This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current main page.
Basicly, a bot that sends a series of pre-typed messages to a number of users.
This could be used when your RFA gets an overwhelming number of support votes, like this one, It will be a pain in the ass to send everyone a thank you; or if you want to update everyone in a big wikiproject about a event or just want to have some sort of newsletter.
remove links to pages that redirect back to current page
It would be beneficial (imo) to have a bot to remove links to pages that redirect back to the current article (i.e. X shouldn't link to YifY is a redirect to X). I'm no programmer (otherwise I'd work on this myself!) but I'd have thought that it should be possible to do this by checking the what links here information and then unlinking anything that links to any redirect pages in the list. This might need to be a human-supervised bot (or AWB function?) as there may be occasions when the link is apropriate. It should only deal with the main namespace - links on talk could be useful and/or out of date relating to page moves/mergers. Possbily it would be of benefit in the Wikipedia: namespace if WP: shortcuts are explicitly ignored. Thryduulf23:04, 6 May 2006 (UTC)
Some notes on this and why it would need to be human-supervised and what else it could do. There are three types of links that link back to the article itself:
Alternate names, where the linking should be replaced just with bolding (if it's mentioned for the first time)
Accidents, where the link (but not its text) should be simply removed
Subtopics which haven't been created yet, and so link back to the main article. These redirects may be marked with {{R with possibilities}}
A bot like this would have to take these things into account, and so would need a human to discriminate the type of link. The bot might also mark redirects with {{R with possibilities}} too if that's the appropriate choice. I'm not volunteering to write it.—Pengo21:52, 27 May 2006 (UTC)
Right now, I have code that keeps the semi and full-protection sections of WP:PP up to date. I may add userpages later, but it is quite usefull now. I will make a copy of the code that does not have any messages and submitts the newform automatically. I can then get that times for User:VoABot.Voice-of-AllT|@|ESP19:48, 7 May 2006 (UTC)
I will begin running User:VoABot from the monobook...though I would greatly appreciated if someone good with programming can tell me how to run it of script on my computer, and not my script on Wikipedia.Voice-of-AllT|@|ESP03:24, 8 May 2006 (UTC)
This task would be as simple as examining Special:Log/protect and storing the most recent entry for each page encountered. Examining each line would yield:
Disregarding articles where the most recent action was "unprotected", the bot could examine the others and sort them chronologically in a table like this:
Yeah, it probably should be built in. Anyway, what you said is already what my bot does, see here [2]. If you want me to generate a chart, I would have to make it in userspace, and take some extra time to code in the table formatting.Voice-of-AllT|@|ESP11:56, 8 May 2006 (UTC)
Category:Council elections in England - wikify dates
I have tools to deal with most problems concerning dates. Leave a list of articles on my talk page and I'll have a look. RichFarmbrough 22:50 11 June2006 (GMT).
Request for bot to do simple daily maintenance of Current events
Over at Current events we're floating around a couple of new systems with several goals in mind, a couple of them being simplifying the monthly archiving process (when Current events is moved to April 2006, for instance) and then solving the issue of having only one day's worth of events on May 1 because all the previous ones were archived. My current proposal is to create a page for every date (May 1, 2006 for instance) and then just listing the last <<fixed_number>> date pages in Current events as templates. The hassle of monthly archiving would be virtually eliminated as the only necessary maintenance would be creating a new monthly page before the end of the month, creating 31/30/28 blank daily pages, and inserting them all into the monthly page as templates. The monthly page would begin blank but will fill up as each date's window of exposure passes at Current events (or of course if anyone chose to edit them afterwards).
Where we need your help is that every night we would need the oldest date removed from Current events and then the current date template inserted at the top of the list. I'm wondering if any current bot owners would be willing to add to their bot's daily activities the nightly rollover at midnight, and then the monthly mass creation of new, blank date pages (should this reach consensus as the new method which with a bot owner on board I think it would)? — GT14:27, 11 May 2006 (UTC)
Maybe we can create a template that does that automatically, instead of needing a bot? I'm not quite good at that, but someone could try. FetofsHello!21:28, 11 May 2006 (UTC)
Unless I'm misunderstanding what you're saying, I think the system I'm describing would requre either a bot or a diligent user(s) to do the rollover at midnight and the monthly mass page creation would just be a matter of convenience. — GT00:20, 16 May 2006 (UTC)
I'm talking about the rollover at midnight. See for example {{day+1}}. If we could tell the system to use day-18, day-17, or something like that, it would automatically change the pages, although I don't know if this is possible. FetofsHello!12:46, 16 May 2006 (UTC)
I am aware that this proposal was not popular here, but I would like to request one to run on the Commons. The Commons does not have such a strong community as en.wp and it is quite important that we keep trying to promote essential information such as licensing requirements, to save both users and admins a lot of hassle. I brought up this idea on commons:COM:VP and it was well-received.
On the 14th May we had 163 new users register; on the 13th, 185. So probably 150-200 is a typical daily number. The possibility to restrict welcomes to users who have made at least one edit or one upload would also be welcome.
If you get it successfully working, would you mind to leave me a note too? We've discussed this on Wikiquote, and if there is a working copy of one, I'd like to be able to offer it for demonstration. Essjay (Talk • Connect)14:39, 21 May 2006 (UTC)
Here is the tasks of my robot: Make Category, Raplace a letter with another, Make redirect, Interwiki (Especially Farsi language), Disambiguation pages, etc. --MehranVB 18 May 2006 (UTC)
If you plan to create and run this bot yourself, you should ask for approval on Wikipedia:Bots/Requests for approvals rather than here; in both cases, you should probably be more specific, especially since several bots already perform these kind of functions. Schutz11:44, 18 May 2006 (UTC)
Hi, is there already a bot for adding {{linkless}} to all pages on the long, long Special:Lonelypages list? If not, could someone write one? It would not have to run all the time, just for a short time each time the Special:Lonelypages page is regenerated. This would be beneficial for Wikipedia: the Siegenthaler page only lasted as long as it did because it had almost zero links in. When someone sees an article with {{linkless}}, though, they will be encouraged to add links to it from elsewhere. Cheers, --unforgettableid | talk to me22:59, 24 May 2006 (UTC)
It seems to me that Jimbo should have taken this excellent idea as the solution to the Siegenthaler issue rather than the restrictions he decided upon. Just a thought. — Stevie is the man!Talk | Work23:32, 16 June 2006 (UTC)
I am currently doing that with User:MarshBot, but I have been swamped lately and haven't run it (eventually I'll get a bot flag for it so it's easier to do). Since it ignores pages with that template already on it, I have no objection to other people running similar bots/AWB sessions. You will want to avoid disambiguation pages, {{deleted}} pages, and some others that should NOT have inbound links, or should have only very specific ones. --W.marsh23:38, 16 June 2006 (UTC)
This idea needs a bit more work yet. For example it doesn't cover the no-sources and the no-licenses and the bad-licenses. It will also affect images that are decided to be kept. I think there needs to be more discussion about whether or not it would actually help, before implementing it. pfctdayelise (translate?) 23:33, 25 May 2006 (UTC)
Adding a new image can always be rolled back; deleting an image can never be rolled back. What are you talking about, not helping? Right now, no sister-project has any notification that something is on the chopping block. Once the image is deleted, there is no simple way of getting to Commons, where it was deleted. --Connel MacKenzie07:42, 26 May 2006 (UTC)
For a start uploading over an image in no way 'notifies' the projects where it's being used. In fact rather dangerously, they get no notification at all. Weeks could go by where they are displaying ugly crosses completely unaware. Is that going to help? Won't it look really unprofessional? Shouldn't we really get permission from all the projects that use Commons before we implement this? I'm not saying we should do nothing, but I think the m:User:Duesentrieb/CommonsTicker is going to be way more helpful and relevant, that's all. Anyway, I hope we can continue this on the commons:COM:VP?pfctdayelise (translate?) 09:11, 26 May 2006 (UTC)
I can see your reasoning for wanting approval on each project, but it is seriously flawed. Right now, projects are assulted by images dissapearing, with absolutely no notification whatsoever. The big red X concept is worlds better, as simply everyone who sees it is given an indication that there is a problem, and an avenue to comment about it. That is an order of magnitude nicer than leaving broken (null) image links everywhere. Even if CommonsTicker worked, it would still not give indication to anyone (e.g. sister-project sysops/janitors such as myself) who might be inclined to do something about the situation. --Connel MacKenzie23:25, 26 May 2006 (UTC)
Do we have an existing bot that checks for the existence of hyphens in spelled-out fractions? It would look for common word pairs I guess like "one third", "two thirds", etc. and add in the hyphens between them.--Hooperbloob05:40, 26 May 2006 (UTC)
In 2001, Justin Wilson wins the FIA International Formula 3000 Championship with three wins, six second places, one third place and two pole positions - he broke the records for the most points in an F3000 season and the most podium finishes.
Looking through the Wikipedia:Most_wanted_articles page it struck my attention that there are really a lot of requests for the 19xx Pulitzer Prize or Oscars. Now, for the Pulitzer at least, the data are available at http://www.pulitzer.org/. Would it be possible to make a bot to do the drudge work of capturing the data from there and wikifying it? Copyright is ever troublesome, of course, but the format of the existing 19xx articles does not seem to vary significantly from the hypothetical output of such a bot. - Pthag00:15, 31 May 2006 (UTC)
Lists of facts, such as winners of an award, would not be copyrightable. Commentary in paragraph form would be, obviously. — Jun. 2, '06[17:48] <freak|talk>
I deleted Image:Perfecto.icon.svg (redundant with Image:Flag of Canada.svg). Someone was using it in their signature... so I don't know if it's worthwhile replacing all of the link references or deleting them all. If nobody wants to set a bot on them, I don't see a problem with leaving them alone. ~MDD469601:24, 1 June 2006 (UTC)
Personally I think this sort of thing us better to run on a direct database query but that means troubling someone w/ shell access -- Tawker04:35, 1 June 2006 (UTC)
There's only 3 pages including this one linkig there now, so I huess we can forget this one.RichFarmbrough 18:57 16 June2006 (GMT).
A few days back I found a featured picture that did not have a {{FeaturedPicture}} template on it. Thinking as a one-off case, I added the template. Today, I found another such image. This made me think that there might be many more such cases. Can we have a bot which will do the following:
Check whether all the articles/pictures/lists/etc. featured on Wikipedia (as indicated by WP:FA, WP:FP, etc.) have the required templates on them.
Check whether all the articles/pictures/lists/etc. that have the featured template on them are really featured on Wikipedia (Rare, but it might be possible that some vandal can do it and escape undetected).
I actually already have a bot, approved without a bot flag, NedBot. However, I'm still learning how to use phyton, pywikipedia, regex, and all that fun stuff. But I've been pretty busy with work and all, and so far have just used the basic text replacement and category functions that are included with pywikipedia.
I would be really greatful if someone could help me get the bot to place a message at the top articles found in a category, and possible in targeted section as well. The bot was approved for providing assistance to WP:DIGI, and we wanted to add a message to editors to help combat a large amount of anon users who were adding unsorted information. Something to the extent of <!-- Please cite character attack blah blah, such information without a source could be deleted at any time-->. (the exact message, if you are interested, is being discussed on our WikiProject's talk page) Any help would be greatly appreciated. -- Ned Scott05:18, 7 June 2006 (UTC)
At the top of the articles? You would just get the text, save it in a variable and add the text you want to it. Something like:
It seems you already got the most difficult part sorted (looking into the category). If you want to place it in the top of a specific section, use the replace tool:
Is there a bot for use in the field of chemistry? One that can place and edit chemboxes? If not there should be, this is getting really tedious... - jak(talk)21:04, 11 June 2006 (UTC)
What information is available? It would be possible to do something along those lines. RichFarmbrough 23:17 11 June2006 (GMT).
Notification to editors of Wikipedia entity up for deletion
As an editor of some articles and other things that got put up for deletion, I think it would be great if there was a a "courtesy bot" that informed past editors of the thing. Perhaps the bot would insert a post on the editor's talk page. Note: I don't think watching is a good solution, because the decision to watch something isn't (in my case) related to all articles I care about, but rather ones I feel are especially critical to watch at any given time. — Stevie is the man!Talk | Work16:48, 14 June 2006 (UTC)
This might be semi-practical, if:
The alerts are kept short and sweet, not some huge colorfully formatted template.
It only notifies people who have asked to be notified.
It ignores anonymous edits, minor edits, and edit summaries which suggest it's merely a vandalism revert.
It ignores blocked accounts whose block is not expected to expire before the VFD closes. A large portion of these will be accounts whose only edits to the page were vandalism.
All of those points are well-taken and IMHO, agreeable. The second point sounds like it would require a new preference setting, but is that normally done for bots that are not built into the wiki software? Perhaps there should a generic setting where the user decides whether bots can leave messages on their talk pages. The third point probably would be difficult to perfect as far as anything outside of anonymous and minor edits. But if somebody can handle parsing out posts to determine which are merely vandalism reverts, all power to them. — Stevie is the man!Talk | Work19:53, 14 June 2006 (UTC)
Another thought: Instead of requiring a change in user preferences, another approach could be to have a "registry page" specific to the bot for those who want to sign up for such notifications. That is, if you've put your username on that registry page, the bot will notify you of any deletion efforts being established on articles/etc. you have edited. — Stevie is the man!Talk | Work21:25, 14 June 2006 (UTC)
Yerse, some of ous don't really watch out watch lists anyway. RichFarmbrough 18:19 15 June2006 (GMT).
Wow, before you go any further, you might look into m:ParserFunctions. The {{expr:}} notation might make this task a lot easier. — Jun. 14, '06[18:35] <freak|talk>
I think this gets proposed a lot... and it kind of defeats the purpose of a personal welcome. Also it would welcome a lot of vandals, which could confuse RC patrollers. And considering the number of accounts, and a 1k message, well it would only take up about a gigabyte a year, but still... it would be 1,000,000 edits/year more and that's a lot of server load in actually leaving the messages. I believe various helpful links appear as you create an account anyway. --W.marsh22:40, 16 June 2006 (UTC)
It's really impersonal for a bot to do this... if it's a person, it means that someone noticed your edits and took the time to say hello. In reference to W.marsh's comment, I see a lot of people welcoming vandals or even people with no edits already! It just makes me shake my head... ~MDD469623:39, 16 June 2006 (UTC)
I think that a WelcomeBot could complement, not undermine, the personal touch to welcomes. Typing something like:
{{subst:Welcome4}} --~~~~
isn't exactly very personal. After the WelcomeBot has left its inital message, a user-left welcome would probably look more like this:
"Hello and welcome to Wikipedia! Feel free to [[User talk:32902304892|drop me a line]] if you have any questions. --[[User:32902304892|32902304892]] 06:06, 06 June 2006.
As for welcoming users with no edits... why not? A welcome message pointing them to useful pages will encourage them to contribute and hopefully improve the quality of their inital contributions. Also, I created an account, and not a single one of the Welcome message links popped up. If there's a definite "no" consensus about a WelcomeBot, we should at least add some key links to the account creation process. --Alex S00:13, 17 June 2006 (UTC)
How to distinguish between an account and a user? We have the best part of 2 million accounts, most of which I suspect will never be used again, either they're nonce accounts or vandal accounts. "Wasting" over 1 million edits and the associated growth of the databse and dumps, seems like a bad idea. RichFarmbrough 10:24 23 June2006 (GMT).
It would be better to implement this in the MediaWiki software. On every user talk page that does not exist, it would be automatically redirected to a template, although the redirect will not be shown on screen. That deals with the database size problem. When the page is edited, the initial wiki code text would be loaded directly from the template also. Upon saving, the page is created and it does not redirect anymore. A new user will see a "You have new messages..." screen when they first create their account. This is similar to a Welcome message as the first message in a new E-mail box, although the text is not duplicated. Invitatious16:38, 1 July 2006 (UTC)
Replacing single digits with their written equivalents
Doing a text search for the phrase " 2 types" I found numerous cases where the digit should be spelled out (for numbers less than ten (eleven?)). If a human-assisted bot could include searches for " 'n' types" and " 'n' categories" this would help. If there are other grouping names that could be included that would be great, I've only considered of those 2...er. two! --Hooperbloob05:25, 22 June 2006 (UTC)
Kia Ora, I was wondering if there is much chance of someone getting a bot thingy to change the uses of "Maori" to "Māori", which is more correct. Many thanks. --Midnighttonight10:24, 22 June 2006 (UTC)
Someone changed "Sasha" into a disambig page which broke like a hundred links. Can someone run a bot to change all links that currently point to "Sasha" to now be piped links to "Alexander Coe"? Thanks a bunch! Wickethewok16:02, 24 June 2006 (UTC)
Bot requested to replace Image:Vatican coa.png with Image:Emblem of the Papacy.svg
Could some please program a bot to replace all instances of Image:Vatican coa.png (which is being used under a "fair-use" claim) with Image:Emblem of the Papacy.svg (which is public domain)? There must be well over a hundred pages (including categories, where fair-use images have no business anyway) it's being used on. Thanks. User:Angr17:40, 18 July 2006 (UTC)
I didn't really see a point in doing this by bot, so I have completed your request manually. 208 images replaced. All done. Alphachimptalk07:06, 19 July 2006 (UTC)
The Mediation Committee has for some time been interested in a bot to open new cases and a few other tasks. We had arranged with one coder to program the bot, but the project fell aside, and we're now interested in finding someone with the time to program it for us.
A few requirements:
The person should be fairly patient, as I will be the one setting it up, and I am by no means a programmer. Something fairly easy-to-install and run would be greatly appreciated.
It should be something that can be run on Zedler (the toolserver); my preference is to have it execute hourly via crontab, do what it's supposed to do, then terminate until the next run is started by crontab. I want to make sure it's not going to be running in the background somewhere, or have multiple sessions of it going at once.
Would prefer it be something that could easily be expanded as the Committee determines new tasks, preferably something that another programmer could work with if the original one is no longer available.
The basic task is opening cases on RfM. When activated, the bot should search the pages transcluded to Wikipedia:Requests for mediation/Pending for pages containing {{Accepted case}}. (If doing so is difficult, I'm open to listing cases in another spot for the bot to check, in addition to /Pending. I'd really prefer that the template be the activator, but another similar option is acceptible.) If it doesn't find the template, it should shut down and do nothing until the next activation by crontab. If it finds that template, it should:
Add an entry to Wikipedia:Requests for mediation/Open Tasks in the "Unassigned" section. The format for entries is {{Wikipedia:Requests_for_mediation/Open_Tasks/U|Case name (subpage name)|date accepted (~~~~~)}}.
Notify all the parties listed in the section of the subpage marked "Involved parties." The parties should always be listed with {{user}}, and I can make a special point to be sure they are correctly done. The notification should be a simple subst: of a prepared text located at User:MediationBot/Accepted case. The bot will need to put the case name in as a parameter, the exact formatting can be easily changed.
That is the basic task set for the bot. We would also like to be able to:
Notify parties that a case has been assigned to a mediator, closed, reopened, etc. This would be a simple subst: of a prepared text, and would not even require the bot to include the case name as a parameter. This should be something that could be hand-activated, rather than being done by crontab. I'm open to ideas from the programmer on the best way to notify the bot that this should be done, and where to put the text.
On specific occasions (very rarely), send notifications to all Committee members. This would always be a simple subst: of a prepared text. A full list of current mediators is always maintained at Template:Medcom, so that could be hardcoded. As above, I'm open to suggestions on how to accomplish this in the most user-friendly and least-programmer-time-exhaustive way.
If it would be easier to code two separate bots, one to do the regular case opening and a second to do periodic notifications (I know there are existing notification bots, like the Signpost's bot and the Esperanza newsletter bot), then we can have MediationBot and MediationBot2 or something.
Anyone with the requisite experience to code the above, and the patience to deal with a non-programmer, is invited to contact me directly via my talk page. This has been approved by the Mediation Committee, and is being undertaken as an official Committee project. If successful, we would like to be able to offer a similar bot to our sister committee, the Arbitration Committee, obviously customized to thier specific needs, and subject to thier desire to use it. For the Mediation Committee,Essjay (Talk • Connect), Chairman, 12:16, 11 July 2006 (UTC)
In the interests of avoiding duplicated efforts, I wanted others to know that Tangotango is currently working on this. If others are interested in assisting, please contact either of us. Essjay(Talk)13:59, 13 July 2006 (UTC)