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 talk page.
About 20% of the 2200+ categories in
Wikipedia:Database reports/Categories categorized in red-linked categories are year-related (sort by Member category to get an idea - the three pages have different biases, page1 has a lot of 1st millennium dates, page3 has the most in total). Some of these cats have been red links for 2+ years, but are potentially quite amenable to creation by bot and would allow human editors to concentrate on the more demanding categories. I'm thinking of pseudo-code along the lines of
Scan the right column of
Wikipedia:Database reports/Categories categorized in red-linked categories and reject everything that doesn't have a year (including BC years) - say between 1000BC and (CURRENTYEAR+10) - quite a lot of cats deal with the near future, upcoming Olympics, elections etc. Decades, centuries and millennia would be nice too.
Check to see if category has been created - the report is normally created on Sunday morning, so a weekly run soon after this would be optimal
Any non-existent categories named "(dis)establisments in" - check against a list of countries (eg the ones in ‹The
templateCategory link is being
considered for merging.›Category:21st-century establishments by country ) and US states and if the place is on the list, create the category with the following templates - note how the decade ones require you to calculate the century :
1970s establishments in Ruritania has {{estcatCountryDecade|19|7|20th|Ruritania}}
1970s disestablishments in Ruritania has {{disestcatCountryDecade|19|7|20th|Ruritania}}
1979 establishments in Ruritania has {{estcatCountry|197|9|Ruritania}}
1979 disestablishments in Ruritania would have {{disestcatCountry|197|9|Ruritania}}<nowiki>
**'''1970s establishments in New Jersey''' has <nowiki>{{estcatUSstateDecade|19|7|20th|New Jersey}}
1970s disestablishments in New Jersey has {{disestcatUSstateDecade|19|7|20th|New Jersey}}
1979 establishments in New Jersey has {{estcatUSstate|197|9|New Jersey}}
1979 disestablishments in New Jersey has {{disestcatUSstate|197|9|New Jersey}}
I think there's scope to apply some fuzzier logic on the cases of "nnnn in foo", "nnnn foo" or "nnnn–nn in foo" (sports leagues mainly). Since the 2012 equivalent will almost certainly exist, just grab the categories and templates from the (CURRENTYEAR - 1) version of the category, adjust the numbers accordingly, and use that content to create the category. It may not be quite perfect but I wouldn't let perfection be the enemy of the good, at least it gets it into the hierarchy where it is visible to the subject experts, any errors in the year hierarchies tend to be very visible - and to be honest duplicating the 2012 caetgory actually seems to work pretty well as a heuristic when doing it manually, the error rate should be acceptable. Maybe have a "handbrake", to base no more than three new categories on a 2012 category without asking a human to doublecheck that it's OK?
Then check that there's no category red links on what you've just created - it'd be nice to create those in turn, although they could wait for the next run of the report.
Something to watch for in the above is the sorting - BC dates are assigned negative numbers for sort purposes, so eg 270 BC will be given -30 in a 3rd-century BC category. — Preceding
unsigned comment added by
Le Deluge (
talk •
contribs) 18:50, February 18, 2013
Discussion on whether this is a good idea (conclusion: yes it is)
Comment on the above. This is basically sound and getting help from a bot would be nice. However categories for nnnn (dis)establishments in Ruritania and even categories of the form nnnn in Ruritania are somewhat controversial when nnnn is a year in which Ruritania didn't exist. CfDs on that topic have been inconclusive. Le Deluge believes (and it's certainly sensible) that the categories should be created anyways and discussed at CfD down the road. I believe the categories shouldn't be created until a clearer consensus emerges from CfDs. (See
our conversation on that topic) That being said, if someone is willing to write the code, I'd be open to a change of mind if the trial run is convincing.
Pichpich (
talk)
14:05, 19 February 2013 (UTC)
Pichpich has more experience of this but my impression is that the use of modern countries in years when they didn't exist is not a huge problem - after all the bot would not be making this call independently, it's merely creating a parent category when a human being has created a daughter cat like Companies established in Ruritania in nnnn. It would be easy to program the bot with a couple of exceptions - I would guess that just avoiding "in the United Kingdom" before 1707 and "in the United States" before 1776 would get the debatable cats down to acceptable levels, obviously you would throw in some of the other obvious ones like Germany pre-1871, Italy pre-1861, Yugoslavia pre-1929 and the Soviet Union pre-1922 whilst you were about it.
Le Deluge (
talk)
14:35, 19 February 2013 (UTC)
If you can't pontificate when you're old, when can you? <g> Seriously, it's a perfectly reasonable point you're making - it really just comes down to what kind of error rate is acceptable. Obviously 100% perfection is ideal but seldom achievable, you just want the error rate to be manageable and the errors to be visible. With these categories the mistakes are very visible, so that's good. I thought it was worth some analysis to see what we were looking at, bearing in mind that the "problem" countries will be over-represented in the list as people like Pichpich have chosen to work on "easier" categories. I took the full list of 2227 missing categories, converted the first letter to lower case, and then rejected everything that didn't have at least one upper case letter and one digit in it. I then rejected anything that whose only upper cases were "BC" (rejects eg586_BC_in_politics) and which didn't contain "_in_". That left 289 categories.
Of these 52 mentioned one of the
50 main US states and 132 were either in the
list of UN members, a plausible common name thereof (Bolivia, Democratic Republic of Congo, (East Timor), Iran, Libya, Moldova, (North Korea), South Korea, Syria, Tanzania, United Kingdom, United States, (Vietnam)), or a region (Africa, Asia, Central America, Europe, European Union, Middle East, North America, Oceania, South America, (Balkans), (Caribbean), CONCACAF,(UEFA)).
Dependencies accounted for another 9 (Curaçao and Dependencies, Isle of Man, Jersey, Réunion, United States Virgin Islands) plus 1 more for Western Sahara. Eight former entities accounted for 16 categories (Austrian Empire, Austro-Hungarian Empire, British Empire, Ottoman Empire, Korea (1898_in_Korea,1994_in_Korea), Persia, Prussia, Yugoslavia) and fourteen "
colonies" accounted for 48 cats (Dahomey, French Cameroons, French Sudan, French Upper Volta, Kamerun, Malaya, Mandatory Palestine, New Spain, Republic of Upper Volta, Rhodesia, Tanganyika, Thirteen Colonies, Western Samoa Trust Territory, Zanzibar). There were 9 cats from five non-US subdivisions (Catalonia, England, Northern Ireland, Nova Scotia, Sikkim) - former countries that became states of the US,Canada,Aus,India need a bit of thinking about, I'm happy to flag those as risky. 19 cats were national adjectives (2018_in_American_politics, 1990–91_in_Spanish_basketball etc). The only cats left over after that were 2000s_in_Birmingham & 21st_century_in_Birmingham, which were created in error by someone who didn't know what they were doing (and which I've now taken care of).
I don't want to put anyone off - if anything the above has made me a bit happier that the "Ruritania" aspect of the bot is achievable and can be made to be pretty clean, because the UN members and NSGT lists give us the basis of a whitelist of acceptable Ruritanias, which can be amended with a relatively small number of common names, regions, US states and former empires/countries/colonies. I was worried that most of the potential categories would be things like Indian princely states and oblasts on the fringes of Russia, but the above makes clear that most of the categories involve "mainstream" countries. A basic whitelist of countries will help out with a few 100 categories in the first place, and in time it could be expanded to cover things like national adjectives. If the bot was coded to cope with a lookup table of cutoffs like "No UK before 1707" then I don't mind doing a more comprehensive list of cutoff years per nation. I'd start with year of UN accession from that list and then backdate the older countries to foundation.
Le Deluge (
talk)
21:04, 19 February 2013 (UTC)
Alright, there's nothing like good data to settle a debate. I suppose I am being too conservative and a well designed bot would do way more good than harm.
Pichpich (
talk)
21:49, 19 February 2013 (UTC)
Billboard URL repair
Billboard has revamped its site, and we have tens of thousands of dead links. The old link looks like
http://www.billboard.com/artist/<artist name, urlencoded>/chart-history/<magic number>?f=<chart number>&g=Singles with "#" inserted at random locations.
All the magic numbers for the artists changed. The artist's names stayed the same, even though the formatting has shifted. The chart numbers remain the same.
I've done the crawl of Billboard to find the new tokens. The results are now in templates. {{BillboardID}} will return the appropriate number for the artist. {{BillboardChartNum}} will return the chart number from a chart name (which should stay more stable). The new template {{BillboardURLbyName}} will take that data and return a correct URL. The purpose of the template is to keep from having to go through this again when Billboard revamps again. It seems to happen every few years.
For size considerations, {{BillboardID}} is actually broken into 40 separate templates, broken by first character. Take a peek at {{BillboardID/Q}} and {{BillboardID/R}} and it will be obvious.
So, what the bot needs to do:
For each URL of the form http://www.billboard.com/artist/<artist name, urlencoded>/chart-history/<magic number>?f=<chart number>&g=Singles or http://www.billboard.com/#/artist/<artist name, urlencoded>/chart-history/<magic number>?f=<chart number>&g=Singles
extract artist name and chart number
if (artist name not translated by BillboardID)
then log error and skip
else if (chart number not translated by BillboardChartNum)
then log error and skip
else replace URL with {{
BillboardURLbyName|artist=<artist name>|chart=<chart name>}}
When the bot is done, it should provide a log of every time it found an artist that it couldn't handle or a chart that it couldn't handle. I'll take those errors and fix the templates to handle those cases, and we can rerun as necessary.
An excellent idea. Much easier than doing it all by hand. Doesn't seem like a very hard bot to code. If someone can do it, please! This is a big issue we are currently facing. Status01:07, 20 February 2013 (UTC)
I decided to dive in, and I'm at the 95% complete stage on this bot. I'll ask again if I need help crossing the finish line.—
Kww(
talk)
01:58, 21 February 2013 (UTC)
If the page is to work as intended It would need to be checked every 30 mins or so, which unfortunately cluebot does not do. Although adding the cluebot config the the page wouls still probably help the current state! ·Add§hore·Talk To Me!03:51, 21 February 2013 (UTC)
There are hundreds of articles, mostly about settlements in Brazil, that specify 'language=Portguese' as a citation template parameter. There are too many to correct to 'Portuguese' by hand, but it would be an easy job for a bot.
Colonies Chris (
talk)
22:09, 19 February 2013 (UTC)
I am going to file a BRFA with an argument that I believe is good enough and will also run it semi-automatic if required. -- Cheers,Riley02:22, 20 February 2013 (UTC)
Request: a very quick search-and-replace in ~250 articles
Over 200 articles that cite the Cancer Dictionary all need the same trivial change to the links to the original source to make them right, as the base URL, but not the article codes, for those entries has been changed on the cancer.gov website. Example: in
Peritoneal mesothelioma, the link to
I could easily fix this myself with my own bot in a few lines of code, but don't want to have to go through the process of obtaining permission to do so for such a small task: if someone with an already authorized search-replace bot is an a position to do this without putting themselves to any great effort, and would be kind enough to help, I'd be very grateful.
Alternatively, if this is such a small task that it would OK to just to this without going through the usual bot-task process, I'd be happy to do it myself.
I've recently discovered that AnomieBOT will create conflicting references when there's a citation error in an article and there are references that are created by {{singlechart}}. Note
this edit, for example, where AnomieBOT's response to a broken reference named "Hungary" was to treat the reference "Australia" as orphaned, even though the Australia reference was fine. Given the discussion at
User talk:AnomieBOT/Archive 5#breaking references, I'm not expecting a fix from AnomieBOT any time soon. To control the damage it's causing, I need to get a list of articles that are in ‹The
templateCategory link is being
considered for merging.›Category:Singlechart making named refand have a colliding reference defined both through <ref name=xxxxx> and through {{singlechart|...|refname=xxxx}}. Hopefully one of the existing bots with a good search capability can do that for me. I'll do the fixes manually, but I don't want to try and trawl through 450 articles manually searching for something that doesn't even give an easily visible error message.—
Kww(
talk)
07:00, 23 February 2013 (UTC)
That looks like the list I need. Thanks. I'll take it from there, and Anomie did agree to insert code to keep the problem from spreading.—
Kww(
talk)
15:10, 23 February 2013 (UTC)
For generating an HTML page with current category structure
We on Sanskrit wiki want a bot to generate an html page to contain all our categories in form of html tree-view control,
something like this. I know that there is a special page on wiki for category viewing,but the problem is that it doesn't work readily (loading time is required on each click); while if we can download a single html page, we can view the category structure easily by collapsing & expanding as needed, even offline, and i want to review the category structure using that method. We have manageable number of categories on sawiki. So please tell if that is feasible. -
Hemant wikikosh (
talk)
11:58, 17 February 2013 (UTC)
Featured pictures are (with one or two exceptions) stored on Commons but have a local page that links to their FP nomination page and indicates the date they were
picture of the day, if applicable. Sometimes, however, the FP templates are replaced due to vandalism, and then the pages are deleted per
WP:CSD#F2 (
example 1;
example 2). These bother me and I would like to fix them; those local templates are the primary advertisement the FP project gets. I find these at random, but it seems like it would be fairly straightforward to write a script that would list all FPs with no local page, just by going through the subpages of
the directory (one way might be to compare images linked in the directory to
Category:Featured pictures; if they're not in the category they're missing the local page or its templates). Piece of cake, right? Thanks.
Chick Bowen19:14, 25 February 2013 (UTC)
Update languages refs
{{Ethnologue}} has been used as a shortcut to {{Ethnologue16}}. However, with the publication of the new edition of Ethnologue today, this is no longer appropriate. Please convert all transclusions of {{Ethnologue}} to {{Ethnologue16}}, then delete {{Ethnologue}}.
Create with the simple template "{{Sockpuppet category}}"
That should be good enough - if you read {{Sockpuppet category}} you can see that there's a bit of scope to get cute about encoding certain non-alphanumeric characters but the template is smart enough to sideline "problem" names into a maintenance category so it's not very necessary. It's a simple enough little task that could be done manually with AWB, but it might as well be automated, sockpuppet cats make up 20% of those red-link categories and anything that helps out the anti-sock guys is worth doing. — Preceding
unsigned comment added by
Le Deluge (
talk •
contribs) 18:50, February 18, 2013
The consensus
here was for "a bot that blue-linked all redlinked categories that were named after an account and contained at least one other named account, and deleted tags from any IP account that was associated with a red-linked category after that creation was completed." Since this is beyond my technical skills, I have withdrawn my bot request. I hope another bot operator can pick up this request.
GoingBatty (
talk)
01:45, 26 February 2013 (UTC)
At
Talk:Main Page#Today's article for improvement on the Main Page, there appears to be consensus (pending an uninvolved party's discussion closure) to proceed with the addition of a
TAFI section to the main page, with the dynamic display of three article links from a pool of ten (
example). A concern is that the links, generated via {{random subpage}}, change only when the page's cache is purged. So I'm requesting that a bot purge
Main Page's cache with whatever frequency is feasible and acceptable (once per minute, perhaps?). —
David Levy18:27, 21 February 2013 (UTC)
The server admins would have the operator's head on a pike before you could blink. The main page is cached for a reason, it is the most viewed page on wikipedia. Breaking the cache system that often for something that trivial would cause some backlash.
Werieth (
talk)
03:05, 22 February 2013 (UTC)
Right. Just display 3 articles at a time, and then switch them out whenever DYK gets updated or something. Don't dynamically change them.
Legoktm (
talk)
03:14, 22 February 2013 (UTC)
For the reason noted below, the community approved the section's addition on the condition that the article links be randomly pulled from a large pool. (For the record, I wasn't involved in that discussion, but I would have expressed the same concern.) —
David Levy03:39, 22 February 2013 (UTC)
I understand the importance of caching, particularly when a page is viewed millions of times per day, but I'm unclear on how the proposed setup constitutes "breaking the cache system" or poses a problem. In terms of overhead, how would it differ from any other cache purge (a common occurrence across the encyclopedia)? In the subsequent minute, whether a page is requested once or 7,000 times, isn't the same cached version being sent? (Please forgive me if I've misunderstood how this works.)
Note that we already include a purge link on the main page. The TAFI proposal previously called for an additional one to be included in the new section (with readers encouraged to use it repeatedly), which probably would have resulted in significantly more than one purge per minute.
Also note that the purpose of varying the links is to avoid sending too many editors to articles at the same time (thereby causing endless edit conflicts and potentially driving away new contributors), which isn't a trivial matter. —
David Levy03:39, 22 February 2013 (UTC)
What did we do for the 2008 election FA? That was two articles displayed together, with the order rotated randomly; presumably we hit the same problems then, but we may also have solved it :-).
Andrew Gray (
talk)
09:44, 22 February 2013 (UTC)
That relied on JavaScript code, which
failed gracefully for users without JavaScript enabled (who saw the blurbs in a static order). I don't know whether something similar is feasible in this instance. —
David Levy11:57, 22 February 2013 (UTC)
Does a cron job which requests http://en.wikipedia.org/?title=Main_Page&action=purge every hour, say, really require bot approval? If it's not editing, it doesn't require a bot, does it?
Neo Poz (
talk)
19:09, 23 February 2013 (UTC)
Probably not, technically, but I did go through the process for
User:Joe's Null Bot, which does roughly four times that many purges per day. *shrug* The original request was for one per minute, not once per hour, of course. --
j⚛e deckertalk19:54, 23 February 2013 (UTC)
How do you feel about adding the 1/hour to Joe's Null, and if so going though the approval for the addition or not?
Neo Poz (
talk)
20:08, 23 February 2013 (UTC)
"Once per minute" was merely a suggestion. I've requested "whatever frequency is feasible and acceptable", meaning that which can be programmed without inconveniencing the bot's operator or causing any server-side problems (though I suspect that only a ridiculous rate would have the latter effect).
The page needn't be purged once per minute, but I hope that it can occur more often than once per hour (as that seems long enough for the articles to be flooded). —
David Levy20:36, 23 February 2013 (UTC)
Just as a question - does this HAVE to be done by something hard-coded that gets refreshed periodically? Since the aim is just to present 3 out of 10 links, could you not slice it by some other way than by a cache-purge? I'm thinking of using Javascript to do something like If CURRENTTIME-in-milliseconds ends in 1, show links 1,2,3; if CURRENTTIME-in-milliseconds ends in 2, show links 2,3,4 and so on. Doesn't have to be done on time either - it could be done on the ASCII value of their user name, sum of their IP address (with a bit of help server-side), whatever. It seems a better way of doing something that is fundamentally quite simple, rather than messing with the caching on such a heavy-traffic page.
Le Deluge (
talk)
19:46, 25 February 2013 (UTC)
My concern is that not all users have JavaScript enabled. Assuming that the code were to fail gracefully (as in the instance discussed above), perhaps the number of affected users would be acceptably small.
But unless I've misunderstood the caching system, the fact that
Main Page is "a heavy-traffic page" is irrelevant. A cache purge causes the page to be rebuilt once, after which the same version is sent until the cache is purged again. Whether a page is requested zero times or 100,000 times in the following 15 minutes, it's been rebuilt only once. So the cost of purging
Main Page's cache should be no greater than that of purging any other page's cache. A 15-minute interval would result in 96 cache purges per day, which is negligible. —
David Levy20:23, 25 February 2013 (UTC)
Usual assumption is that about 1.5% of web users have no Javascript - in this case one might assume that the percentage of likely editors is less than that, but we'll go with it. If your NOSCRIPT page just presented people with links 1,2,3, that would mean that 11.35% of people see pages 1,2,3 and 9.85% see each of the other combinations. I don't think that's unreasonable. If you had 3 out of 100 links, then you might set aside one combination of links specifically for the NOSCRIPT people. As for the caching, you're need to think like a user - or rather like the cache nearest the user, rather than at the Wikipedia end. The fact that the en.wiki main page is so popular means that almost every cache on the planet will need a current copy. Let's assume one of their users requests it once every minute for the sake of this argument. If MP is purged once an hour, then 59 users can be supplied with the version on the cache, and only one has to wait for the cache to request the new version up through the hierarchy of other caches until a request gets made to Wikipedia. Purge it once a minute and all 60 users think "Gee, Wikipedia is slow today". OK, an extreme version. The real problem is the extra network traffic you're incurring, if this is a popular page it will be on 100,000's of servers throughout the internet. Compare that with say the main page of Welsh Wikipedia, where not only are the numbers much smaller, the likely users can all be "fed" by getting a new copy to a few servers in places like Cardiff (and one in Patagonia). A cache in Hong Kong or Romania isn't likely to need a new copy of the Welsh page every minute. Apologies to any network engineers out there, but you get a flavour of some of the argument.
Le Deluge (
talk)
21:42, 25 February 2013 (UTC)
I don't recall the issue of third-party caching arising in this context before. I'd be interested in reading others' opinions on whether this is a major concern.
Either way, I'd like to see a working example of the JavaScript-based approach, which seems like it might be a better solution. —
David Levy23:00, 25 February 2013 (UTC)
Well I have the one line cron job ready to go, and am thinking about being bold about it, but I would much rather Joe add the 15 minute interval to Joe's Null Bot. Joe?
Neo Poz (
talk)
22:13, 25 February 2013 (UTC)
Hey, sorry I haven't seen this. I'll drop a request at BRFA in the next 12 hours. I could get hit by a bus too, but I'm happy to at least ask BAG what they think about a 15 minute purge. That sounds very impact to me, but let's ask the gurus. --
j⚛e deckertalk00:12, 26 February 2013 (UTC)
Wow, I meant "low-impact". What an embarrassing omission, my apologies. My intuition is that the cost of rerendering an uncached page every few minutes is quite minimal, but I'd be lying if I claimed any detailed operational experience with server-side Wikipedia. Sorry for the lack of clarity, I'm going to spend some time writing, then stepping back, then reading, then rewriting the BAG request, to avoid any similar lapses of clarity. --
j⚛e deckertalk01:08, 26 February 2013 (UTC)
No worries. It's just a minor typo. :)
My intuition is the same, but I'm not a server expert either, and I certainly agree that it would be helpful to consult those more knowledgeable in that area. —
David Levy01:43, 26 February 2013 (UTC)
See the "Category:Wikipedia files with no copyright tag" section of the
current revision of WP:AN. DumbBOT persists in creating something that was once useful but has now been made useless, and because the operator is AWOL, there's no way to stop the bot, as far as I can tell, except for the horribly unhelpful method of blocking it. Could someone please produce a replacement bot? Its functions are:
removes protection templates from non-protected pages
One problematic category isn't a big deal, but if we make broader changes in the future, this bot may start having serious issues; imagine if the AFD nomination process were have a fundamental change in its page structure, for example. Basically, the only reason I'm requesting this is so that we'll again have a bot doing this under the control of an operator who's active here at Wikipedia, so that changes could be made to its operation if necessary; if you create such a bot, we can softblock DumbBOT until the operator returns, which probably isn't going to happen.
Nyttend (
talk)
16:31, 27 February 2013 (UTC)
You could just add the category title to the title blacklist, and see if there is any negative fallout. Probably the easiest thing to do.
Werieth (
talk)
18:39, 27 February 2013 (UTC)
Recently I found two articles in the same category, which were sorted differently. After my question on WP:HD
User:John of Reading fixed them
[1], however there's still plenty of mis-sorted bigraphies in
Category:Icelandic people and its subcategories. Possibly some bot, when supplied with a list of categories (or just searching them by regexp 'Iceland|Icelandic'), might correct such biographies sorting, just by picking the DEFAULTSORT key, swapping it, dropping a comma and adding it explicitely to appropriate categories' links. Of course it must also recognize biographies among other articles. --
CiaPan (
talk)
12:16, 1 March 2013 (UTC)
Bot to add missing merge tags
This sounds like something that may have been proposed before, but can we run a bot to add missing merge tags on all the articles in
Category:Articles to be merged? I would say about half of the articles proposed to be merged do not have the accompanying tag on the target article. What I mean is that, if the article has {{mergeto}} or {{mergefrom}}, the bot would check that target article article has the reciprocal merge tag. My thinking here is that if there's an article missing one of them, and the merge is clearly not supported, someone would remove the tag from both articles.
Also, on a similar train of thought, a more permanent bot would be one that could check that these tags are in place until the merge is resolved. I think many proposed merges are inactive simply because 50% of the editors have no idea there has even been such a proposal. Usually the merge tag winds up on the crappier, less viewed article rahter than the highly viewed target. --
NickPenguin(
contribs)01:49, 25 February 2013 (UTC)
anti-spam bot to show how many times the same spam has been added to an article and removed
I need a bot to check certain pages which have had, for years now, IP addresses with no other edits, spam links to commercial sites.
Microsoft Ants for example, if you check the history, you will find voobly.com has been spammed there many times over the years, and reverted by me or others. I filed a report on some cases of them spamming articles at
[2]. I would like to be able to run a bot on any page I select to investigate, which would check every single edit in the page's history, scanning the text in each edit to determine when certain text was added, and then list that information. Be able to see the edit, date, and name of editor who put it there, clicking on a link to see their contributions. It makes it easier to show a constant pattern of abuse by commercial sites that do this across Wikipedia. If looking over that, one finds that dozens of different IP addresses have edited only to add the same spam link to an article or other articles, it proves the Wikipedia Spam filter needs to be updated to stop that, and those pages protected from IP address edits. When scanning for something, should be able to do multiple scans at the same time, typing in the names of various often spammed links, since sometimes they have the same owner, and they rotate which link they put in there, or more than one site is regularly doing this to the same place.
DreamFocus07:55, 4 March 2013 (UTC)
I think this whole tree needs to be deleted, and some related trees as well. I really do not feel like tagging the whole thing, so I was hoping it would be possible to have a bot assist.
John Pack Lambert (
talk)
18:45, 7 March 2013 (UTC)
I can do it with AWB. However, it's been almost 5 days, so I'm not sure it's a good idea at the moment. (We need a category, AWB users willing to submit bulk nominations.) —
Arthur Rubin(talk)16:28, 12 March 2013 (UTC)
A bunch of articles
[3];
[4] are using the deprecated {{Coor dm}} or {{Coor dms}} for their coordinates; they should use {{Coord}} instead, to which the former templates redirect. As we don't simply fix redirects, it would be good if this could be added to some routine cleanup bots'/ AWB's tasks (is there a better place to request this?) Here is an example edit. Andy Mabbett (Pigsonthewing);
Talk to Andy;
Andy's edits20:11, 7 March 2013 (UTC)
In general, we do not approve bots to go around simply replacing a redirect with its target, unless the redirect is about to be deleted. It is also against the AWB rules to use AWB to perform that sort of cosmetic change. — Carl (
CBM ·
talk)
17:07, 11 March 2013 (UTC)
I acknowledged that "we don't simply fix redirects". By "some routine cleanup bots'/ AWB's tasks" I mean that the change should be applied at the same time AWB is being used to make other changes. Andy Mabbett (Pigsonthewing);
Talk to Andy;
Andy's edits20:49, 15 March 2013 (UTC)
The categories I have checked have made odd usage of {{Year by category}}, and not used the parent parameter, so it seems to me that the best thing to do is to replace the entire content of the category.
The search string only needs to select the whole page while matching the year, and split it into 4 parts: ^.*([12])([890])([0-9])([0-9]).*$
... the replace it with:
{{Cat main|Motorsport in Italy}}
{{Year by category
|m=$1
|c=$2
|d=$3
|y=$4
|cat = in Italian motorsport
|subcat = $1$2$3$4 in Italian sport
|sortkey = Motorsport
|parent = Motorsport in Italy by year
}}
[[Category:$1$2$3$4 in motorsport|Italian]]
Webcite generation for set of domains that are scheduled to be shutdown
A publisher is set to shut down three major RS sites for video games: 1UP, UGO, and Gamespy. (The sites effectively have stopped publishing but their content is still there, but we have no idea how long that will last). We want to try to webcite these links, a list is given at
[5]. We had a previous situation where we knew a web site was going dark in a matter of days and had a bot run through and webcite the links and update the articles, but I'm having problems finding that request from before. Here we don't think we're in as big a rush to fix this but certainly have the issue that we have no idea of the timing here. Is there a bot set up to handle this or would a new one need to be made? --
MASEM (
t)
17:20, 18 March 2013 (UTC)
I have put in a nomination
here to rename all of the by-year categories which currently are subcategories of
Category:Events_in_the_Thirteen_Colonies to
Category:Events in the British colonies of North America. I put a note on the CfD talk page (per the instructions) for help tagging all of the relevant categories, but someone suggested I try here instead. Can you help? In short, I need to add a CfD note for a rename to all of the 180+ subcategories of [:Category:Events in the Thirteen Colonies] -- Dis/establishment by Year/Decade/Century. There are one per year from 1607 to 1775, plus 18 decade ones, and two centuries. Thankfully, most of the "Disestablishment" categories haven't been created yet. Thank you!!
JRP (
talk)
07:07, 19 March 2013 (UTC)
Insert at the top of the page: {{
subst:cfr|Formula One race reports}}
The resulting notice won't state the renaming target, and it would be better if the rename target was specified.
Specifying the rename target
Insert at the top of the page: {{
subst:cfr|$1 Formula One races|Formula One race reports}} The $1 value for the year can be picked up a regex which by matches the YYYY string at the start of the title of the parent category YYYY in Formula One (e.g. ‹The
templateCategory link is being
considered for merging.›Category:1990 in Formula One).
Template:CongLinks, used in External links, has a washpo parameter, for Washington Post. The broken links are all alphabetic (plus underlines and possibly commas) but the working ones are a mix of alpha and numbers in two versions. For example,
Blanche Lincoln is broken,
Dick Durbin works, and anything that looks like f9d0a3fa-4bbc-11e2-8758-b64a2997a921 works. I need a bot to go through and blank out only the broken (old version) instances.
184.78.81.245 (
talk) —Preceding
undated comment added
19:10, 10 March 2013 (UTC)
Could you please clarify? It appears that an all alphabetic parameter on
Dan Coats works just fine. Do the broken links all contain underscores? Thanks!
GoingBatty (
talk)
03:12, 11 March 2013 (UTC)
Should the |nndb= and |findagrave= parameters also be removed at the same time, since the template no longer supports them? Thanks!
GoingBatty (
talk)
03:20, 11 March 2013 (UTC)
Deleting would remove the incorrect links that are currently on articles, and would still give the ability to repair later on. However, would you prefer to have a report of all of the parameters that need repairing?
GoingBatty (
talk)
16:52, 11 March 2013 (UTC)
GoingBatty, you are absolutely correct about Dan Coats. Would it work to delete washpo parameter values which include an underline? I wouldn't delete findagrave and nndb, as those were removed with no discussion, and Template:NNDB and Template:Find a Grave are widely used. There's often confusion between links acceptable for EL and links which are acceptable as reliable sources for references. Perhaps after the cleanup is done, a list of current Senators and Representatives who lack a washpo parameter would be a useful addition to the relevant Wikipedia Project, or perhaps post an example of the code needed to scan for a missing or valueless parameter in templates, starting from a list of articles such as the members of the 113th Congress. I've now remembered a related broken link problem with the votesmart parameter. I believe those which include alpha characters no longer work.
184.78.81.245 (
talk)
17:23, 15 March 2013 (UTC)
I have a massive number of redirects for a bot to create. Bellow is a list (from
Former counties, cities, and towns of Virginia) of former Virginia counties, now part of West Virginia and Kentucky. The list also includes some cities and towns that are now part of West Virginia. The list has been modified so that it links to X, Virginia instead of X, Kentucky/West Virginia
The vast majority of entries on this modified list are redlinks. I would like a bot to redirect these redlinks to the correct Kentucky/West Virginia article. Be aware that, as the article says "Many of these [Kentucky] names were later reused to name other new Virginia counties. Some of those were "lost" again when the state of West Virginia was formed in 1863." so in those cases where Kentucky and a West Virginia county have the same name , I would strongly recommended the bot create a disambig instead of a redirect to eater article. In the case where a current Virginia location has the same name as one of these old counties/cities/towns, I would like the bot to create a hatnote.
Emmette Hernandez Coleman (
talk)
01:53, 24 March 2013 (UTC)
Update: I've taken care of the Kentucky counties manually, including the ones that needed a disambig for a West Virginia county having the same name. That just leaves the West Virginia ones for the bot to take care of.
Emmette Hernandez Coleman (
talk)
02:21, 24 March 2013 (UTC)
I'm hoping I understood your request, as it stands with the current updates. You want the rest of the redlinked pages (<county>, Virginia) redirected to West Virginia pages (<county>, West Virginia), considering the such target exists? If so, I could do it. If not, please clarify. Hazard-SJ ✈ 04:37, 24 March 2013 (UTC)
A couple of months ago,
Yobot fulfilled a request that I made on behalf of
WP:JAZZ; the details are available in the
archived discussion. It appears that at least some of the edits need some cleanup:
The {{WikiProject Jazz}} banner did not inherit the class= rating from the existing WikiProject banners (or derive it from article's length, or presence of a stub template in the article)
The 'bot always placed the banner at the top of the page, regardless of whether {{WikiProjectBannerShell}} or {{Talkpageheader}} was already present. (In at least one case it fixed a redirect link to {{WikiProjectBannerShell}} but didn't place the new banner inside the shell; in another it actually added the shell but didn't include the banner in it: [7].)
For pages in the
/Songs list, the parameter songs=yes was added; it should have been song=yes (I'm not sure if it affects all the song articles, but I've noticed it with those pages that were originally tagged with album=yes and then re-tagged a few minutes later).
I have spot-checked a few dozen edits from Jan. 4-5 and I'm consistently seeing these issues. For example, [8], [9]. Of course where there were no existing assessments, or no banner shell, this is a non-issue (other than not identifying stubs).
I apologize that I didn't spot this and bring it up any sooner. I'm not sure if Yobot (or another 'bot) can simply retrace its earlier steps, or whether it needs to go back through the categories again, but the category scheme remains largely the same (cat. list is at
Wikipedia:WikiProject Jazz/Categories. I'm making a few edits to remove
red links, add at least one known new category, etc.). If it does need to go through the categories again, maybe tag any new articles along the way, as per the previous request. --
Gyrofrog (talk)20:20, 13 March 2013 (UTC)
I am sorry my edits weren't perfect. I hope someone helps to fix these because I am busy in real life to deal with this task. --
Magioladitis (
talk)
16:17, 16 March 2013 (UTC)
Magioladitis, my request for auto-assessment may not have been clear in the first place – I very briefly mentioned it in my request and then placed a link to the 2010 request, so it was very easy to miss. --
Gyrofrog (talk)15:33, 18 March 2013 (UTC)