Wikipedia talk:Bots/Archive 14

This is an current talk page .

Hi! I have a question about the difference between using a bot and the very useful

You will require a bot account because your edits would still be coming in fast enough to clog up Recent Changes. In order to avoid that, you need a bot flag, hence; we need to see a separate account. Rob Church ^Talk 21:31, 4 November 2005 (UTC)

See! I knew there was a logical reason :-). Thanks, RobChurch. >: Roby Wayne ^Talk 19:34, 7 November 2005 (UTC)

We are Luigi30. You will be assimil... wrong bot.

I (Luigi30) am looking for approval on a bot to clean out

Approved for a week's trial run. Please list it on Wikipedia:Bots under the section for bots without a flag and keep us posted here. Throttle your edits to one every 60 seconds max. and if all goes well, you'll have no problems getting consent to apply for the bot flag. Rob Church ^Talk 21:29, 4 November 2005 (UTC)

I hereby request approval for running bot MalafayaBot. This bot will exchange interwikis between the English and Georgian Wikipedia using the pywikipedia software. Malafaya 18:01, 4 November 2005 (UTC)

Approved for a week's trial run. You can apply for a bot flag afterwards if there are no objections, but check back here first. Please list it on the

WP:BOTS page, under the section for bots without a flag, and throttle edits to one every 60 seconds max. Rob Church ^Talk

21:41, 4 November 2005 (UTC)

I am looking for approval to run my WikiBot: User:Aweebot . I will start the bot nightly or every other night at 8 PM PST. It will run on wikipedia en and it is running pywikipedia. This bot will do Solve Disambiguation, Categories, Redirects, interWiki. Why do I need the bot? Well I want to help out in the Wikicommunity to make wikipedia even better. Is it important enough for Wikipedia to allow my bot? I think another bot helping out would make wikipedia better, as there is a growing ammount of articles, we want to keep growing and not have mistakes that never get fixed on articles. Thanks --Actown 02:55, 3 November 2005 (UTC)

You'll have to be a bit more specific on what you want to do with disambiguation, categories, etc. --AllyUnion (talk) 01:10, 5 November 2005 (UTC)

Better? If not can you explain what you mean? I changed this: This bot will do Solve Disambiguation, Categories, Redirects, interWiki. Thanks, --Actown 01:24, 5 November 2005 (UTC)

How will it perform those tasks if it is unattended? Martin 10:43, 5 November 2005 (UTC)

Oh Then I will watch over it. --Actown 15:44, 5 November 2005 (UTC)

The scripts cited require user intervention by default. You will need to monitor the thing running. Exactly what will be it be doing with "categories, redirects and interwiki", however? We have a policy whereby users running bots modifying interwiki links must understand the languages being changed. Rob Church ^Talk 20:01, 6 November 2005 (UTC)

When talking about disambiguation, what you will have to do is not just 'watching over it'. The disambiguation bot is highly interactive; it is actually the user who decides for each separate page being changed. - Andre Engels 09:19, 9 November 2005 (UTC)

What will you be disambiguating? Everything? You'll have to be a bit specific on what you plan to disambiguate. What will you be doing with categories? Merging them? Editing them? Moving or changing the categories? What would you be doing with redirects? Correcting the redirects? Changing the redirects? --AllyUnion (talk) 12:37, 7 November 2005 (UTC)

I would focous on Video game disamb and redirects. --

wn ^e

20:34, 3 January 2006 (UTC)

On basis of what is the interwiki bot policy being made? The policy as it is now makes it impossible to actually run the bot, because it will get all interwikis. Apparently it was created by someone who has not first checked the actual working of the bot. Who was it, where has this been decided, and what can I do to get the policy under discussion again? - Andre Engels 09:16, 9 November 2005 (UTC)

You may propose a new policy here. The consideration was after a considerable screw up made by Flacus and his FlaBot (FlaBot) which removed interwiki links on several articles. Flacus made no attempt to continuously check his bot at the time, and his bot was subsequently banned. He also attempted to, apparently, get his bot unblocked by stating that his bot was confused for a vandal bot with FIaBot (FIaBot).

According to the block log:

• 00:22, 11 May 2005 AllyUnion blocked "User:FlaBot" with an expiry time of 3 weeks (Bot not operating as intended; Removing zh-min-nan or zh (chinese) links.)

And according to the contributions: Contributions of FIaBot (FIaBot)

• 18:48, 11 May 2005 (hist) (diff) m Japanese cuisine (robot Modifying:de)

Flacus was also repeatedly warned, both on his German userpage and his English userpage to fix his bot, but to no avail. See: Wikipedia_talk:Bots/Archive_interwiki#FlaBot. However, since his unbanning, his bot seems to be operating properly. The main issue of concern here is that we have bots that seem to operate, and end up removing interwiki links. Shouldn't the removal of interwiki links be done by someone who familiar with the language? --AllyUnion (talk) 21:39, 11 November 2005 (UTC)

Furthermore, shouldn't they reasonably understand English if they plan to run their bot here or at least have someone who can speak for them? --AllyUnion (talk) 22:03, 11 November 2005 (UTC)

It's hard for me to check the facts here; the edit you give above is indeed by FIabot rather than Flabot as he claims, but there are other complaints on the history page of the user discussion page which do seem to be valid.

However, I still am of the opinion that restricting bots to only edit languages known by their operators is too strict. First, the basic working of the bot includes the addition of interwiki links on pages that interwiki links exist to. Secondly, a lot of information can be found in languages that one does not speak. As an example I show this recent edit by Robbot - I do not speak Swedish, but I understand enough of it and of Wikipedia to realize that sv:Färgelanda is a disambiguation page, and that of the three meanings on the disambiguation page, Färgelanda Municipality corresponds to sv:Färgelanda kommun. - Andre Engels 10:45, 13 November 2005 (UTC)

One thing I do, but which I guess is hard to get into rules, is that I check through the 'normal browser' before allowing Robbot to remove a page or before making a decision where there is more than one page found for a language. - Andre Engels 13:20, 13 November 2005 (UTC)

How about we amend the policy to the following:

If using the pywikipedia framework:

• Please run the latest version
• Update on a daily basis

Users who run such bots must prove themselves to be responsible and harmless, and work closely with the community of interwiki bot operators. Users who come from other language Wikipedias must demonstrate some proof of credibility, such as being an Administrator of a major language Wikipedia, Bureacrat, developer for the pywikipedia framework, or being vouched by a credibile user on the English Wikipedia.

Would that be much better? Instead of focusing on whether a user can speak the language, how about focus on whether the user is responsible enough to keep on eye on their bot, and stop their bot if it goes awry. The issue that I want to try to avoid is that an operator of an interwiki bot refuses to stop his or her bot on the basis they believe it is working correctly, when there is evidence that is not. --AllyUnion (talk) 00:15, 14 November 2005 (UTC)

It would be acceptable to me, yes. The important thing in my opinion would be that people are careful when doing removals or non-trivial changes to interwiki links with their bots. My problem with your rule was that I think 'speaking the language' is too strict a rule for that, especially when it's also extended to the standard interwiki-bot additions. - Andre Engels 09:19, 17 November 2005 (UTC)

Please see Wikipedia:Administrators' noticeboard#A cut and paste bot. Who was running this? jni 08:37, 10 November 2005 (UTC)

Likely to be a vandal bot. --AllyUnion (talk) 21:39, 11 November 2005 (UTC)

Correction, apparently run by Infohazard who has not registered for a bot. --AllyUnion (talk) 22:06, 11 November 2005 (UTC)

I'm wondering if it would be worth having a standardised means of flagging unauthorised bots. How about a template to stick on the user talk page of a suspected bot? It would serve a triple purpose:

• to tell the user to confirm if he/she/it is a bot or not, and where to request approval before continuing (just in case the human behind it sees it) (not to forget the possibility of the same user account being used for both bot and human edits)
• to inform others that it is a suspected bot
• to produce a list of unauthorised bots that admins can examine, either by What links here or a category.

-- Smjg 12:04, 11 November 2005 (UTC)

Unauthorized bots should be blocked, without exception, unless they have requested permission here, or at least indicated that it is a bot. For example, Kakashi Bot is not necessarily an authorized bot, but it is listed under "Bots running without a flag" on the project page. Kakashi Bot is subject to human scrutiny, and is intentionally left without a bot flag. I was blocked for running unauthorized bot as well. Typically, it is asked that editors do not run bot-assisted or bot edits under their own account. Exceptions are made to this for administrative oriented bots, like Curps who runs an autoblocker. --AllyUnion (talk) 21:47, 11 November 2005 (UTC)

True, but my points are:

• Not all of us have the power to block.
• Sometimes one may be unsure whether a user making bot-like edits really is a bot.
• There's always the possibility of a random person, especially one who hasn't gone through the bot approval process, using a single account (or raw IP address) for bot and manual edits, either to try and disguise that he/she/it is running a bot or out of not knowing better, and that this needs to be taken into consideration when verifying bot accusations.

-- Smjg 11:36, 14 November 2005 (UTC)

Then find someone who does, like at

WP:AN/I. --AllyUnion (talk)

00:57, 16 November 2005 (UTC)

User:DHN-bot - an Interwiki bot to exchange interwiki with the Vietnamese Wikipedia. DHN 22:18, 12 November 2005 (UTC)

Please be more descriptive with your bot's user page. --AllyUnion (talk) 09:41, 14 November 2005 (UTC)

I've updated the bot's user page accordingly. DHN 06:52, 29 November 2005 (UTC)

What computer language do one uses to create Wikipedia bots? CG 22:35, 12 November 2005 (UTC)

Any language may be used to create a Wikipedia bot, but most typically people rely on the Python pywikipedia framework. --AllyUnion (talk) 01:01, 13 November 2005 (UTC)

Thank you. CG 12:01, 13 November 2005 (UTC)

This is a request from my talk page: --AllyUnion (talk) 01:04, 13 November 2005 (UTC)

Would it be possible for you to add to sandbot (or some other regularly-run bot) the unconditional deletion of

The answer yes, but requires community approval. --AllyUnion (talk) 01:04, 13 November 2005 (UTC)

Why not use a more obscure article title to keep red. If the more obscure article title keeps getting created (which would probably be vandalism to mess around with

Wikipedia:How to edit a page) could we protect the page from creation? Using a bot to regularly delete a page is an abuse of Wikipedia resources.--Commander Keane

09:35, 13 November 2005 (UTC)

The page we link to must be intended as a redlink, and must remain as a redlink. Protecting the page from creation would cause the red link to show up blue. --AllyUnion (talk) 13:29, 13 November 2005 (UTC)

This is the text from

• You can create it by clicking on the link (but please do not do so with this particular link).
• To create a new page:
  1. Create a link to it on some other (related) page.
  2. Save that page.
  3. Click on the link you just made. The new page will open for editing.
• For more information, see
  naming conventions
  .
• Please do not create a new article without linking to it from at least one other article.

Sorry about my ocnfusion over protecting a page. Will creation of the page show up on a Watchlist. If so, then a bot is hardly necessary (but won't do any harm), if a couple of admins have the page on their watchlists it will get deleted fast enough.--Commander Keane 05:34, 14 November 2005 (UTC)

Deleted pages can be placed on watchlists... although, the matter being discussed here is having my user account check the page's existence every so often and go off to delete the page. --AllyUnion (talk) 08:34, 14 November 2005 (UTC)

This is a new user who appears to be using some sort of automated or semi-automated spellchecker script, which appears to be systematically violating spelling conventions by switching to the British style. If someone could politely handle this matter, it would be appreciated, as I have little technical knowledge in this area. — MC MasterChef :: Leave a tip — 09:25, 13 November 2005 (UTC)

We do not allow automated spellchecker bots. We do permit semi-automated spellchecker scripts/bots so long as they include US and internationalized versions of spelling. --AllyUnion (talk) 13:35, 13 November 2005 (UTC)

This bot has been used by its owner

1. References are usually primary resources, and an encyclopaedia is a secondary resource
2. References are places you are directed to to check the veracity of the text in the article: in this case, the article is based on the 1911, so you would go elsewhere to check the validity of the 1911 itself.

Is there concensus for Bluebot's actions? This particular activity is not cited in the request (above) which is for other valuable stuff. (I have also asked this question on the template's talk page, and will do elsewhere as well.) Noisy | Talk 10:09, 13 November 2005 (UTC)

Please read the

Wikipedia:Cite sources, which say to put sources under a references heading. Also, the bot isn't really "moving" the template, it is adding the heading, because at present the 1911 tag often just floats around at the bottom somewhere, which is definately wrong. Plently of articles already use the 1911 tag under a references heading as well. Plus I might add that I have been monitoring all edits this bot has been making. Martin

10:15, 13 November 2005 (UTC)

Bluebot has been blocked temporary to resolve any issues raised by user Noisy. I also have confirmed that the bot is operating out of its specified parameters, in regards to adding the 1911 template. --AllyUnion (talk) 13:56, 13 November 2005 (UTC)

Additionally, User:Beland's question was still left unresolved above. --AllyUnion (talk) 13:58, 13 November 2005 (UTC)

Is there a specific list of replacements that folks can check for potential problems? -- Beland 07:08, 26 October 2005 (UTC)

The author is requested to resolve all questions, and problems before the bot is blocked. --AllyUnion (talk) 13:58, 13 November 2005 (UTC)

As is pointed out above, Noisy is completely wrong, there is no issue to resolve. I answered Belands question where he asked it, why do you say it was unresolved? Since then I have moved on to do other things with my bot. I have made ~60,000 edits with my bot with no valid compaints and many compliments. I have unblocked my bot now. Martin 14:12, 13 November 2005 (UTC)

I have no objection to the bot being used for the other purposes listed above, because I think they are valid and useful. If Martin desists from using it for the 1911 purpose, and proposes its use for 1911 in the proper way in which bot announcements are supposed to be used on this page, then I have no objection to the other work being done, and the bot can be unblocked.

As AllyUnion says, a more explicit listing of the actions undertaken by the bot would be welcome, as would a direct response to the two objections that I have listed so far. Pointing to an old Wikipedia page does not mean that I can't challenge the position, and ask for a new opinion from other Wikipedians before wholesale changes are made. Noisy | Talk 15:09, 13 November 2005 (UTC)

You don't have to build a concensus to stick to the guidlines, you have to build a concensus to change the guidlines, which is what you seem to want to do. You are allowed to challenge the guidline, but not revert my correct changes, and as it stands, no one agrees with your oint of view so I don't see the guidline changing.

As for listing what I want to do with my bot (note that I update my bot user page occasionally with a list of tasks) I don't mind doing it out of political correctness, but I don't see the point as lots of people don't bother mentioning it every time they do a new task. Anyway here is what I have been doing;

• Fixing numerous ==See also== and ==External links== errors. (complete)
• Re-categorising articles per
  WP:CFD
  .
• Subst'ing
  templates.
• "Touching" articles on request.
• Requests from users (such as updating template names).
• Ensuring 1911 tags are in a ==References== section.

Martin 15:50, 13 November 2005 (UTC)

Any operation you plan to make that isn't something you originally applied for needs to have a passing mention here. This gives the community some time to prepare for, or improve your idea that you plan to implement. It's an insurance type of thing. --AllyUnion (talk) 23:41, 13 November 2005 (UTC)

Ok, Note that things such as the subst'ing were discussed on the

Subst'ing page, and changes to do with the manual of style were discussed on the manual of style page. thanks Martin

23:44, 13 November 2005 (UTC)

You still need to make a note here, either a link to the discussion or something. --AllyUnion (talk) 09:36, 14 November 2005 (UTC)

Users using bots should be aware that they should make every effort to run their bots off a server, on a separate IP address. Due to changes in dealing with sockpuppets in the MediaWiki software, a block of a user with a bot that runs off the same computer that the user edits, will cause the user unable to edit. --AllyUnion (talk) 23:57, 13 November 2005 (UTC)

..."changes"? I thought autoblocking always worked like this. (Or at least, as long as I can remember.) In any case, this is more motivation to make bots halt if their talk page is edited, as mentioned at

Cryptic (talk)

00:47, 14 November 2005 (UTC)

Does the pywikipedia bot stop when the talk page is edited?--Commander Keane 05:05, 14 November 2005 (UTC)

No. At the current moment, it does not. Although it does check whether you have new messages... --AllyUnion (talk) 09:37, 14 November 2005 (UTC)

First, I want to apologise for running a semi-automated edit tool from my own account without even realising that there was a procedure for seeking approval. Someone soon put me right on that one, so I've stopped using it and come here.

I want to run a program that helps me correct common errors in cricket articles by identifying candidate errors for me to consider. I'm not even sure whether this counts as a bot. One proposal earlier on this page suggests to me that some people might not consider it to be a true bot, although I would tend to think of it as one; but let me give you more details.

So far, I've been running it to correct "test" to "Test" (in the sense of Test match, an international cricket match). The bot is written in Perl and it works as follows:

1. User (me) launches the program
2. It reads List of international cricketers, and downloads the first few that it hasn't read before
3. It finds ones with "test" in, changes "test" to "Test", and pops them up in a browser for me to read
4. When it's found 20 candidate articles, it stops
5. Then I go through the suggested changes and commit the correct ones

So it doesn't make any edits itself. I run it occasionally, it downloads maybe fifty pages, I make twenty edits myself in the space of a minute or two, and then I wait a few minutes or a few hours before running it again. Have a look at my last 500 contributions for the typical impact.

My ideas for future tasks that this could do

1. Link "Test" and "Test match" to Test cricket, not to Test match or Test.
2. Make sure all short biographies have {{cricketbio-stub}}
3. Add country-bio-stub to all cricketbio-stubs (ask on country page first, because this could generate a lot of new stubs in their lists).
4. Find articles entitled "N (cricketer)" but not linked from "N".
5. Find very short (sub-stub) articles, because an anonymous user has recently been making a lot of these.

We currently have something over 1500 cricket biographies, and a few hundred other cricket articles, so a program like this is a natural solution to improve conformance to

So my questions are

1. Is this a bot that requires approval?
2. If so, may I have approval?
3. Even if not, should I run it as a separate user instead of as me, and should that user have the bot flag?

Thanks very much. I hope that's enough information, but please ask if you need to know anything more.

Stephen Turner 12:54, 15 November 2005 (UTC)

We also need a bot to change everying in categories such as Category:English test cricketers to Category:English Test cricketers :) jguk 13:41, 15 November 2005 (UTC)

That should go on

WP:CFD as a speedy request, I can sort that out with my bot. '''''Martin'''''

13:52, 15 November 2005 (UTC)

jguk: Agreed, although I think that would be a job for a bot that committed its own edits (there are too many of them, and no risk of false positives), and I'm not sure I want to get into that yet. My program is deliberately ignoring those at the moment. Stephen Turner 13:54, 15 November 2005 (UTC)

This script looks pretty safe, as you are manually checking all the edits I don't think anyone can complain. '''''Martin''''' 13:52, 15 November 2005 (UTC)

Approved for test run for one week. --AllyUnion (talk) 21:41, 15 November 2005 (UTC)

Thanks very much, but I'm still unclear about one thing. Should I make the edits from a bot account or from my own account? Given that the bot is not making the edits itself but just suggesting edits to me. Sorry to be slow, but I want to be sure what the correct practice is. Stephen Turner 08:04, 16 November 2005 (UTC)

Bot account. --AllyUnion (talk) 21:14, 16 November 2005 (UTC)

I am requesting that AFD Bot to purge the page cache of Wikipedia:Articles for deletion at 00:01 UTC every day. --AllyUnion (talk) 21:53, 15 November 2005 (UTC)

I have listed a series of redirects that start with "2005 English cricket season/" here: Wikipedia_talk:WikiProject_Cricket#Redirects_that_need_fixing. Do you want a bot to go and remove the pages linking to them, then mark them for speedy deletion? --AllyUnion (talk) 18:18, 14 November 2005 (UTC)

Yes please, much appreciated. If you could turn Yorkshire v Worcestershire 7-10 September 2005 and similar pages to redirects as well, that'd be lovely - the list is here. Sam Vimes 18:43, 14 November 2005 (UTC)

Redirects to... where? --AllyUnion (talk) 07:20, 15 November 2005 (UTC)

Either the home team's page, or the period pages (1-14 June and so on). Whichever is easiest to program Sam Vimes 08:05, 15 November 2005 (UTC)

Kakashi Bot will be used to perform this task.

1. Take the list of Wikipedia_talk:WikiProject_Cricket#Redirects_that_need_fixing, and make certain no pages link to it (except for the pages Wikipedia_talk:WikiProject_Cricket and any AFD subpage), and mark the redirect for deletion. As it requires manual checking, whether or not the bot did the task correctly, I am inclined not to auto-delete. Will use: {{db|Requested deletion via bot per [[User:Sam Vimes|]]. The bot is suppose to remove everything linking here. Also specified as a task at [[Wikipedia talk:Bots]]. Please check almost nothing [[Special:Whatlinkshere/{{PAGENAME}}|links here]] before deleting.}}
2. Attempt to redirect articles in Wikipedia:Articles_for_deletion/Cricket_matches_articles/list to an appropriate page. (season date) --AllyUnion (talk) 22:04, 15 November 2005 (UTC)

I had an idea this evening - what about a bot or similar program (perhaps running on a local copy of wikipedia). It would cross reference between articles in different languages, looking for images that are in one, but not another.

Images frequently break the language barrier, and who knows what hidden gems might be found using this methodology? I propose that the results of this program would be put in a Wikipedia namespace page, so that individuals could work on cross-dissemination of images in a decentralized fashion (in the spirit of wikipedia).

A much more aggressive approach (I'm not sure what I think about this) would be to have a bot take this output and tag the bottom of every page for which an image was available in another language.

What do people think? I wouldn't mind providing some of the programming talent for this, but this kind of wikiwide action would need some pretty hefty official support. - JustinWick 01:33, 17 November 2005 (UTC)

Do you mean the images that reside on WikiCommons or just images that are missing in general? Please remember, there are fair use images on the English Wikipedia that can not be transferred to other projects. Also, we have Commons images, which are used across several projects. --AllyUnion (talk) 03:46, 17 November 2005 (UTC)

Well, Fair Use is a very gray area (at least in the united states) but... I am only talking about taking images from, say, the English version of wikipedia, say for an article on robotics (how apropos) and suggesting its inclusion in articles about robots in other languages of wikipedia (if they don't have it already). This way if any language gets a nice image for an article, there's a good chance that it can be reused in other languages. It seems unlikely to me that this transfer would be considered illegal (wikipedias of all languages have the same legal status, right?) but I admit I am no legal expert! - JustinWick 16:26, 17 November 2005 (UTC)

I had a similar thought to this as well, it would be best done off-line on the database dumps rather than by bot. I imagine it would look at the inter language links on an en article, then see if any of the linked to foreign articles (or a maybe just the fr and de articles) have more images than the english one, and generating a list of these articles, and the missing images. Generating a list off-line would not need any community support, and would make a nice community project. Hope that makes sense. Martin 16:46, 17 November 2005 (UTC)

I agree that that's the most computationally efficient way to do it, that's actually, in my mind, the easy part. To me the hard part is - people in other languages are trying to improve their articles, how do they know someone has compiled this great list? Moreso, how do they know the particular article they are looking at has some alternate images? - JustinWick 17:05, 19 November 2005 (UTC)

Such suggested images are okay for suggestions, compiled as a report from an offline database dump and should rather be placed on meta or something, and so long as the suggested image has a free license. However, it be more beneficial for suggestion images to be moved to the Commons and placed on a gallery on the subject. --AllyUnion (talk) 16:54, 19 November 2005 (UTC)

I can see it being put in the Commons, however the big thing I'm worried about is, how do people editing articles know this resource is available for a particular article? Maybe if there was some way of getting a Wikipedia Project started for media cross-distribution, such that people in different languages would just take articles off the list and put whatever images are appropriate into each article for which they are available. (marking them off when they are done). - JustinWick 17:05, 19 November 2005 (UTC)

Perhaps you can link to the commons page on the talk page of the article which the two subject matters relate to? Or use one of those boxes that says, "The Commons has more images on <subject>" --AllyUnion (talk) 09:32, 9 December 2005 (UTC)