that I was wrong. Correcting the overlinking will help unclutter pages and make new users aware of what the MoS actually says regarding wikifying dates. Gheorghe Zamfir 10:50, 26 February 2006 (UTC)
• Neutral. Separating bobblewik's "bot edits" from his "real edits" will be a step in the right direction, though I disagree with the principal of many of the types of edits he is interested in performing, only one of which has been mentioned in this discussion. — Feb. 26, '06 [13:06] <freakofnur_xture|talk>
• Support. — Matt Crypto 13:09, 26 February 2006 (UTC)
• Support. older ≠ wiser 13:45, 26 February 2006 (UTC)
• Oppose. While I support the contention that multiple data links in an article are unneccessary, unless very widely separated, date links per se provide a useful mechanism for establishing historical context for events. I resist the 'automatic' de-linking of dates, and contend that this should only be done in the overall context of reviewing an article, and not on a 'fly-by' basis. Noisy | Talk 14:50, 26 February 2006 (UTC)
• Weak opppose. I sometimes use links to years and dates to search for events that happened on a particular date. I also like years being linked, since that visually differentiates them from other numbers. All of this based on my personal tastes, hence weak oppose. Zocky | picture popups 18:41, 26 February 2006 (UTC)
• Oppose. A solution in search of a problem, and Bobblewik's demonstrated carelessness when supposedly doing this sort of thing manually is likely to be magnified if done automatically. --Calton | Talk 20:58, 26 February 2006 (UTC)
• Strong support. KillerChihuahua^?!? 23:51, 26 February 2006 (UTC)
• Support so long as the edits are well-checked, and if a bot-flag is needed (see Rich Farmbrough's vote above). ··gracefool |☺ 00:11, 27 February 2006 (UTC)
• Neutral The idea of the bot to assist in making wikipedia more consistent is worthwhile. However, there doesn't seem to be consensus on what the appropriate level of date-linking should be. I'd like to see that sorted out, first. -- Ch'marr 00:15, 27 February 2006 (UTC)
• Support. I fully support the cutting back of date links (assuming it is done with care, and an understanding of the Article involved). I believe a date should be linked only if it enhances my knowledge of the substance of the Article I am researching. Michael David 00:55, 27 February 2006 (UTC)
• Support –Joke 00:58, 27 February 2006 (UTC)
• Support - links that are not useful to someone reading the article should be removed. Most date links fall under this category. So long as it is monitored by humans to catch the exceptions to this rule then I am firmly behind this attempt to remove unconstructive link clutter. --HappyDog 01:13, 27 February 2006 (UTC)
• Strong Support Excessive blue linking is one of two linking scourges on Wikipedia today. VirtualSteve 03:12, 27 February 2006 (UTC)
• Oppose Most date links are excessive; but, without a clear concensus on what must be monitored (ie, are historically significant years ok to link — and how does Bobblewik propose to sift those from his bot edits), allowing this to go at a fast rate is going to end up in several reverted pages which should be avoided. Neier 11:23, 27 February 2006 (UTC)
• Support a bot to bring articles closer in line with recommended style but I note Rich Farmbrough's discussion above.Thincat 14:01, 27 February 2006 (UTC)
• Strong support. However, the code should be reviewed and tested against all cases to ensure a minimal amount of errors. Gflores ^Talk 14:51, 27 February 2006 (UTC)
• Oppose. Example diff given above by Robth shows why these chamges should not be made by an automated or semi-automated process. android79 15:19, 27 February 2006 (UTC)
• Support, with the proviso that Bobblewick should be careful to look for years that may be of interest to readers of the article, and leave them linked. As others have suggested, ancient dates and the first appearance of a year in an article are more likely to be useful to the reader. Otherwise, year links are a plague and I welcome an effort to remove them.--Srleffler 13:04, 28 February 2006 (UTC)
• Support. Dates are massively overlinked and distract from valuable links. Drastic action is needed. Haukur 16:05, 28 February 2006 (UTC)
• Oppose why remove information that does no harm? If someone took the time to manually link stuff they probably had a reason. -Ravedave 19:05, 28 February 2006 (UTC)
• Weak support - good idea to get all pages into general conformity, but having some links gives context to many articles: I found it very useful for early history, and when The Guardian of 24 October 2005 had a panel of experts review articles, one positive point was "And when I click on click on the year 1922, I get a page telling me what else happened that year. [TS] Eliot is at the centre of a whole web of other references." So please try to keep enough linked years to give context. ...dave souza, talk 19:23, 28 February 2006 (UTC)
• Support, so long as all edits are monitored by a human and not ALL date links are removed (i.e.- typically leave one link per year per page, and ones that are actually significant). EWS23 | (Leave me a message!) 04:34, 1 March 2006 (UTC)
• Support - all in favour of anything to improve consistency of formatting in articles. --Ali@gwc.org.uk 12:44, 1 March 2006 (UTC)
• Oppose - excessive year linking looks ghastly, but frequently the context of an article means a year should stay linked. I feel uncomfortable letting a bot loose on what will be very high volume changes. -- Ian ≡ talk 07:48, 3 March 2006 (UTC)
• Oppose - links should not be repeated but I don't believe anyone would agree that no year should ever be linked to. It appears the aim of this bot is to remove all year links without discrimination, including the first occurence in an article, which I think would be severely to the detriment of many articles. Worldtraveller 13:08, 3 March 2006 (UTC)
• I think it's a terrible idea to delink years automatically, and besides, Bobblewik spammed me. anthony 15:59, 4 March 2006 (UTC)
• Strong support: trivial date links are a scourge on WP that have somehow crept in on the back of the mechanism for formatting full dates. I can't imagine why anyone would object.Jclerman 13:20, 5 March 2006 (UTC)
• Support, plan is to operate in accordance with guidelines; if guidelines change then bot can be adjusted.--A Y Arktos 19:18, 5 March 2006 (UTC)
• Strong Support - He even states "Each edit will be checked visually." invalidating half the oppose votes rationales. --Quiddity 22:47, 5 March 2006 (UTC)

  I didn't want to get into point/counterpoint here, but I don't think it is fair to say that people's voting rationale is "invalid". Bobblewik's made this claim from the word go, yet even a cursory look at his talk page shows numerous examples of where he hasn't bothered to check his edits, leaving his mistakes in place - and this isn't just the case with his date modifications. Astonishingly he's even left mistakes in a section barely a few sentences long. He's made these promises in the past but he's demonstrably not keeping them. Talrias (t | e | c) 23:39, 5 March 2006 (UTC)

  Is permission even needed if you are going to check every edit manually? Can't that be done using your regular user account?

• Comment - when and how should this poll be closed? – Quadell ^{(talk) (bounties)} 14:51, 6 March 2006 (UTC)

I would like to know too. bobblewik 15:34, 8 March 2006 (UTC)

I think one legitimate oppose is all that should be required to deny a bot operating approval. We got that at 19:54, 25 February 2006 (UTC). I suggest we close the comments now.--Commander Keane 15:40, 8 March 2006 (UTC)

• Oppose. While I am not opposed to Mr. Wik himself, I am really tired of all the controversy surrounding the stupid bot antics to begin with. ... aa:talk 02:09, 9 March 2006 (UTC)

• Oppose this user is reckless in his delinking of dates. Jooler 01:53, 10 March 2006 (UTC) - example TWICE on the same article within the space of two weeks! - [3] and [4] Jooler

• Oppose - Although some date delinking done by Bobblebot is beneficial, some edits, such as to Abu Ghraib prison (diff: [5] (removing links to dates significant as part of the US invasion of Iraq) clearly detracts from the completeness of the encyclopedia. I will support this bot if more careful supervision if enforced. - Tangotango 06:50, 10 March 2006 (UTC)

Application withdrawn (see section below). bobblewik 07:03, 10 March 2006 (UTC)

Depending on the how I counted the votes and the close date, I counted:

• Support: 70 to 80%
• Oppose: 20 to 30%
• Neutral: ~5%

As the proposer, I have a vested interest. If somebody independent wants to count, that would be welcome. bobblewik 16:56, 13 March 2006 (UTC)

Wikipedia on the whole operates by consensus, not by voting. However, the bot policy is that if there is a significant objection, then the bot request is denied - because bots are potentially damaging and we want all bases covered. Talrias (t | e | c) 17:00, 13 March 2006 (UTC)

Thanks for the reply. The count has no hidden message. It was not a request for information about policy, or a challenge to my own decision to withdraw. Some people are interested in the count. I agree with your point that consensus and counts are two separate things. bobblewik 18:24, 13 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

WatchlistBot

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

Mulder416sBot

I would like to request to run a bot (AWB) to simplify tasks that I come across that are way too repetitive, or would otherwise take foreever. Mainly mass template substitution. -Mulder416 00:52, 28 February 2006 (UTC)

Bot status is granted --Walter 20:20, 24 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

Approved. Martin 09:17, 11 April 2006 (UTC)

User:Slobot

I'd like to request permission for running the pywikipedia bot Slobot. It will not do much edits here. It will mainly work on nds.wikipedia and only add interwiki links to en. --::Slomox:: >< 13:46, 2 March 2006 (UTC)

Approved for a trial run for one week. Rob Church 01:13, 30 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

Approved. Martin 09:19, 11 April 2006 (UTC)

FairuseBot

This will be a bot for maintainence tasks relating to the use and abuse of fair-use images. The initial task will be removing any image in Category:Disputed fair use images that's been tagged for a few days from articles it's in. Future tasks may include going through Category:Fair use images and tagging images with no source information as {{nosource}}, and other tasks as determined by Wikipedia:WikiProject fair use and Wikipedia:Fair use review.

I'll be using OrphanBot's code as the basis for this bot's code.

The reason for using a separate bot account, rather than having OrphanBot do the editing, is that working with fair-use-tagged images is more contentous than no-source or no-license images, and I want to keep OrphanBot's activities separate from those of FairuseBot. --Carnildo 07:07, 3 March 2006 (UTC)

Approved for the week trial run with one caveat; let me know how it goes. ;-) Rob Church 01:14, 30 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

Approved. Martin 09:08, 11 April 2006 (UTC)

Tawkerbot2

Tawkerbot2 is a fork of the anti-vandalism aspects of Tawkerbot as it was suggested that vandal fighting be removed from Tawkerbot (1)'s duties. Tawker 10:26, 3 March 2006 (UTC)

The bot runs various anti-vandalism checks on suspicious pages flagged by a set of algorithms that has been closely QA'd to ensure there are no false positives (the release of the algorithms will be considered upon request). It will run without a bot flag so any edit it can make can be double checked. Tawker 10:50, 5 March 2006 (UTC)

I imagine checking even a small fraction of recent changes could be load intensive. Where does it get the data to check the edit? And I assume it will remain indefinitely user-assisted (apart for sqidward), since a bot will always make some mistakes when looking for vandalism?--Commander Keane 11:52, 5 March 2006 (UTC)

Source data comes from pgkbot the IRC bot used by the CVU and once it grabs an alert from there, it runs an extensive check on the page (simply being flagged by pgkbot will NOT result in an automatic revert, it only touches article space edits and it will not revert twice. In addition, any chanop on #wikipedia-en-vandalism as well as any sysop can kill the bot -- Tawker 12:52, 5 March 2006 (UTC)

This bot appears to be doing a trial run (approved?). Out of a current 7 edits, the 5th was a mistake - the bot reverted a good edit[6]. --Commander Keane 12:40, 5 March 2006 (UTC)

Noted, and fixed -- Tawker 12:52, 5 March 2006 (UTC)

The bot has multiple shutdowns and we're working on a couple more to make it even more safe.

Sample output (what it shows on console) is here - if anyone wants to monitor the bot please send me a message.

We just rewrote the bot and tested it for a while but somehow a codechange was commited when the bot was in live mode (over debug mode where it simply reports what it would do), I've locked out the live codebase from the development version so that better not happen again, I stopped the bot as soon as it happened and I fixed all of it's "damage" -- 02:15, 6 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

Zirland = PorthosBot

I am the admin of Czech Wikipedia and my main activity here on English is doing interwiki links. I run the multi-purpose robot on cs.wiki. I'd like to enable the robot doing interwiki to czech articles.

So I am asking you for support to granting the bot status to my user account here on en.wiki. -- cs:User:Zirland Zirland 11:24, 3 March 2006 (UTC)

I changed my mind and I'll keep my account non-bot. So I registered special account for the bot - User:PorthosBot. If there is no objection, I will ask for the tag on Meta tomorrow. --Zirland 15:37, 16 March 2006 (UTC)

1. You haven't read our policies well enough, it seems
2. "Tomorrow" was too soon

   Rob Church 01:15, 30 March 2006 (UTC)

Per wikipedia talk:bots#PorthosBot (and other talk linked from there) I think it is clear:

1. PorthosBot should immediately be de-flagged at en:wikipedia,
2. ... so that this bot approval request can be assessed on its own merits.

For my own opinion (in the "assessment of bot merits" meaning):

• Probably the interwiki-linking jobs of the bot are by themselves not problematic, but the bot operator should address issues, e.g. the issue mentioned at user talk:PorthosBot#Incorrect addition of interlanguage links;
• I hope communication in English isn't a problem... as mentioned by Rob above Zirland apparently didn't yet really understand implementation of current bot policy at WP:BOTS#Current policy on running bots.

--Francis Schonken 11:59, 9 April 2006 (UTC)

The robot is pyWikipediaBot, on English Wikipedia adds links to (primarily) czech articles. Due to pyWikibot standard functionality it can however add/change other links. This bot is manually operated. --Zirland 06:41, 14 April 2006 (UTC)

Gnome (Bot) I reproposed with a new mission below

The code for this bot is nearing completion, it is capible of several things, what I will use it for is still up in the air. (when I do deciede I will ask for bot status)

As a result of my code nearing completion I would like to test it, its editing capabilities.

1)Can it write correctly to wikipedia

2)Can it write to the right fields.

3)Can it not make a mess while doing so.

Is a sandbox alright, (on my Userpage)? What do I need to do? Thanks Eagle (talk) (desk) 02:30, 4 March 2006 (UTC)

NOTE: I do not want a one week trial period, if that can be avoided... I just want to double-check my coding.Eagle (talk) (desk) 02:32, 4 March 2006 (UTC)

What I did with OrphanBot while I was developing it was have it make one edit, checked the result, reverted any mistakes, and fixed the bugs I found. As long as the bot isn't editing too fast, isn't messing anything up, and you're checking every edit, short testing runs are fine. --Carnildo 04:10, 4 March 2006 (UTC)

Hey thanks, no trial period is required...or notification? I really don't want to get people mad at me right away:-)Eagle (talk) (desk) 04:19, 4 March 2006 (UTC)

User:Heikobot

I'd like to request permission for running the pywikipedia bot Heikobot. It will use up-to-date pywikipediabot software and in en.wikipedia.org it will only add/correct interwiki links. Heiko Evermann 22:03, 4 March 2006 (UTC)

Between which projects? Rob Church 01:15, 30 March 2006 (UTC)

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

Pathosbot

I've created Pathosbot on the Pywikipedia framework to automate tedious or slow tasks, such as tagging blocked proxies, template substitution, recategorisation, template conversion, cleaning up template code, et cetera. Since the bot is intended for various uses including requests, there is no specific limit to the amount of time it's expected to be used. It's manually activated and runs under my supervision. The bot is actively being run on Wikisource as Pathosbot, where it has met with approval thus far. // Pathoschild (_admin / ^talk) 09:39, 6 March 2006 (UTC)

Since there's no opposition so far, I'll begin a trial run. The bot will be working on User:Pathoschild/Projects/Template substitution for now. // Pathoschild (_admin / ^talk) 22:54, 13 March 2006 (UTC)

For the sake of the redtape lovers, trial run endorsed. Rob Church 01:41, 30 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

User:AIDbot

Situation on WP:AID (imense increase in number of nominations in just few weeks) forced me to write this bot as maintaining this project became imposible by humans. I wrote it without consulting this page properly, so I started runing it without putting a notice here. But, now I read this page and now I'm putting te notice here :-) Please, don't block this bot for not requesting it's approval first as you can see that nobody complained about it, and moreover, it recieved only compliments. I'm sure we can settle any disputes about this bot (if any souch dispute should arise, I don't see why would that happen) without stoping it, as it is essential in trying to keep AID up to date. --Dijxtra 09:37, 7 March 2006 (UTC)

• The bot puts templates on user talk pages AIDbot's user contributions:

1. 10:56, 6 March 2006 (hist) (diff) m User talk:Lukobe (rollover) (top)
2. 10:56, 6 March 2006 (hist) (diff) m User talk:DanielCD (rollover)
3. 10:56, 6 March 2006 (hist) (diff) m User talk:Raghu.kuttan (rollover) (top)
4. 10:56, 6 March 2006 (hist) (diff) m User talk:Hahaandy1 (rollover) (top)
5. 10:56, 6 March 2006 (hist) (diff) m User talk:SpacemanAfrica (rollover) (top)
6. 10:55, 6 March 2006 (hist) (diff) m User talk:Vir (rollover) (top)
7. 10:55, 6 March 2006 (hist) (diff) m User talk:Pschemp (rollover)
8. 10:52, 6 March 2006 (hist) (diff) m User talk:Ugur Basak (rollover)
9. 10:52, 6 March 2006 (hist) (diff) m User talk:Jmabel (rollover)
10. 10:52, 6 March 2006 (hist) (diff) m User talk:Radufan (rollover) (top)
11. 10:49, 6 March 2006 (hist) (diff) m User talk:Wikiacc (rollover)
12. 10:49, 6 March 2006 (hist) (diff) m User talk:172 (rollover) (top)
13. 10:48, 6 March 2006 (hist) (diff) m User talk:Adammathias (rollover) (top)
14. 10:48, 6 March 2006 (hist) (diff) m User talk:Resistor (rollover) (top)
15. 10:48, 6 March 2006 (hist) (diff) m User talk:Bhadani (rollover)
16. 10:48, 6 March 2006 (hist) (diff) m User talk:Ashibaka (rollover)
17. 10:48, 6 March 2006 (hist) (diff) m User talk:Silence (rollover)
18. 10:47, 6 March 2006 (hist) (diff) m User talk:Quadell (rollover)
19. 10:47, 6 March 2006 (hist) (diff) m User talk:Carwil (rollover) (top)
20. 10:47, 6 March 2006 (hist) (diff) m User talk:Tombseye (rollover)
21. 10:47, 6 March 2006 (hist) (diff) m User talk:NeoJustin (rollover) (top)
22. 10:47, 6 March 2006 (hist) (diff) m User talk:TachyonP (rollover) (top)
23. 10:47, 6 March 2006 (hist) (diff) m User talk:Khoikhoi (rollover)
24. 10:46, 6 March 2006 (hist) (diff) m User talk:Stevecov (rollover) (top)
25. 10:46, 6 March 2006 (hist) (diff) m User talk:Neutrality (rollover) (top)
26. 10:46, 6 March 2006 (hist) (diff) m User talk:Cuivienen (rollover)
27. 10:46, 6 March 2006 (hist) (diff) m User talk:Wackymacs (rollover) (top)
28. 10:45, 6 March 2006 (hist) (diff) m User talk:Waltwe (rollover) (top)
29. 10:45, 6 March 2006 (hist) (diff) m User talk:Fenice (rollover)

• The templates thus posted are not in template namespace (e.g.: [7]), so: unnecessary overhead
  • Care to explain in more precise manner? What is the actual problem here? --Dijxtra 13:27, 7 March 2006 (UTC)
    • Suppose your project has an average of 100 contributors soon; A year later the wikipedia databases contain approx. 5000 times this string:
      ==[[<pagename>]] won!== <div style="text-align: center; margin: 0 10%;"> {| class="notice noprint" id="{{{id}}}" style="background: #ffccFF; border: 1px solid #ff33FF; margin: 0 auto;" |- | [[Image:Aidlogo.png|none|50px| ]] | Thank you for your support of the '''[[Wikipedia:Article Improvement Drive|Article Improvement Drive]]'''.<br>This week '''[[<pagename>]]''' was selected to be improved to [[Wikipedia:Featured articles|featured article]] [[Wikipedia:What is a featured article|status]].<br>Hope you can help… |} </div>
      — i.e. approx. 5000 times over 200 characters (didn't count); it's possible to do the same requiring less than 10% of these resources, by typing the above in a template once (replacing <pagename> by {{{1}}}), so the database only has to digest 5000 times a string of this length: {{AIDnotif|<pagename>}}.
      Note that even then I don't think this a good idea. Afaik, all projects sending out invitations on user talk pages by bot stopped doing that (the last project I knew in this sense was Esperanza). --Francis Schonken 17:25, 7 March 2006 (UTC)
      • Good point! Didn't remember of that. If we decide that it's OK to leave messages on talk pages, I'll make the new template and use it. And the decission whether to leave messages on talk pages will be made somewhere else. I myself don't plan on entering that discussion, I'm interested just in the result of the discussion, so I know how to program my bot. I'll inform the people on AID project to discuss that. For now, I'm dropping all of the features of the bot that are disputable, so we can use the features which are not disputable. --Dijxtra 18:41, 7 March 2006 (UTC)
• The bot adds non-content (or "project-related") templates to articles http://en.wikipedia.org/w/index.php?title=Roma_people&diff=prev&oldid=42461727 — whether or not such templates are desired in article namespace, should not be decided by a vote on the bot performing these changes: I mean: it is questionable whether such templates are called for in article namespace: I'm merely saying the discussion about these does not have to take place here, on the talk page of the "bots" page. Maybe there was some prior discussion (and approval) of the use of the {{AIDcur}} template, if that is the case: please give a link where such discussion was concluded.

  (PS by Francis Schonken 12:09, 7 March 2006 (UTC):) until such approval is a fact, I suppose:

  • Such templates should only occur in talk namespace, like templates of similar initiatives, see Wikipedia:Template messages/Talk namespace
  • The model of the template should better conform to the generic template shown at Wikipedia:Template messages/Talk namespace#WikiProject generic notice, i.e. {{WikiProjectNotice}}

  • The bot was formed on rules which exist on AID (you can see those here). This bot just does what was previously done by hand. Therefore, this is not the place to discuss such policies. If you do not approve this bot, the templates will be added by hand, as they have been for last few months. This bot does not impose new rules, it just automates the old ones. --Dijxtra 13:27, 7 March 2006 (UTC)
    • "rules which exist on AID" — but apparently as rules for a small-scale project. You indicate yourself the project has grown quite a bit. Apart from the question whether the involved tasks need to be automated, the question whether it is a good idea to have AID as a large scale project needs to be answered too. Not on this page, as we both agree. But apparently the question wasn't answered elsewhere either (as you don't give a link to such discussion with conclusive decision - note that I didn't ask you to give a link to the page with the rules, but a link to the place(s) where it was decided that there is broad community approval of the mode of operation of the project). I don't see the need to give permission for bot automatisation as long as the other question isn't answered
      Apart from that, I don't think it a good idea to have templates like {{AIDcur}} in article namespace, per wikipedia:avoid self-references. And even less when a bot would be given permission to place such templates in article namespace. --Francis Schonken 17:25, 7 March 2006 (UTC)
      • As we agreed, this is not the place to discuss that, so I'll do the following: drop that feature untill the thing is discussed somewhere else. --Dijxtra 18:41, 7 March 2006 (UTC)
• The AID project is only one (of many) projects on article improvement: I don't say this project is necessarily worse or better than the other projects/initiatives, but maybe it shouldn't have a competitive advantage by being served by a bot (yet). The discussion about the various article improvement initiatives should take place elsewhere, not here on the bot talk page, IMHO. Was there any prior discussion (e.g.) in village pump? announced at RfC? etc... --Francis Schonken 11:24, 7 March 2006 (UTC)
  • This project has currently 73 nominations. Every nomination has to be checked at least once a day (preferably every 6 hours). To check the nomination you have to count the votes, detect the anonymous ones, remove expired nominations and update header information on nominations. This is a tremendous amount of work, tremendous enough to make me learn python and learn how to write a bot withouth any how-tos. I insist that this project will colapse under it's own weight if this bot is suspended as the number of people willing to check the nominations is very small and shrinking. --Dijxtra 13:27, 7 March 2006 (UTC)
    • I made no objections to that part of the bot operation, did I?
      • Sorry, I misinterpreted you. I felt that "it shouldn't have a competitive advantage by being served by a bot" was an objection. If it isn't, then everything's cool! --Dijxtra 18:41, 7 March 2006 (UTC)
    • I objected to the serialized talk page posts and to the namespace where the bot places the AIDcur template. I didn't see you would be prepared to modify your bot on these two points (well, are you?). I think it a bit curious I had to find out about these operations myself (you didn't mention them on the bot's talk page, apart from using the word "rollover", which I didn't know implied all that), so for the time being I'd disable these two operations of the bot (that is: the posting of invitations on user talk pages, and the posting of a project template in article namespace instead of in talk namespace). Unless these objections can be taken into account (or remedied by proof of wide community acceptance), separately from the project page check & update functionality (which I don't object to), I'm opposed to the bot as a whole. --Francis Schonken 17:25, 7 March 2006 (UTC)
      • Objection taken in account, the two disputed features dropped utill consensus on those is reached. The bot will for the time being do just operations like these: [8] [9] [10] Is this acceptable? --Dijxtra 18:41, 7 March 2006 (UTC)

Fetofsbot2

This will be the bot that I'll use for general fixes, such as substing and disambiguating. All edits by this bot will be manually checked. Fetofsbot2 22:57, 12 March 2006 (UTC)

Who is running you? Fetofs (talk · contribs), I imagine. You shouldn't edit with your bot - edit under the operator name for comments such as these. :) Talrias (t | e | c) 23:07, 12 March 2006 (UTC)

I forgot to log in as I had just created my account. Sorry! Fetofs ^Hello! 23:11, 12 March 2006 (UTC)

Disambiguation is fine, but substituting what? Rob Church (talk) 22:15, 31 March 2006 (UTC)

IW-Bot-as

I'd like to run bot User:IW-Bot-as, which will place interwiki links (mostly to lithuanian wiki). I would like to request bot-status for this bot. I use interwiki.py which is periodically updated from CVS. --Laurinkus 17:50, 27 February 2006 (UTC)

No problems, approved. Rob Church 01:39, 30 March 2006 (UTC)

User:ShinmaBot - The Wikipedia:Articles for creation maintainance bot

For a while, User:Uncle G's 'bot has done archiving maintainance on Wikipedia:Articles for creation. However, recently, his bot has stopped functioning. I'm currently working on a bot that will take over these activities as well as add new activities for helping maintain AFC. Specifically:

Auto-archiving of requests.

• Actions: Moves the current day's request to the appropriate archive, starts a fresh request page, and updates the archive listings.
• Frequency: Once per day.

Stub/category cleanup.

• Actions: Looks for stub entries and categories added by users in their requests and converts them to simple links {{tl}} and {{cl}} links. This is at the request from the stub/category sorting communities.
• Frequency: 3-6 times per day.

Empty request cleanup.

• Actions: Looks for dead/null requests (that is, requests that consist of nothing but the request template with no added text) and removes them.
• Frequency: 3-6 times per day.

At first, the bot will be manually started until it is determined that all bot functions are working as expected. At that point, the bot will run on as a cron job from a Linux server. I'm currently writing the bot in Java (not using python framework). -- ShinmaWa^(talk) 19:33, 13 March 2006 (UTC)

Mmm, I anticipate no problems. Trial run approved. Rob Church 18:51, 30 March 2006 (UTC)

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

User:CmdrObot

I've been running a manually assisted bot on the English wikipedia for the last couple of months under my normal user account. It does spelling and punctuation fixes; HTML entity conversion; URL tidying and other miscellanea (I used it to bulk convert a bunch of TinyURLs recently, for example). I manually inspect every edit before submitting it.

It's been brought to my attention that I should really be doing this from an account with the bot flag enabled, so I've created a new account for this purpose. The bot has been fairly heavily tested at this point (the best part of 30,000 edits) and should be fairly safe by now. I'd like to get bot-status for this account if I may. Cheers, Cmdrjameson 03:42, 14 March 2006 (UTC)

No reponds. Is this bot approved by default or not? --Walter 23:44, 22 March 2006 (UTC) (steward)

By old guidelines (see intro above), a bot couldn't be approved - it could only be rejected by the community. This bot probably fall into this category, and you can see no bots before it got any votes, including mine. If any user had seen this they would probably care to comment. Fetofs ^Hello! 00:13, 23 March 2006 (UTC)

• if it is still open for votes I support this bot. ILovEPlankton 00:49, 24 March 2006 (UTC)

• Support Pegasus1138^{Talk | Contribs | Email} ---- 01:29, 24 March 2006 (UTC)

Bot flag approved. Rob Church 01:21, 30 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

User:Terminatorius - automated blanking of the vandal anonymous IP talk pages

The problem

It is usually good to give for a vandal a warning, adding one of the standard warning templates to the talk page (vandals then frequently stop they activity without blocking). However if the vandal uses the anonymous IP, another user (not vandal) may later see that warning, being unpleasantly surprised. In Lithuanian Wikipedia there was a discussion that the person who left the warning to the anonymous user should delete it after several days, but remembering IP's and removing warnings manually takes too much time. Also, the page with 20 previous warnings and the new warning at the bottom may be less impressive to a vandal. From the troll concept, such page feeds the troll, as seems showing a lot of attention to vandal (multiple warnings, threats, sad words from the innocent users of that IP and so on).

There are talks on Wikipedia that the found vandal warning templates must be automatically substituted by they content, and there are some bots doing this. We think that there is no need to keep old warnings and old anonymous talk pages in general: the page can be just blanked instead, saving even more resources. The real discussions (not warnings) should go on the talk pages of the registered users, or on the article talk pages.

The solution

The proposed bot blanks the anonymous vandal IP talk pages that are older than 240 hours (since the last editing). It does not blank the normal user talk pages and also does not blank pages without the standard vandal warning templates.

This bot

• Obtains the list of IP pages where the vandal warning template is included (via "what links here"). We do not use the subst: in anonymous pages as they do not last longer than two days and vandal seems not deserving needed Vikipedia resources. There are anonymous IP talk pages with such unsubstituted templates in English Wikipedia also (see [11], for instance).
• Checks the page name, it must be the anonymous user talk page (User talk:N.N.N.N).
• If these two matches, checks the page history (visits the history page). The page must be older than 240 hours (10 days since the last editing session).
• If all previous conditions are true, replaces the page content to empty, also explaining the action in the edit summary.

The bot was tested in Lithuanian Wikipedia, first under strictly manual control and later in automated mode. It typically runs daily and just for several minutes. The program is written entirely in Java, without using any additional libraries.

The expected running frequency could be once a day, and (if required) it can be a limit of the deleted pages per session.

The bot could probably be useful for doing the same work in English Wikipedia. The bot operator is Audriusa 20:51, 14 March 2006 (UTC)

• Oppose the possible misunderstandings can be easily explained by to anyone who accidentally gets one of these templates and indeed several templates are solely around to warn people of the possibility. There is also a great need for other users and admins to be able to see warnings left on people's pages so they can determine A) whether an {{ipwelcome}} is warranted to welcome the person and/or whether the IP in question has a history of vandalism. Pegasus1138^{Talk | Contribs | Email} ---- 19:39, 15 March 2006 (UTC)

• Oppose While I accept Andriusa's good faith, my main problem with his rationale is that not all IP's are transient. Many IP's, probably most, are consistent vandals or non-vandals. For consistent vandals, warnings indicate their lack of innocence (first-time vandalism is presumed to be a test until warned) and lack of credibility on dubious edits. Conversely, for transient or shared IP's, my problem is that the bot is removing notices that IP's are AOL, for which warnings are pointless, as well as contact information about shared IP's at schools etc, sockpuppets, etc. Wuzzy 00:41, 16 March 2006 (UTC)

• Oppose. Some IPs are chronic vandals, and when I'm dealing with them it's very useful for me to be able to see that. When I'm going to an IP talk page to add a vandalism warning and I see that that IP has 20 previous warnings stretching back over six months, I handle it differently than if the IP has received one or no warnings. I'm sure that many other RC patrollers work in the same way. Hbackman 01:03, 16 March 2006 (UTC)

• Strongly Oppose as per the above. Elf-friend 07:09, 16 March 2006 (UTC)

• Oppose. Per above. Some IPs are static and therefore warnings should not be blanked. Anyway if no new warnings are added other users of that IP will not get 'new messages' so there isn't too much risk of them being coming across them. Editors can manually remove old warnings when adding new messages to an IP talk page if appropriate (what is appropriate depends a lot on the IP and therefore should be left to an editor not a bot to decide). In addition warning templates should be subst. Petros471 09:50, 16 March 2006 (UTC)

Consensus has been reached

I am convinced by the negative arguments and do not plan to continue with this idea. The discussion on this bot can be closed. The idea of time-dependent editings (not necessarily blankings) probably can be reused by someone suggesting something completely different in Wikipedia:Bot requests. Audriusa 08:17, 16 March 2006 (UTC)

Just for completeness, adding a reference to User:Tawkerbot which has a similar blanking function and ran into similar problems. Femto 15:06, 16 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

User:NongBot Interwiki backlog

I'm using meta:interwiki.py from meta:pyWikipedia. Mainly doing the backlog from Thai Wikipedia. --Manop - TH 21:54, 14 March 2006 (UTC)

Support good idea. Jtkiefer^{T | C | @} ---- 22:50, 16 March 2006 (UTC)

Trial run approved. Rob Church 01:22, 30 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

Pegasusbot

I'd like permission to run a bot, preferably getting a bot tag tag using the [[replace.py script tofind and replace to subst templates as defined by Wikipedia:Template substitution to find statements like {{test}} as well as the other defined templates and replace them with {{subst:test}}. Pegasus1138^{Talk | Contribs | Email} ---- 19:33, 15 March 2006 (UTC)

Neutral expanding several character vandal warning templates into they much longer content also seems vaste of Wikipedia resources and bring benefits only if the page is server frequently enough. The bots like Pegasusbot should optimize the frequently used pages only. Audriusa 20:17, 15 March 2006 (UTC)

The Wikipedia:template substitution guideline page states that these templates should always be substed and this bot would only help enforce that. If you have a problem with that I suggest you take it up there because it can (and is) done without the help of a bot every day, though a bot would be of great help due to the fact that running through the lists for transcluded templates by hand takes a very long time even when using a tool like AWB. Pegasus1138^{Talk | Contribs | Email} ---- 20:24, 15 March 2006 (UTC)

Note: His oppose vote is related to his request above which is pretty much the opposite of mine, conversley my oppose vote on his bot proposal is for some of the same reasons. Pegasus1138^{Talk | Contribs | Email} ---- 01:16, 16 March 2006 (UTC)

I am convinced by the negative arguments against the User:Terminatorius idea but still do not support yours. It may be pretty good as it is. Audriusa 08:30, 16 March 2006 (UTC)

I have recently tested the bot on a very small sample of pages (see Special:Contributions/Pegasusbot) and it works perfectly at substing pages, this can also be done using one template at the time substing or doing multible at a time using exported wikilink lists. The bot would be run no more than once a day and possibly less than that if there is no current need for a given time period. The edits could also be reasonably spaced out in time. Pegasus1138^{Talk | Contribs | Email} ---- 20:24, 15 March 2006 (UTC)

Isn't there already a bot that does this? If I'm mistaken, I think it's a fine idea. If there's already another bot that subst:s templates, I don't think that we need redundant bots. Hbackman 01:22, 16 March 2006 (UTC)

There's User:Bluebot but to my knowledge he hasn't been extremely active in substing templates so it wouldn't be a redundancy as far as I can tell. Pegasus1138^{Talk | Contribs | Email} ---- 01:31, 16 March 2006 (UTC)

• Support definitely a needed bot. Jtkiefer^{T | C | @} ---- 22:51, 16 March 2006 (UTC)
• Support, but isn't it better to use template.py (a script designed to actually subst templates)? Fetofs ^Hello! 22:48, 18 March 2006 (UTC)

I read through it and I think template.py is more useful for changing from the old style of template to new style of template and replace.py works just as well. Pegasus1138^{Talk | Contribs | Email} ---- 02:36, 19 March 2006 (UTC)

Comment: I have also tested it out doing taks such as emptying out categories and moving articles in categories to other categories as per [[WP:CFD}] which I plan on adding to my bot's routine if it gets approved, all overseen and started manually by me of course. Pegasus1138^{Talk | Contribs | Email} ---- 10:01, 19 March 2006 (UTC)

• Support ILovEPlankton 20:22, 19 March 2006 (UTC)
• Neutral. We already have User:Tawkerbot(1) to do that, no? --Rory096 02:30, 20 March 2006 (UTC)

We do but there's still a large backlog of that and it's a big task, I also (see my comment above) plan on doing other work with it including dealing with WP:CFD tasks. Pegasus1138^{Talk | Contribs | Email} ---- 02:55, 20 March 2006 (UTC)

• Comment since it's been a week and there seems to be no real objections can an admin close this when they get a chance and alert someone to give my bot a flag (I've already tested it out and proved that it works as can be seen by the account's history). Pegasus1138^{Talk | Contribs | Email} ---- 22:56, 22 March 2006 (UTC)
  • Comment: The place to do that is m:Requests for bot status, and you don't need to be an admin to do it, it's even better that you request it yourself. Even though you said you analyzed both scripts, replace.py has to be told where to look for the templates, while template.py finds them automatically. Fetofs ^Hello! 23:17, 22 March 2006 (UTC)
    • Done, but now they're asking for approval here. Pegasus1138^{Talk | Contribs | Email} ---- 23:47, 22 March 2006 (UTC)
      • This bot is has now bot-status. Mix up with the request of #User:CmdrObot --Walter 00:03, 23 March 2006 (UTC)
        • Archiving... Fetofs ^Hello! 00:09, 23 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

Drito -> Drinibot

I'm requesting feedback on User:Drito which is my bot for grunt work (mostly substing templates). I plan on request the bot flag next week, so please comment on it. I ran it a few days without flag so the contibs list got populated and you can see what it did. -- ( drini's page ☎ ) 03:01, 16 March 2006 (UTC)

Since the purpose of the bot is shown now, I won't run it anymore until I get enough feedback and the flag. -- ( drini's page ☎ ) 03:02, 16 March 2006 (UTC)

Neutral Despite the fact that I have a bot request currently pending for the same type of thing (see above) I see nothing bad from this bot except possibly your choice of edit summaries which is not quite as informative as they could be. As I have my own request pending for the same thing I am not giving a definite one way or another as to avoid a clear conflict of interest. Pegasus1138^{Talk | Contribs | Email} ---- 03:09, 16 March 2006 (UTC)this is mostly for substing but there are lenty of grunt jobs out there for admins...

Well, the edit summaries can be changed, can you point me what's wrong with them? How to improve them? I also don't see the problem having two bots. As I said, -- ( drini's page ☎ ) 02:35, 17 March 2006 (UTC)

And also, well, I'had announced since more than a week before I planned to do this bot, and I was just waiting to do the formal request, see [12] dated march 10

• Support although it may be repetitive if the above bot also gets approval. Jtkiefer^{T | C | @} ---- 22:52, 16 March 2006 (UTC)
• Support Sounds like a good idea to me. ILovEPlankton 22:08, 22 March 2006 (UTC)

Oh it seems that in norwegian, the word "drito" is very obscene, so I'll stop editing as "drito" and the account will go inactive, I've just created "Drinibot" and I'll be using Drinibot instead of Drito as a bot. So this request should be about Drinibot. EVerything that was said about Drito should be now about Drinibot. -- ( drini's page ☎ ) 07:39, 22 March 2006 (UTC)

You can't start the thing off doing one task and then state that it'll do whatever the hell else you feel like setting it doing, I'm afraid. You need to be up-front about what it is doing, when, and get approval for individusl tasks. Rob Church 01:34, 30 March 2006 (UTC)

Ok I've put a more specific statement on the bot's talk, I copy here:

• substing templates (using either AWB or pywikimedia) and (when I figure out how
• create capitalization redirects (like Giant steps for Giant Steps or Albert einstein for Albert Einstein)

and that's all the bot will do. -- ( drini's page ☎ ) 00:05, 1 April 2006 (UTC)

About this Giant steps for Giant Steps thing etc, that goes against Wikipedia:Redirect: Don't fix links to redirects that aren't broken.--Commander Keane 00:34, 1 April 2006 (UTC)

No, you're misunderstanding me. Let's say for instance United states didn't exist. People typing "united states" on the earch box wil lbe sent to a blank page. So United states was crated as redirect (same happens with the 2 ones above (I'm actually the creator of Giant steps since I got puzzled on a search). So what the bot would do is to lookwhen those capitalizations redirects are missing, and then creating them. I'm not fixing redirects that arent broken. -- ( drini's page ☎ ) 01:00, 1 April 2006 (UTC)

See [13] for my request at the pywikipedia mailing list for that. Meanwhile (since I havne't completely figured out how to do it) I'll only be doing substs. -- ( drini's page ☎ ) 01:02, 1 April 2006 (UTC)

Indeed, as I've stated in other places on this page, even non-broken redirects should be fixed if they (while still being helpful), actually exist to accommodate an incorrect usage. An article writer's failure to capitalize "United States" or use the accent mark in "JonBenét Ramsey" or properly spell "Ashlee Simpson" (by mistakenly using the more common "Ashley") would fall into this category. Links to such redirects, being clear and present typos, should always be replaced with the correct link text, and never made into a piped link, (which, unfortunately in this case, is the default behavior for both solve_disambiguation.py and popups.js). — Apr. 10, '06 [11:02] <freakofnur_xture|talk>

You're also misunderstanding me. The bot would NOt change any current entry. It would create new ones. For isntance, a few minutes ago I was looking for Once Upon A Time In America, but I didn't know the proper capitalization, so I went and put Once upon a time in America which sent me to a blank page. There should be a redirect from there instead of a blank page. What would the bot do will be to create such "capitalization redirects" i.e. missing redirects where only capitalization changes from the real entry. (I didn't fix this particular instance so you would know what I'm talking about). -- ( drini's page ☎ ) 17:40, 10 April 2006 (UTC)

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

Planktonbot

I would like to get my bot approved. It will edit using the AWB and the list of common misspelling to find and correct words that are incorrectly spelt.(Note if it is approved it will need to be unblocked.) ILovEPlankton 04:38, 18 March 2006 (UTC)

• Support sounds good; just be careful with it Where ^(talk) 04:49, 18 March 2006 (UTC)
• Support as per Where. - File:Ottawa flag.png nathanrdotcom ^{(T • C • W)} 04:53, 18 March 2006 (UTC)
• Support!--Tdxiang 陈 鼎 翔 (Talk)ContributionsContributions Chat with Tdxiang on IRC! 04:57, 18 March 2006 (UTC)
• Comment -- whatever you do, don't leave it running on automatic. There are exceptions to every misspelling, and they're easy to miss. I've mis-corrected enough of them, despite being human – a bot wouldn't stand a chance – Gurch 11:04, 19 March 2006 (UTC)

I won't, I will only run it when i can watch it.

• Support assuming it's being actively overseen all the time by a human. Pegasus1138^{Talk | Contribs | Email} ---- 19:56, 19 March 2006 (UTC)

I have to question the reason the bot is blocked prior to approving a trial run? Rob Church 01:24, 30 March 2006 (UTC)

I forgot i needed to get my bot approved. But the bot has already been unblocked. ILovEPlankton 03:11, 30 March 2006 (UTC)

Approved for a trial run for one week. Rob Church 16:10, 30 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

Modulatumbot

I've created Modulatumbot to automate general tasks, such as misspellings, tagging orphan images, and stubbing very short articles. The bot itself will be run for short periods of time on a personal computer, so excessive resource usage will be a non-issue. I'll also be working on an AI engine that determines what an ambiguous link should link to based on the content of the article, but that's not anywhere near completion. MOD 00:12, 20 March 2006 (UTC)

• Support ILovEPlankton 02:01, 22 March 2006 (UTC)

Thanks much, the bot has been operational now for the current week as under my username with the comment "modbot." MOD 02:35, 25 March 2006 (UTC)

• Had some serious concerns I should have posted here, but are now at Wikipedia talk:Bots#modbot? (didn't know this bot request was running - the requested name also does not correspond with the name used in edit summaries - neither does the bot run under its own account). After the second incident on the same encyclopedia page, I seriously, seriously oppose this bot. Modulatum seems rather clueless as to what the bot is actually doing. Not the kind of bot we need. --Francis Schonken 19:12, 25 March 2006 (UTC)

I coded the bot from the ground up and I know precisely why it's messing up. It's called debugging. I profusely apologize for my own human mistake of overriding subsequent edits, but the bot has absolutely nothing to do with that. MOD 23:45, 25 March 2006 (UTC)

• No, "clueless" included being clueless about wikisyntax for image tags; and in general that the bot messes with things between double square brackets, which is a completely different thing than doing spelling corrections in text outside double square brackets.
  Re. "debugging": not if it disturbs wikipedia: if your bot needs that kind of debugging, please contact MediaWiki developers and ask if they can provide an alpha or beta testrun environment for your new software. Until it tests positively (which it apparently does not do yet) it should not be used on life wikipedia. --Francis Schonken 09:21, 26 March 2006 (UTC)

• This bot activity must be suspended. First of all, you need a seperate bot account. Secondly you need to discuss why the bot has been making these errors that Francis Schonken indicated. The bot must not run until these issues are address and you are given the go ahead. If it does run the account will have to be blocked.--Commander Keane 20:00, 25 March 2006 (UTC)

Faulty dictionary entry is the reason for the mistake. The bot has now been redirected to User:ModulatumbotMOD 23:45, 25 March 2006 (UTC)

• Oppose as the owner is using his own account for the bot edits, he is misrepresenting his edits, and he appears to not even understand what his bot is doing. Pegasus1138^{Talk | Contribs | Email} ---- 20:41, 25 March 2006 (UTC)

I don't know where "doesn't know what he's doing" came from, and I understand the implications of using my own account for the bot. I've now changed the user-config.py to go to User:Modulatumbot. MOD

• Strongly oppose. Bots should never be used for fully-automatic spell-checking, and it's busy stub-tagging images. --Carnildo 20:42, 25 March 2006 (UTC)

If you look on my userpage, you'll see that the stub mechanism has been taken out due to false positives. MOD 23:45, 25 March 2006 (UTC)

Please halt this bot at once. It was not approved for a trial run and automated spell checking bots are prohibited on this Wikipedia. In addition, legitimate concerns about the bot's operator and the bot's programming and purpose have been raised, and these must now be addressed. Thank you. Rob Church 00:58, 30 March 2006 (UTC)

ZsinjBot

I created a user account, ZsinjBot, that will be running solely via Martin's AWB. The only thing it will do it subst out templates. The templates I currently have lined up are the vandalism warning templates (test, test1, test2, test3, and test4). I plan on starting with "test" as there are over 1500 un-subst-ed "test" templates. I have tested that it will do only this, as is evident in the short contributions I made with this account in the past few minutes [14]. Currently, the timer is set for 30 seconds (to comply with the trial period), however due to page load times, it usually comes to about a 40 second delay. I welcome all comments and questions. I will publically announce everything the bot will do on both its user page after gaining concensus with peers to do so.

In the future, once all vandalism warning templates are subst-ed, I may expand to common misspellings, etc, however I have no intention of doing so at this time. --Zsinj^Talk 01:48, 22 March 2006 (UTC)

Addendum: I will only be running ZsinjBot when I have the time to commit to supervising it. If it is deemed appropriate and, most importantly, safe to do so, I would not mind running it for extended periods of time unsupervised (with a longer delay, of course).

2nd Addendum: Hm... I just read the three requests ahead of mine and it seems this is a very popular use for a bot. Whether or not ZsinjBot is granted a trial and/or the flag, keep in mind the uses of AWB and the fact that I would never do anything near controversial without consulting others, espically admins. --Zsinj^Talk 01:54, 22 March 2006 (UTC)

• Support i like the idea, but most people are very sensitive about just substituting templates. ILovEPlankton 01:56, 22 March 2006 (UTC)

• Hence the reason I'm leaving it's possibilities open-ended. Disambiguation repair is a semi-automatic task that ZsinjBot can do a lot faster than I can by hand. Image tagging, such as where an image has a blank summary, is also something that can be done semi-automatically. New page patrol and stub sorting are other tasks I enjoy that ZsinjBot can assist with. Thanks. --Zsinj^Talk 02:11, 22 March 2006 (UTC)

• then it sounds like there is no reason for it to fail. ILovEPlankton 02:21, 22 March 2006 (UTC)

Approved for a week's trial run for the substitution part. Rob Church 01:33, 30 March 2006 (UTC)

Report: After almost two weeks, I have not had any complaints that have not been resolved. ZsinjBot was blocked once due to a bug that is explained on User:ZsinjBot. Should I go ahead and request approval for the bot flag? Thanks. --Zsinj^Talk 18:02, 12 April 2006 (UTC)

Ganeshbot

I've created Ganeshbot to create city/town articles in India based on on data provided by Census India 2001. I converted the data into a comma-seperated file. Bot will read the file line by line and create article stubs. Please see examples, Aadityana and Aambaliyasan, that I had created in the sandbox using the bot. It will be manually run by me. There are 5161 towns listed. It should take a couple of hours to complete. This is similar to the User:Rambot that created U.S city/town articles. Thanks, Ganeshk (talk) 08:33, 26 March 2006 (UTC)

A few questions:

• Do you have more information you can put into the articles besides the name and rough location? Part of the strength of the Rambot articles is the amount of information they contain: basic geography, a basic demographic profile, and frequently a map showing where the place is.
• How fast is the bot editing? To keep from overloading the servers, bots shouldn't edit more than once every ten seconds. Creating 5161 articles should take at least 15 hours.

--Carnildo 09:08, 26 March 2006 (UTC)

Reply:

• The excel file had the following columns,

City Name,Urban Status,State Code, State Name, District Code and District Name

Let me research if there is anyway possible to expand further.

• I have added 30 seconds delay between each edit. Couple of hours was just a guess. Out of 5161, many exist already. I have not run it on the entire file yet. So I don't have a time estimate.

- Ganeshk (talk) 09:22, 26 March 2006 (UTC)

Per suggestions by User:Carnildo, I had made changes to the bot. Could you please look at Aadityana, Aambaliyasan and Kodumudi and give approval for the bot?

Changes:

• Added population count
• Added Geo-coordinates and Altitude

I would like to run it for all towns.

Thanks, Ganeshk (talk) 08:20, 1 April 2006 (UTC)

I think it looks good, I would give it permission, but I think it might need greater community approval first, as it is such a large project. Maybe you could mention it at the village pump? Martin 12:56, 1 April 2006 (UTC)

We should only be happy to have something like this, and eventually for all the countries in the world. Hearily support, except perhaps the demographics could be moved to another paragraph, like Rambot did. But if this is all the info you have then this is probably fine. --Golbez 20:05, 1 April 2006 (UTC)

Martin, Thanks for the tip. I put up a message on the village pump.

Golbez, I have added two new sections, Geography and Demographics. Please check Kodumudi. Thanks, Ganeshk (talk) 22:00, 1 April 2006 (UTC)

How do you feel about exploiting Geographic references like Rambot, instead of adding the same reference to all the articles? Melchoir 21:52, 1 April 2006 (UTC)

Melchoir, Could you please explain with an example on how Rambot used Geographic references? Thanks, Ganeshk (talk) 22:00, 1 April 2006 (UTC)

Done. Added geographic reference similar to Rambot. I used Template:GR format since the number will not be constant. - Ganeshk (talk) 22:40, 1 April 2006 (UTC)

Please don't go ahead Bot articles are horrible, especially the awful rambot ones about American localities. A couple of human-written sentences are always better and they will all get done in the end. I see huge long term problems with the rambot articles. They are going to get very out of date, but when new census data is ready how are they going to be updated, especially those where people have added proper content? Is is going to be wiped along with the old rambot bilge, or is the rambot data going to be left in place forever. Please be patient and wait for Wikipedia to be written by people not machines. Hawkestone 23:03, 1 April 2006 (UTC)

I have to disagree; having the stubs help greatly, I find the information generally useful, and I prefer to have machine-made articles than none at all. Many of the Rambot articles have since been improved, sometimes vastly so, by editors. Treat them as stubs. --Golbez 23:53, 2 April 2006 (UTC)

The examples provided look pretty good, as for becoming out of date, this is applicable to all articles, regardless of how they were created. Creating the articles like this is good, as it provides a base for humans to build on and create much better articles, as is the case with many of the rambot articles. Martin 23:17, 1 April 2006 (UTC)

It should round to the nearest percent, rather than truncate (to avoid the gender proportions totaling to 99%). It should give the literacy rates for males and females, rather than the proportion of the literate people. TimBentley (talk) 19:21, 2 April 2006 (UTC)

I fixed the rounding issue. I rephrased the the literacy line. If you feel it still does not sound right, Could you please write the exact line how it should show using Kodumudi as an example. Thanks, Ganeshk (talk) 23:32, 2 April 2006 (UTC)

SmackBot and AWB operated by Rich Farmbrough

Please see:

• Above, #Bot flag for SmackBot (concluded approval to run under bot flag)
• Wikipedia talk:Bots#SmackBot not reviewing edits
• Wikipedia:Village pump (policy)#Rich Farmbrough another Bobblewik?

For a variety of incidents listed by (primarily) William Allen Simpson on the "Bots" talk page, and Village pump (policy), I formally request:

1. removal of "bot flag" from Smackbot;
2. Smackbot (or AWB and other bots/semi-bots by Rich Farmbrough) should no longer be run until reported issues have been solved. This includes (but not exclusively) the issue of delinking dates, presently marked as "no-consensus" at Wikipedia:Manual of Style (dates and numbers)#Avoid overlinking dates, and requested even by pro-delinking supporters not to create disturbance about it by bot or otherwise before the issue is solved (see Wikipedia talk:Manual of Style (dates and numbers)#Request to stop delinking dates in order to avoid additional distraction)

--Francis Schonken 10:21, 26 March 2006 (UTC)

Oh boy! There is I think no real issue over de-linking excess month names and days of the week. In any event SmackBot is currently blocked (and stopped) so no need to panic. I shall get round to commenting on those other pages in due course Rich Farmbrough 13:05 26 March 2006 (UTC).

See also SmackBot (talk · contribs · deleted contribs · nuke contribs · logs · filter log · block user · block log):

• 14 March 2006 -- User talk:Rich Farmbrough#Wikified Dates -- Wikipedia talk:WikiProject Days of the year#Linking dates in lead paragraphs

Continues through today. Replacing the bold around the title to

<!-- Do not change [[{{PAGENAME}}]] it preserves the date correctly formatted 
and stops robots from delinking it. -->
[[{{PAGENAME}}]] 

To start with, self-linking to get bold date preferences isn't a policy or guideline, it was off on a talk page.

But June 21 wasn't linked, it was already bold! The edit summary and comment make no sense. Of course, robots should not de-link dates, so there's nothing to stop!

YOu are quite right that June 21 was already bold. Virtually every other day of the year page was self linked, this one had been unlinked by an anon. And yes the decision was on a talk page Wikipedia talk:WikiProject Days of the year, in a discussion in which I was involved, and I'm sure that Wikipedia:WikiProject Days of the year doesn't have the same force as Policy or MoS. But it is really a rather specialised area to go onto an already bloated MoS. In terms of maintaining this style it is not unusual for either users or bots to delink and bold this per the normal procedure, which results in minor breakage. I have been through the entire year manually correcting it, very recently, and already some have been changed back by well inteniotned people or bots. People can be dissuaded with a comment, bots can't. Hence the solution adopted. If there's a better one almost certainly there is) I would be happy to adopt that. And if necessary I will edit all the 368+ date pages manually to implement it! Rich Farmbrough 13:20 26 March 2006 (UTC).

Would that be "at bot speed" and/or "using a (semi-)bot like AWB or other"?

I think the last sentence of your retort particularly nasty: it sounds like a declaration you'd rather start a revert war over this, than applying wikipedia consensus-seeking processes. Apparently the decisions taken at Wikipedia talk:WikiProject Days of the year haven't reached a broad consensus yet – otherwise other people wouldn't object to it so strongly. So I'd suggest you work on finding such consensus, e.g. publicizing in wikipedia:current surveys or in Village Pump that you think the "Days of the year" WikiProject has developed some "best practice" recommendations, you'd like to see accepted by the community.

Unless finding such consensus, I don't think the issue re. SmackBot can be considered to have been properly addressed. If you stick to the "whatever happens I'll implement it" this is a belligerent attitude that might get you in trouble sooner or later (just drawing your attention that we're not only speaking about the bot account any more in that case, so this discussion would have to move to another page).

Just as a side note regarding the solution you defend: there are strong feelings by some wikipedians against including HTML commentary tags in wikipedia pages. --Francis Schonken 16:33, 26 March 2006 (UTC)

Hey, [[WP:AGF}} and all that! I was offering to wade through several hundred pages and put in a better solution if anyone had one! And I don't think people object strongly, although I may be wrong, one user has taken exception to two edits to List of two letter combinations and List of three letter words (titles not quite right). Rich Farmbrough 18:12 26 March 2006 (UTC).

• 16 March 2006 -- User talk:Rich Farmbrough#Smackbot and US Articles

A request for SmackBot to do some work - I don't see the relevance. Rich Farmbrough 13:34 26 March 2006 (UTC).

• 16 March 2006 -- User talk:Rich Farmbrough#Enlighten me

A question not about SmackBot at all. Rich Farmbrough 13:34 26 March 2006 (UTC).

• 18 March 2006 -- User talk:Rich Farmbrough#Errors in SmackBot

One person thought that we should talk about the 2000 census in the present tense, I explained why I disagreed, and that person seemed OK w with it. Two people point out a capitalisation mistake (immediately corrected both in the process an all articles.) Another apparent error is a bug in the wikimeedia diff routines. Rich Farmbrough 13:34 26 March 2006 (UTC).

• 18 March 2006 -- User talk:Rich Farmbrough#Smackbot escaping apostrophies

A simple error only affecting "Playmates of the Month" if I remember correctly, simply and quickly fixed. Rich Farmbrough 13:34 26 March 2006 (UTC).

• 19 March 2006 -- User talk:Rich Farmbrough#Smackbot grammar changes
• 21 March 2006 -- User talk:Rich Farmbrough#SmackBot not reviewing edits -- reverted the table changes several times, and the bot would just go back and do it again! Obviously, AWB was not being monitored, as no human would make this mistake. It's visually clean and clear, with no sentences or grammar involved.

The problem that started this off! Rich Farmbrough 18:10 26 March 2006 (UTC).

• 23 March 2006 -- User talk:Rich Farmbrough#Smackbot edit to delink month
• 23 March 2006 -- User talk:Rich Farmbrough#Delinking months and days of week

"I really appreciate your bot..." Rich Farmbrough 00:52 27 March 2006 (UTC).

• 24 March 2006 -- User talk:Rich Farmbrough#Smackbot and Chemistry

Small problem, quickly fixed. Rich Farmbrough 00:52 27 March 2006 (UTC).

• 25 March 2006 -- User talk:Rich Farmbrough#Capitalizing the first letter of all chemical names is incorrect

Small problem, quickly fixed. Rich Farmbrough 00:52 27 March 2006 (UTC).

• 25 March 2006 -- User talk:Rich Farmbrough#Bot changes to U.S. cities
• 25 March 2006 -- User talk:Rich Farmbrough#Smackbot block

• Block In error Rich Farmbrough 18:10 26 March 2006 (UTC).

• 25 March 2006 -- User talk:Rich Farmbrough#Links

Request for assistance Rich Farmbrough 18:10 26 March 2006 (UTC).

• 25 March 2006 -- User talk:Rich Farmbrough#Smackbot blocked

The block occsioned by this complaint. Rich Farmbrough 18:10 26 March 2006 (UTC).

I'd suggest that the bot be permanently blocked until he certifies that he's personally reviewed and fixed every single edit ever done by the bot.... At a rate no faster than 1 every 2 minutes.

--William Allen Simpson 12:55, 26 March 2006 (UTC)

William seems to have dumped all the header from my talk page here implying they are all major problems. Some are the reverse, requests for more bot changes, some are queries and some are minor problems. Rich Farmbrough 00:54 27 March 2006 (UTC).

Geni has blocked the bot indefinitely, pending more information from the bot operator. Talrias (t | e | c) 12:59, 26 March 2006 (UTC)

• This page is for the requesting of approval for a bot flag, not for filing complaints against bots and/or their owners. Please file a Request for Comment or put a notice on the Administrators' Noticeboard if you would like to make a complaint. Thanks. --Zsinj^Talk 16:08, 26 March 2006 (UTC)
  • Hi Zsinj,
    1. This request is about SmackBot's bot flag in the first place;
    2. See also Commander Keane's reply here: "Wikipedia:Bots/Requests for approvals#SmackBot_and_AWB_operated_by_Rich_Farmbrough seems to be a good place to allow discussion." - please convince Commander Keane too if you think this has to move elsewhere. --Francis Schonken 16:33, 26 March 2006 (UTC)

I have to agree with the original request for removal of the bot flag. The flag was approved without any explanation from Rich Farmbrough as to what his bot was actually going to do, it was just a very vague request to be able to automate things which were tedious to do manually. That's not a request that should ever be honored. And the removing of date links does not have consensus. User:Zoe|^(talk) 17:33, 27 March 2006 (UTC)

I'm not removing date links, just links to months and days of the week. If anyone thinks they should stay linked, please say so and I'll stop. (I had a request to delink years associated with months, which I turned down for precisely that reason.) In general you can stop SmackBot any time, using the Big Red Button (tm) on the user page. Rich Farmbrough 23:18 27 March 2006 (UTC).

I also must reluctantly support removing Smackbot's flag status. I made two specific complaints about Smackbot edits a couple weeks ago and received absolutely no reply from Rich. I support most of what Smackbot does, but Rich is a bit too liberal with his use of Smackbot and he seems unwilling to engage in discussion. Such behavior seems irresponsible to me. Kaldari 02:01, 28 March 2006 (UTC)

I think I replied to almost every other post to my talk page. And I didn't miss you out for any unkind reason - I wrote a reply, but it never got posted - I don't know whether it was a browser crash or WP problems - my apologies for that. Nonetheless I'll talk to anyone, me, and often have. I'll also try to discuss, understand and explain, and will change my behaviour to accomodate others - viz slowing down, avoiding certain groups of pages, reverting or redoing, generating lists of problem pages, helping with tricky markup, adding a stop button. On the other hand this whole thing was kicked off by a user who hasn't replied to my messages, who has made misleading representations, and posted them on various admin pages without contacting me first - and has caused me (perhaps foolishly) to spend a lot of valuable time defending myself (on the other hand it's all a good learning...). Cest la vie. Rich Farmbrough 23:03 28 March 2006 (UTC).

I'd agree to a go-ahead as a de-flagged bot. Following Zoe's reasoning above, I'd say either make this a "various tasks" AWB bot (in that case not bot-flagged); or a flagged bot, but then only if positive after a new approval request on the basis of one or a few precisely described tasks, and new approval requests if new types of tasks are added to that.

Rich, is that a choice you want to make? Or am I too narrow-minded here? --Francis Schonken 15:20, 29 March 2006 (UTC)

It's a good deal closer. What I would suggest is leave it flagged, and I will look for approval for each task of over 100 edits. This is the way my thinking has been going with bots anyway, especially with the new streamline approval process. It won't impact single task bots, but blanket approval is a little bit OOT, anyway. What do the approvals group think? Rich Farmbrough 00:47 30 March 2006 (UTC).

To be honest, I've seen a lot of concerns raised above and I'm uncomfortable for this bot to retain a flag. I'd be open to Rich making some changes and sorting out problems (as he seems willing to do) and then approving another trial run, i.e. starting afresh; what do the others think? Rob Church 01:28, 30 March 2006 (UTC)

• Was SmackBot de-flagged? (I don't know where to check that) – there appears to be pretty good consensus in this section (Zoe, Kaldari, Rob, myself,...) that this bot should be deflagged. Was that done?
• SmackBot is "active" again ([15]), just realised the bot was unblocked a few days ago, with a rather mysterious unblock message ("I'm an idiot, unblocking actual username") [16] - is SmackBot an "actual username", then it should be non-botflag, of course.

I asked the admin who blocked it to unblock, rather than some random admin. That admin simply "unblocked" "User:SmackBot" (i.e. User:User:SmackBot in error, before unblocking the "actual user name". Hence the comment. Rich Farmbrough 19:22 30 March 2006 (UTC).

• Don't agree SmackBot changing "Chopin's" to "Chopin's" in the Claude Debussy article (diff - this was done 13:17, 28 March 2006). Was this "addressed" somewhere by Rich? If so, please let us know where. The contentious edit was not yet reverted anyway (and that's a task for Rich, if he wants to continue running a bot I suppose)

Changed to Chopin's. Rich Farmbrough 19:56 30 March 2006 (UTC).

• "100 edits" rule proposed by Rich is one of the most nonsensical things I ever heard. A bot should not do a single contentious edit under a bot flag. It's specifically the small tasks, in the contentious/non contentious border zone for which no bot flag should be applied. These should be done as ordinary logged-in user, and not at bot speed (so that they are easier to detect and one gets a user-user interaction in case of disagreement, not a user vs. a "I have permission to do this" bot operator).

Yes your right about this ("these should be done as ordinary logged in user"). Maybe it's me but "one of the most nonsensical things I ever heard" is not perhaps the most tactful way of putting it. Rich Farmbrough 19:54 30 March 2006 (UTC).

• So I don't think SmackBot should do a *single* edit any more, until de-flagged or having acquired an approval on the basis of a *new* bot request.

See m:Requests for bot status#en:User:SmackBot --Francis Schonken 10:00, 30 March 2006 (UTC)

Complaints procedure

I need no convincing that this (above) is a mess, and Zsinj has an excellent point. So please comment at a new discussion: Wikipedia_talk:Bots#Complaints_procedure.--Commander Keane 17:39, 26 March 2006 (UTC)

Tawkerbot (1) Fair Use Rationale Request Request

This is a request for approval for an addition to Tawkerbot's work (as the bot is idle right now)

The bot would leave a friendly message on an fair use image uploader's talk page to kindly leave a fair use rationale for the image.

It would essentially

1. Grab a category of fair use images
2. Use a regex to see if the worlds "rationale" or "fair use" do not appear (outside of the template)
3. Post a message on the image uploader's talk page requesting that they rationalize the image. This could be done for all uploaders of an image or just the first one.

I haven't seen anything that does it specifically and it might get us a little more than the 0.5% of images with fair use rationales. -- Tawker 15:39, 27 March 2006 (UTC)

Are you assuming that if the words "rationale" and "fair use" do not appear then there is no rationale? Martin 15:45, 27 March 2006 (UTC)

The problem is that most people don't understand fair use, and rationales are likely to be copy-and-pasted from some other image, leading to situations like List of Presidents of Portugal where the "rationales" for the images claimed that the photos showed "how the event depicted was very historically significant to the general public", when most of them were random portraits. --Carnildo 20:53, 27 March 2006 (UTC)

Essentially I was planning on just looking for the words "rationale" and "fair use" not in the template category. It wouldn't be perfect but it would be better than nothing. What I'm thinking is a "fair use rationale help page" which would have examples etc, essentially a template that the bot would subst in. I haven't written this bot yet, its just a proposal but I want to see if people want it before I commit to writing it. -- Tawker 23:18, 27 March 2006 (UTC)

My concern is that if people start copy-and-pasting random rationales, we'll end up with worse than nothing. --Carnildo 02:48, 28 March 2006 (UTC)

Certainly it would be useful to create a list. Perhaps a a database scan would be better. Then I think the License Cleeanup project (or whatever it's called) should be asked to suggest the next step. Clearly we would like proper rationale on each page, but even asking the uploader for a rationale would show good faith. Rich Farmbrough 13:09 29 March 2006 (UTC).

I'd like to see a pledge to eliminate too many false positives; also, please be aware that if this starts accelerating copy-pasting of fair use rationales as noted above, then it will need to stop. Rob Church 00:50, 30 March 2006 (UTC)

I support this request, but I'd like this to generate a list of all images where the user was warned, if at all possible, so that users could watch for bad rationales. This could be easily achieved with one list linking to all images warned; in this way, the "related changes" special page could be utilized. If this is possible, it would be much appreciated. Ral315 (talk) 14:36, 30 March 2006 (UTC)

Sounds like a reasonable request to me. Rob Church (talk) 19:38, 8 April 2006 (UTC)

{{subst:discussion top}}

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section.

BOT-Superzerocool

Hi everyone, I want a bot flag to run my interwiki bot in en.wikipedia. I have 3 flag bots in all proyects (spanish, estonian, ilokan) and I'm requesting in others wikipedias. (see my bots page in es).

The software to use is pywikipedia and is controlled by me or run in autonomous mode (in this mode, the problematic interwikies doesn't change).

I will run this bot for one year or more time.

Congrats Superzerocool 03:37, 28 March 2006 (UTC)

Sorry for the english, but I write some english and none es (!) :P

Just to note, the bot page is User:BOT-Superzerocool.--Commander Keane 04:31, 28 March 2006 (UTC)

I believe our policies state that users running interwiki bots on this Wikipedia need to be able to understand English, plus the languages of the wikis they will be linking to, no? Rob Church 19:58, 28 March 2006 (UTC)

Did that ever beome "policy"? Rich Farmbrough 13:11 29 March 2006 (UTC).

Yes I can understand english. Sorry for the english, but I write some english and none es (!) :P (this is a joke!). I'm sorry for the joke :'(. Superzerocool 02:34, 29 March 2006 (UTC)

As far as I'm concerned, this sort of matter requires common sense, so if it's not a policy, it bloody well ought to be one. Rob Church 00:53, 30 March 2006 (UTC)

A comment, the bot Yurikbot have task from many wikies, including arabic wikies. Superzerocool 02:43, 29 March 2006 (UTC)

Approved for a trial run. Rob Church 01:25, 30 March 2006 (UTC)

Ok, I use the trial time. Thanks :P Superzerocool 02:44, 30 March 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.

Werdnabot

Werdnabot (talk • contribs) will be used to maintain various aspects of my .NET Bot Framework. This includes opt-in announcements for users of the .NET Bot Framework (opt-out for approved users of it) regarding new versions, auto-subst:ing on my talk page and other areas, and similar tasks that I will program it to undertake. It is written using my .NET Bot Framework, in C# (.NET Framework 2005). Most of the time, it will undertake its tasks attended, although I may leave it to run a batch job unattended occasionally. It is currently programmed to delay 30 seconds between edits, pending approval - at which point it will probably run at one edit per five or ten seconds. Werdna648^T/_C\^@ 06:29, 28 March 2006 (UTC)

Wikipedia:Spam says of talk page spam: "Don't use a bot". I'm not sure why auto-subst'ing on your talk page requires a bot.--Commander Keane 06:45, 28 March 2006 (UTC)

The reference to not using a bot for internal spam is referring to spam to those who have not requested the spam. Spamming to an opt-in list with a bot is okay. Werdna648^T/_C\^@ 08:59, 28 March 2006 (UTC)

"And similar tasks that I will program it to undertake" - we need these disclosed up front. Please clarify what "opt-out for approved users of it" means. Rob Church 00:54, 30 March 2006 (UTC)

All users who are approved to use the .NET Bot Framework are automatically listed on the list for messages (here. They can remove themselves from the list and receive no further messages. As for "other tasks", I haven't the foggiest at the moment. At this stage, most likely stuff like cleaning up double-redirects, perhaps stub sorting, and other repetitive tasks. Werdna648^T/_C\^@ 23:06, 30 March 2006 (UTC)

Trial run for one week is approved for the stated items. Please come back and check when adding new functions to the bot. Rob Church 21:41, 31 March 2006 (UTC)

Thanks very much, although I won't get to use it much - I'm overseas for the moment. Werdna648^T/_C\^@ 16:59, 4 April 2006 (UTC)

Just letting you know, I've added and tested a new functionality (which I will hold off on using until I get approval for it), whereby i give it the name of a user to approve for the .NET Bot Framework, and it parses the list of requests, removes the request, adds a note that it was approved, then moves it to the Approved page. It then adds the user to the announce-list, and leaves a message on their talk page. To see this in action, see the diffs for removing the request, listing that the request was approved, with a note that it was approved by the bot, adding the user to the announce-list, leaving a message, and finally logging its actions. Another functionality I've added is logging. For every action that this bot does, it leaves a note on User:Werdna648/Werdnabot/Log, as demonstrated above. If possible, I'd also like permission to run the bot at faster than one edit per 30 seconds, as this makes it rather sluggish currently. Werdna648^T/_C\^@ 17:30, 7 April 2006 (UTC)

Requesting a bot flag.. Werdna648^T/_C\^@ 19:24, 10 April 2006 (UTC)

Eskimbot

I'd like to get bot status for my bot, Eskimbot, which has now been running for 6 mounths. It adds interwikis and fixes double-redirects. It currently has more than 19000 edits on the English Wikipedia. It is already flagged on fr:, ja:, sv:, nl:, it:, de:, es:, pl:, no:, os: and eo:. ▪ Eskimo ☼ 10:59, 1 April 2006 (UTC)

Question Are you fluent enough in the listed languages to recognize a mistake by this bot and/or do you have someone you can run language links by on those wikis that would be able to recognize and deal with such mistakes? Pegasus1138^{Talk | Contribs | Email} ---- 03:23, 3 April 2006 (UTC)

I speak English, users from foreign wikis contact me if they see I made a mistake on their wikis, I fix them. There is no "user must be fluent in our language" policy on the Wikipedias I work on. An admin from os: even asked me to run my bot on his wiki, knowing that I didn't speak a single word of his language (it's similar to Russian). ▪ Eskimo ☼ 16:05, 3 April 2006 (UTC)

On this Wikipedia, we require that users running interwiki bots speak enough of the languages involved that mistakes will not occur. Please convince us that this is the case, otherwise there will be no bot flag. Rob Church (talk) 19:38, 8 April 2006 (UTC)

OK, so no bot flag :(. Although I don't understand your policy, since a bot run on your Wikipedia will be checking ALL Wikipedias, even if it doesn't have accounts there. ▪ Eskimo ☼ 07:51, 9 April 2006 (UTC)

CJBot

I'm wanting to get approval for my bot, CJBot. It's main tasks will be to bypass redirects (for example change links from Ask Jeeves to Ask.com. Also, the bot will apply AWB's general cleanup fixes, such as simplifying wikilinks. Also, eventually, it will fix typos.

The bot will run on AutoWikiBrowser and will be automated. Computerjoe's talk 11:19, 2 April 2006 (UTC)

Okay let's see how you do. Be careful about which links you bypass. Many of them should remain as-is, and I'm not sure about the "Jeeves" one being a good example of either. I suppose, if in doubt, leave it. — Apr. 2, '06 [14:00] <freakofnur_xture|talk>

Please be aware that this Wikipedia has a policy against automated spelling correction bots which run unsupervised. I don't want this thing "eventually" fixing typos if it's going to be running on its own. Rob Church (talk) 19:31, 8 April 2006 (UTC)

I when he says "fix typos" he is referring to bypassing redirects tagged with {{R from misspelling}}, {{R from title without diacritics}}, and other redirects that involve decidedly non-ambiguous typographical errors. For example:

[[Jose Canseco]] should always be changed to [[José Canseco]]

The entire link text should be replaced, and never piped as [[José Canseco|Jose Canseco]] as this would remove the referring page from Special:Whatlinkshere/Jose Canseco but keep the typographical error visible to the reader. Other cases where the intent of the misspelling is so obvious that such a tagged redirect exists, should also be bypassed in this fashion. — Apr. 9, '06 [07:19] <freakofnur_xture|talk>

This bot has really malfunctioned, screwing up around 30 articles. I think this should be blocked, which is a shame as I put so much time in it. Computerjoe's talk 08:53, 11 April 2006 (UTC)

Tangobot

I'm requesting approval to run Tangobot, which will run interwiki.py from the pywikipedia framework to maintain interwiki links between the English and Japanese Wikipedias. It will remain manually assisted for the time being. (If this changes, approval will, of course, be sought). - Tangotango 14:09, 2 April 2006 (UTC)

Question: Does the policy on the japanese wiki allow this type of running and will you seek approval there pre-emptively as you appear to have done here so that there are no issues in running the bot? Pegasus1138^{Talk | Contribs | Email} ---- 03:21, 3 April 2006 (UTC)

Japanese Wikipedia policy requires that bot operators "Run [their] bot without a bot flag at least for a week, at intervals of 30 seconds or longer" before requesting approval. (their Bots Request for Approval page is in English) I will be adhering to both policies by requesting permission here first, and running the bot at intervals of 30 seconds or more for the approval period. - Tangotango 03:42, 3 April 2006 (UTC)

Sorry, I don't think I answered your question fully. Japanese Wikipedia policy allows running of inter-language bots, and indeed one such bot has been approved on the page linked above. - Tangotango 03:45, 3 April 2006 (UTC)

• Support on a trial basis for now though I'm optimistic that this should get full approval unless something catastrophically bad happens during the trial period. Pegasus1138^{Talk | Contribs | Email} ---- 03:45, 3 April 2006 (UTC)

Approved for a trial run for one week. Rob Church (talk) 19:26, 8 April 2006 (UTC)

Pearle doing wikify-date

By request, I'm doing a test run of Pearle sorting to-be-wikified articles by date. Since this is nearly identical to the cleanup-date task she is already performing, this isn't really something new, but I thought I'd mention it here and make it official. -- Beland 03:11, 3 April 2006 (UTC)

• Support Pearle has been doing nothing but good work so far and I trust that this is more of the same. Pegasus1138^{Talk | Contribs | Email} ---- 14:01, 3 April 2006 (UTC)

No problems. Rob Church (talk) 19:29, 8 April 2006 (UTC)

No reason to object, obviously, but do you think you could easily peruse the article's history (not every edit, just the last edit of each month), searching for the relevant tag in the wikitext, to determine which month the template was actually added, rather than dating them all to the current month? That would avoid overpopulating one category, and give a better perspective on our cleanup/wikification priorities. — Apr. 9, '06 [07:05] <freakofnur_xture|talk>

That's what it's doing now. -- Beland 16:16, 9 April 2006 (UTC)

Ok. Cool. For all I know it was doing that all along and I didn't notice, because all the ones I saw appeared to be dated as whatever the current month was, no worries. Go for it. — Apr. 10, '06 [10:50] <freakofnur_xture|talk>

User:Zbot370

Ok, I ran this for about a month now, and I wonder if there is anything else I could do to improve it before I submit it for a bot flag. Thank you. User:Zscout370 ^{(Return Fire)} 02:28, 4 April 2006 (UTC)

I've heard no complaints, and I don't see a reason not to approve this one's bot flag. Anyone got any objections? Rob Church (talk) 19:29, 8 April 2006 (UTC)

CatHeadBot

I'd like to run User:BmearnsCatHeadBot as an automated bot which will scan certain wikiproject related categories and "what links here" for related templates, and make sure that they are in synch. For instance, all articles in Category:WikiProject Musicians articles are supposed to have Template:Musician as a talk header on the article's talk page according to WikiProject Musicians guidelines. This bot will be using the Pywikipedia python library, and will only be run periodically. It will not automatically update anything without user input, instead it will compile a list of conflicts, and await further instructions on which of those pages to update. B.Mearns^*, KSC 15:01, 4 April 2006 (UTC)

I posted this request a week ago, and haven't heard anything. I'm a little confused about whether or not I was supposed to create a user page for the bot before or after approval, but I haven't done so yet. Is that why there're no comments? B.Mearns^*, KSC 13:07, 10 April 2006 (UTC)

Shouldn't these categories be confined only to the talk page anyway? You should add the category to the template so that the talk pages of these articles are categorized when the template is applied. — Apr. 10, '06 [13:34] <freakofnur_xture|talk>

I don't think so. At this point, the category is supposed to be on the article page, the template on the talk page. If you have opinions on that, pelase discuss them on the Project's talk page

Compiling a list is not a bot job, either a database query or simply getting the two lists manually (AWB makes lists from what links here and categories easily) and then comparing them would be all round a much better solution. Martin 13:39, 10 April 2006 (UTC)

Well the idea was that the bot compiles the list, and then asks for feedback from the user on which pages should be fixed, and which shouldn't be, and then fixes the appropriate pages. It could be all automated, but that would make the bot a little dangerous in my opinion. For instance, the project page includes the template as an example, but it's not on the talk page. I don't want the bot changing that, so it tkaes some feedback so I can indicate not to "fix" that instance. B.Mearns^*, KSC 14:34, 13 April 2006 (UTC)

Ganeshbot (AWB)

I have already nominated this bot above as a India town bot. This second nomination is for running AWB to make general fixes on India-related articles. It will pick the list of articles from List of India-related topics and categories related to India. It will run on a automatic mode with a timer of 15 seconds. You can check it's contribs to see a trial run I did yesterday. Could you please give it a bot flag? - Ganeshk (talk) 16:07, 5 April 2006 (UTC)

Bot status for MessedRobot

It's been a week since my bot was granted a trial run, and now I am applying for bot status. —MESSEDROCKER (talk) 21:58, 5 April 2006 (UTC)

Approved. I saw no complaints, and having checked, the bot is useful. User compliant with policies and co-operative. Rob Church (talk) 00:32, 6 April 2006 (UTC)

Pearle to update WP:PNA

By popular request, Wikipedia:Pages needing attention is being overhauled, to bring all maintenance needs pertaining to a particular topic together in one place. Pearle has been enlisted to help keep the lists here up to date. I have debugged the process and you can see the results on the demo page, Wikipedia:Pages needing attention/Ecology and Agriculture. Most of the processing to determine which articles should be posted is done offline, from a database dump. The bot does read all of WP:PNA's subpages, looking for those that have been "configured" to accept its output. (Volunteers are now working on configuring more sections.) More details are on Wikipedia:Cleanup process/Cleanup sorting proposal. Please let me know if there are concerns. -- Beland 03:45, 7 April 2006 (UTC)

Bot status for Planktonbot

It has been over a week since my bot got approved for a trial run, now i would like to get bot status. ILovEPlankton 15:34, 7 April 2006 (UTC)

Cydebot

I would like permission to get Cydebot (talk · contribs) a bot flag. In the beginning I will run him manually, but in the future I hope to eventually have the task be fully automated. As for what the bot actually does ... I think I'll let Cydebot speak for himself.

I am written using the pywikipedia framework. My sole goal thus far is to examine a day's block log, sniff out the indefinite blocks, parse the block reason, and if it is a username block, add {{UsernameBlocked}} to their talk page and {{Indefblocked-username}} to their userpage (if they aren't already there). I feel that this will be helpful because many users end up getting blocked but they do not know why, and although human administrators may be good at sniffing out the bad names, they aren't so good about always about leaving a message explaining the block. My reliable robotic nature will help me to overcome these human limitations. I am manually run for now, but eventually I expect to do my work automatically.

I think this bot is important enough to justify receiving bot status because of the following rationale:

1. In any given day, dozens, if not hundreds, of new user accounts are indefinitely blocked because of their name.
2. Very few, if any, are tagged with any sort of rationale. The only rationale is a cryptic "user..." block reason in the block log; newcomers aren't going to know what that means.
3. Some of the accounts being blocked are legitimate good faith attempts by newcomers.
4. We don't want to bite the newbies and indefinitely block their accounts without a stated reason because that makes it very unlikely they will stick around and actually help the project.

Also, for non-username blocks, Cydebot will just leave the standard indef-blocked template.

--Cyde Weys 17:16, 7 April 2006 (UTC)

Approved for a trial run for one week. Please remember to throttle edits to no more than two per minute to avoid clogging up recent changes and watchlists. If no sizeable objections are raised afterwards, then a bot flag approval will be forthcoming. Rob Church (talk) 19:20, 8 April 2006 (UTC)

User:Pegasusbot expansion of usage

I am planning on expanding Pegasusbot's uses to also deal with redirects as shown by whatlinks here and a few simple find and replaces. I have begun doing this without official approval as this is non controversial and shouldn't be a problem but wanted official input to be gotten. Pegasus1138^{Talk | Contribs | Email} ---- 06:19, 9 April 2006 (UTC)

What sort of redirect work (nothing to do with this etc) and what sort of find and replace. Examples etc.--Commander Keane 06:58, 9 April 2006 (UTC)

Mostly redirects that are misleading or are too ambiguous to be useful, currently I am running a find replace on NCAA to [[National Collegiate Athletic Association making it so it does not change the display name. I am also working to perfect a method to do semi-automated disambig work though that requires more work before it could actually be put into full service. Pegasus1138^{Talk | Contribs | Email} ---- 07:14, 9 April 2006 (UTC)

Ahh ok, you are doing disambiguation link repair. m:Solve disambiguation.py is very effective for that.--Commander Keane 07:25, 9 April 2006 (UTC)

I'm contemplating that but if I do I'll add it as a request but I'm adding one task at a time and the substing I've had to suspend due to some bugs in how I was doing it. Pegasus1138^{Talk | Contribs | Email} ---- 07:31, 9 April 2006 (UTC)

This bot's edits violate Wikipedia:Redirect#Don't fix links to redirects that aren't broken and Wikipedia:Manual of Style (links)#Context. I inquired on your user page, and you replied that because it's a guideline, it can be ignored. I disagree; guidelines are actionable except in special circumstances. Do you feel there's a special circumstance here? --TreyHarris 04:48, 10 April 2006 (UTC)

I feel that as it is only a guideline if it makes navigation easier to get rid of a redirect it should be done otherwise it shouldn't be since it would be a useless edit, I felt that the NCAA redirects were a useful thing to bypass. Pegasus1138^{Talk | Contribs | Email} ---- 19:41, 10 April 2006 (UTC)

You're using circular reasoning here. Why was it a useful thing to bypass in the case of NCAA? What special circumstances obtained to where you "felt" that the guidelines "were a useful thing" to ignore? In what future cases might you make a similar determination? --TreyHarris 06:21, 11 April 2006 (UTC)

This is also probably a good time to mention that Pegasusbot ran around various bot userpages placing the emergency shutoff button without permission from the bot owners. This, combined with TreyHarris' point above, shows rather reckless behaviour (and misunderstanding over the way WP operates), I'd feel more comfortable if this bot didn't operate, let alone had expanded activity.--Commander Keane 07:16, 10 April 2006 (UTC)

Luckily my bot does good work so it will continue operating and you'll note that when it was requested that I stop putting the bot tag on pages I stopped immediately so that would so maturity not immaturity as well as good behavior. I also know exactly how wikipedia operates and I suggest in the future you Assume good faith. Pegasus1138^{Talk | Contribs | Email} ---- 19:41, 10 April 2006 (UTC)

I'd also like official comment on doing [[disambiguation link repair using m:Solve disambiguation.py which by essence of having to do it partially by hand would of course be only robot assisted. Pegasus1138^{Talk | Contribs | Email} ---- 19:54, 10 April 2006 (UTC)

Since there have been no other objections I am going to assume that it's okay if I expand my bot's tasks as listed. Pegasus1138^{Talk | Contribs | Email} ---- 23:55, 12 April 2006 (UTC)

QovulwBot

Would like to submit QovulwBot for approval or for a trial.

This is an automatic bot that will run through a list of 50-100 Wikipedia pages (for now 50-100, more after initial testing) pre-chosen by myself from a google search of wikipedia.org. It will search for these phrases (to be expanded later to the form the University <verb>) :

the University will

the University offers

the Univesrity is

the University does

and it will change them by making the capitalized 'U' into a lowercase 'u' per the Wikipedia:Manual_of_Style#Institutions. It will run no faster than one edit per forty seconds and will terminate as soon as it exhausts all of the pages I give it to search.

QovulwBot is run on Python with the pywikipedia framework .

I realize this isn't the most important bot in some senses but it still does a viable service for en.wikipedia.org by correcting an infraction of the Manual of Style. Also, I think that this type of formatting will very rarely, if ever edit a false positive because of the rarity of such a false positive.

-Snpoj 23:24, 10 April 2006 (UTC)

Looks like a useful function. How will you avoid quotes? Rich Farmbrough 15:45 12 April 2006 (UTC).

Yeah that's a good call I hadn't thought of that. I'll work on that. Also, to exclude: everything in a References section and maybe some other sections of the bottom if they exists. -68.210.211.4 01:40, 14 April 2006 (UTC)

Fixed it. Thanks for the suggestion. -Snpoj 02:43, 16 April 2006 (UTC)

SmackBot task approval

Task: replace "External link" with "External links" where more than one external link is present.
Method: Either manually using Firefox, or automatically using AWB.
Speed: If manually in batches of 20 - 100 each taking a few minutes. If with AWB 1-2 per minute.
Number: Approximately 1400 articles.
Frequency: Recurring: 1400 have occurred in less than a month since the previous cleanup.
Testing: A previous run of about 6000 caused no problems.

Regards, Rich Farmbrough 01:13 11 April 2006 (UTC).

So are we going back to the old "wait a week to see if there are any objections" model? Rich Farmbrough 16:41 13 April 2006 (UTC).

Looks good to me. Feel free to go ahead. --Carnildo 18:15, 13 April 2006 (UTC)

Thanks. Rich Farmbrough 18:45 13 April 2006 (UTC).

Cydebot II

Requesting permission for Cydebot to substitute {{UsernameBlocked}} and other talk page message templates that have been identified by the community as templates that should be substituted. This will all be done with standard pywikipedia template.py. --Cyde Weys 23:28, 14 April 2006 (UTC)

This is also related to the thing Cydebot got permission for a trial run for. --Cyde Weys 23:58, 14 April 2006 (UTC)

Hrrmmm, apparently also Pegasusbot got approval to do something like this, though he's using AWB and I'm using pywikipedia. There's no substantive difference in these two methods though, so since his was approved, I don't see a reason why this one wouldn't be either. --Cyde Weys 00:37, 15 April 2006 (UTC)