Jump to content

Wikipedia:Edit filter/Requested

From Wikipedia, the free encyclopedia
< Wikipedia:Edit filter

This is an old revision of this page, as edited by Mahitgar (talk | contribs) at 17:26, 8 June 2015 (→Persistant spam). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Shortcuts

This page is for people without the abusefilter-modify permission or people without sufficient knowledge about the coding involved to make requests to enact Abuse filters. Please add a new section at the top of current requests using the following format: 

===Filter name===
*'''Task''': What is the filter supposed to do? To what pages and editors does it apply?
*'''Reason''': Why is the filter needed.
- ~~~~

Bear the following in mind:

  • Filters are applied to all edits. Therefore, problematic changes that apply to a single page are likely not suitable for an abuse filter.
  • Each filter takes time to run, making editing (and to some extent other things) slightly slower. The time is only a few milliseconds per filter, but with enough filters that adds up. When the system is near its limit, adding a new filter may require removing another filter in order to keep the system within its limits.
  • There is a limit to what filters can check for. More complex, non-essential tasks, such as those that need to perform a more in-depth check of the page or fetch information that the filter system does not have access to, are better served by separate software, run by an individual user on their own machine or dedicated server such as Tool Labs, rather than those used to actually host Wikipedia.
  • It used to be called the abuse filter for a reason. Contributors are not expected to have read all 200+ policies, guidelines and style pages. Trivial formatting mistakes and edits that at first glance look fine but go against some obscure style guideline or arbitration ruling are not suitable candidates for an abuse filter -- quite apart from performance concerns, if it doesn't harm the project, it is best not to hassle new contributors because of it.
  • To prevent the creation of pages with certain names, MediaWiki:Titleblacklist is usually a better way to handle the problem - see MediaWiki talk:Titleblacklist for details.
  • To prevent the addition of problematic external links, please make your request at the spam blacklist.

Current requests

Turkish IP adding nonsense

  • Task: The filter should reject edits from Turkish IP who is adding nonsense on many wikis.
  • Reason: There are multiple IP's and those edits always need to be reverted. There are already a global filter at m:Special:AbuseFilter/111 and I think that it should be copied to enwiki as well. If you can't see what the filter contains you can contact me or make a new one based on edits by 78.174.123.127 and 88.253.152.154.

Thanks. Stryn (talk) 18:32, 5 June 2015 (UTC)[reply]

Warn against clipboard hijacking

  • Task: Recognize the addition of strings similar to what I removed in this edit, which are caused by clipboard-hijacking scripts on the source sites, and warn users before the edit is saved.
  • Reason: Some websites use clipboard hijacking techniques that add text if someone copies and pastes something, usually the article title in our case. Not only is the text almost always inanely promotional, it frequently breaks citations. It seems like the added text is formulaic enough that it might be targetable by a filter.

- {{Nihiltres|talk|edits}} 21:02, 4 June 2015 (UTC)[reply]

@Nihiltres: Can you give more examples to show what text is added by other sites? Targeting a single site should be easy, and with more examples I assume targeting a bunch wouldn't be too much harder. PHANTOMTECH (talk) 21:57, 4 June 2015 (UTC)[reply]
@PhantomTech: Here and here are some more; there are probably others in my contributions but I'd have to sift through more thoroughly. {{Nihiltres|talk|edits}} 22:15, 4 June 2015 (UTC)[reply]
@Nihiltres: Thanks, here are some regexes for anyone that wants to try out a filter.
  • read more( at|(?!\.|,|\s)\W)
  • title ?= ?[^\|\}]*?read more
  • title ?= ?[^\|\}]*?read more( at|(?!\.|,|\s)\W)
Any of those will catch all the examples given. The first one is the broadest, second refines to reference titles only (as opposed to anywhere in the article) and the third tries to remove any false positives from the second. I'm not sure which will work best, it all depends on the false positives that come up, it should be checked with a added_lines irlike and !removed_lines irlike to avoid bugging people who haven't added the reference themselves. PHANTOMTECH (talk) 22:44, 4 June 2015 (UTC)[reply]

Inexperienced users removing amboxes

Sometimes, inexperienced users try to remove article maintenance tags (amboxes) from articles.

  • Task: Every time an IP or unconfirmed user removes a {{notability}}, {{news release}}, or {{COI}} ambox from any article, please tag the edit.
  • Reason: This filter will help us to notice edits such as these.[1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16] It's happened many times on Wikipedia that COI users have wrongly removed such amboxes. Often, they use a blank edit summary, making the damage hard to catch. Tagging the edits will make it easier for us to notice and undo the damage.
(Dear Wikipedians: Please freely edit and improve this filter request.)

Thank you. —Unforgettableid (talk) 00:43, 18 September 2013 (UTC)[reply]

Not done. Detecting each and every tag that gets removed would put a severe strain on the server resources. Reaper Eternal (talk) 13:13, 16 November 2013 (UTC)[reply]
I thank you for your reply.
  1. OK, we could watch only for the removal of {{notability}} (perhaps the most important ambox) and ignore all other ambox removals.
  2. Let me suggest some non-obvious optimizations (besides the obvious "look at user groups first"). For one thing, we could only look at edits where edit_delta is between -10 and -100. This is an imperfect heuristic, but would save time. Next, we could look only at edits where the edit summary is blank: if there's an edit summary, then this is probably either a section edit or some other non-tag-removal edit. Okay; by now we've already eliminated a huge proportion of edits. Next, we could look at the first byte of either old_wikitext or removed_lines: if it's not '{', we can stop now. Finally, we could search through the contents of removed_lines. We could use contains "otability": a literal search is probably faster than a glob or regexp search.  Question: Would such a filter still be too CPU-intensive? And if so: which condition would be the biggest problem?
  3. This filter is non-crucial and non-private. If the client has JavaScript on, we could theoretically design the infrastructure to offload all non-crucial, non-private filtering work to the client; if the client has JavaScript off, we could just skip those filters.
—Unforgettableid (talk) 07:53, 27 November 2013 (UTC)[reply]


 Discussion ongoing...

John Daker sockpuppets

- TL22 (talk) 23:13, 2 June 2015 (UTC)[reply]

IP disruption from multiple /8s

  • Task: Block the addition of � to user talk pages, coming from the IP Ranges 81.0.0.0/8 84.0.0.0/8, 86.0.0.0/8 (Maybe even broader than specific /8s)
  • Reason: Prolific IP vandal is spamming a number of user talk pages, amongst other things using that character. At least a dozen direct blocks, some semi-protections and at least one small range block have failed to prove effective.

- Monty845 23:57, 30 May 2015 (UTC)[reply]

@Monty845: Seems reasonable, could you provide some IPs which have been doing this for testing purposes? Sam Walton (talk) 23:59, 30 May 2015 (UTC)[reply]
Could start with 86.142.21.50, 86.131.23.71, 81.131.77.180, 86.131.23.67 Monty845 00:06, 31 May 2015 (UTC)[reply]
I can't find the 84. one, so perhaps start with 81. and 86. Monty845 00:14, 31 May 2015 (UTC)[reply]
The vandalism seems to be coming from IPs from BT Total Broadband (UK), maybe an abuse report to the ISP may suffice Already sent one. Esquivalience t 00:40, 31 May 2015 (UTC)[reply]
They have already started changing their behavior, so my initial request wont stop them. New AN thread on the topic at was started by someone else at WP:AN#Rangeblock?. Monty845 17:32, 31 May 2015 (UTC)[reply]

no Declined. Way too broad—that's most of the UK. Reaper Eternal (talk) 03:40, 2 June 2015 (UTC)[reply]

Itailevi00 socks

- Esquivalience t 14:43, 30 May 2015 (UTC)[reply]

@Esquivalience: I don't see blocking usernames being useful (They'll just use a different string and be harder to track), but it's not an edit filter option anyway, you'd want the username blacklist. We could, however, track edits from users with that string in their username, would that be helpful? Sam Walton (talk) 00:00, 31 May 2015 (UTC)[reply]
@Samwalton9: A hidden, log-only filter that tracks account creations and edits from the Itailevi family of sockpuppets would be useful. Esquivalience t 00:06, 31 May 2015 (UTC)[reply]

no Declined. Just use Special:ListUsers/Itailevi. Reaper Eternal (talk) 03:43, 2 June 2015 (UTC)[reply]

Persistant spam

  • Task: put under observation and/or disallow words "Javad_Ramezani" by not autoconfirmed users with a timestamp atleast one year.
Very smart longeterm interwiki spamster, Copies some lines from some other article adds words Javad_Ramezani for self promotion, and creates new articles with new article names from new ips or new user names.


  • Please refer
this google search on en wikipedia
this google search on wikipedia.org to include all possible wikis
Wikipedia:Articles for deletion/Iran man Besides also refer User:Iran man.
Woman (song javad) as of now still not gone for deletion request.
Insertion seems to haves been made also at article List of Iranian musicians

I suppose experinced spam fighters need to look into the issue.

- Mahitgar (talk) 05:10, 27 May 2015 (UTC)[reply]

While this request is pending, still the same spam seems to continue. From google search I came across Aman az Eshgh (album) this page, where in there seems to be three attempt of recreation and admins deleting the same 31st may, 4rth June, 6th June. and also I came across a latest one dated 8th JuneHero (Javad song) which is still to get nominated for deletion process.
Mahitgar (talk) 17:26, 8 June 2015 (UTC)[reply]

Persistent sockpuppetry involving a name

  • Task: Disallow the creation of usernames involving variations of the words "smooth" and "aashu". If that is not possible for technical reasons, prevent IPs and new accounts from adding content involving such words.
  • Reason: Please see this SPI for reference. This indeffed user keeps recreating similar promotional accounts involving his stage name Smoothest Aashu.

- De728631 (talk) 18:41, 15 May 2015 (UTC)[reply]

@De728631: Created, log only for now at Special:AbuseFilter/690. Sam Walton (talk) 11:19, 25 May 2015 (UTC)[reply]
Awesome, thank you, Sam. And it's working, too. See Wikipedia:Sockpuppet investigations/Smoothestaashu007. De728631 (talk) 16:35, 31 May 2015 (UTC)[reply]
@De728631: Excellent. I may set the filter to disallow the edits (they'll still show up in the log for further investigation) as a deterrent. Thoughts? Sam Walton (talk) 16:48, 31 May 2015 (UTC)[reply]

Changing owner field in infoboxes

  • Task: Disallow non-autoconfirmed editors from changing the |owner= field in certain infoboxes, such as {{Infobox NHL team}} (but likely useful to apply to most sports infoboxes)
  • Reason: Periodic vandalism of articles like this, this and this has been going on for years, but seems to be increasing in frequency lately, perhaps in part because media has taken to writing lazy "someone vandalized Wikipedia hyuk hyuk" articles as of late (e.g.: CSN Chicago). I have no recent examples since I don't follow many articles for other sports teams, but I have seen these types of vandal edits on baseball, basketball and football team pages in the past. I could put together a list of templates if this is doable and desirable.

- Resolute 00:08, 15 May 2015 (UTC)[reply]

Sigh... Again and again. Resolute 16:58, 15 May 2015 (UTC)[reply]
@Resolute: A list of templates to monitor would be useful. Tracking all |owner= changes will be too broad I think. Sam Walton (talk) 11:22, 25 May 2015 (UTC)[reply]
@Samwalton9: - Sure. The NHL infobox listed above (and which would allow me to lift protection on three articles), {{Infobox MLB}}, {{NBA team}}, {{Infobox NFL team}} and {{Infobox football club}} would be a good start. All five use the same owner= parameter. Resolute 14:00, 25 May 2015 (UTC)[reply]
@Samwalton9: It keeps happening. Fortunately, the NHL playoffs will end soon, but I think we've had to protect at least seven team articles during the post-season for this exact reason. I'm seeing it in the odd NBA article as well. Thanks! Resolute 14:16, 3 June 2015 (UTC)[reply]

Fixed typo

  • Task: Pick up anonymous edits that have inaccurate formulaic edit summaries like "Fixed typo", "Fixed grammar" or "Added links", often in combination.
  • Reason: We are getting one heckuva lot of them these days, often concealing petty vandalism that ClueBot NG is simply not picking up. Examples: [17] [18]

- Redrose64 (talk) 19:31, 8 May 2015 (UTC)[reply]

We already have the "canned edit summary" thing for that. --TL22 (talk) 01:01, 9 May 2015 (UTC)[reply]
I've seen that tag elsewhere, but it's not on these examples. --Redrose64 (talk) 07:44, 9 May 2015 (UTC)[reply]
Special:AbuseFilter/633 doesn't pick up combinations of canned edit summaries. The regex could be changed to ^(?:\/\* .* \*\/ )?(Fixed typo|Fixed grammar|Added links|Added content)(?:, \g<1>)*$ to detect them as long as recursive subpattern matches are possible, if not then it isn't too hard to change it to still work. PHANTOMTECH (talk) 05:27, 27 May 2015 (UTC)[reply]
This one wasn't a combination of canned edit summaries, but didn't trigger the tag. --Redrose64 (talk) 12:01, 27 May 2015 (UTC)[reply]
The edit summary in that one doesn't have the default space after the section name, ^(?:\/\* .* \*\/ ?)?(Fixed typo|Fixed grammar|Added links|Added content)(?:,? \g<1>)*$ is the modified version (of my already modified version above) that will detect both that and the combinations. Might not be a bad idea to modify the filter to ignore case also. PHANTOMTECH (talk) 05:59, 28 May 2015 (UTC)[reply]

shadow youtube spam

- JacktheHarry (talk) 22:21, 22 April 2015 (UTC)[reply]

@JacktheHarry: Not sure if this is a widespread enough issue to warrant an edit filter, but Special:AbuseFilter/691 is log only for now to see. Sam Walton (talk) 11:28, 25 May 2015 (UTC)[reply]

Copyviocore

  • Task: Due to an OTRS ticket, I noticed that in this edit, Template:Copyviocore was added by an anonymous user, blanking the article for six months for anyone who wanted to read it, and was not noticed by any users.
  • Reason: This filter would allow for the monitoring of the placement of this template by non-autoconfirmed users (or users who normally do not add it, maybe those with less than 1,000 edits or something) to prevent vandalism, as it could be easily used to cover over articles that vandals may not like, with no quick consequences.

- Kevin Rutherford (talk) 22:17, 13 April 2015 (UTC)[reply]

@Ktr101: Log only for now at Special:AbuseFilter/692. Sam Walton (talk) 11:55, 25 May 2015 (UTC)[reply]

IP MOS vandalism filter - culry quotes, decapitalization

  • Task: Filter should detect abusive edits by long-time IP editor who edits counter to the MOS, specifically by adding curly apostrophes and quotemarks, changing spelling out numbers to numerals, and decapitalizing position titles. Can also detect the edit summary "Punctuative corrections". Applies to mainspace articles and edits by IP editors.
  • Reason: I noticed an IP editor (68.98.155.153 (talk · contribs · deleted contribs · logs · filter log · block user · block log)) who was making punctuation and style changes against the MOS. The user's most noticeable actions were to change straight quotes/apostrophes into curly ones (opposite WP:' and MOS:QUOTEMARKS). They also decapitalized position titles like Secretary of Defense → secretary of defense. Also placing punctuation inside quote marks, again WP:LQ. After searching a bit, I found two other IP editors using the same edit summary and making the same changes as far back as 2012 (see 68.98.155.114 (talk · contribs · deleted contribs · logs · filter log · block user · block log) and 68.98.157.20 (talk · contribs · deleted contribs · logs · filter log · block user · block log)). These editors also changed spelled-out numbers into numerals, against WP:SPELL09. Gogo Dodo noted on their talk page that the IP addresses were from Cox.
Diff examples:
Ping me if more explanation or diffs are needed. Thank you!

- EvergreenFir (talk) Please {{re}} 21:43, 4 April 2015 (UTC)[reply]

Hi EvergreenFir, do you know if this has happened with more IPs? Edit filters aren't used for enforcing the MOS but if this is a long-term issue with more than just these three IPs then I could set one up. Sam Walton (talk) 16:47, 5 April 2015 (UTC)[reply]
For future reference the filter currently in my sandbox should be able to detect the curly quote and logical quote changes. Capitals would be impossible to check as far as I know. Sam Walton (talk) 16:58, 5 April 2015 (UTC)[reply]
@Samwalton9: I found two more in the 68.98.155.* range: 68.98.155.48 (talk · contribs · deleted contribs · logs · filter log · block user · block log) (from late 2014) and 68.98.155.223 (talk · contribs · deleted contribs · logs · filter log · block user · block log) (currently active). They use the same "punctuative corrections" edit summary.
I honestly am not too sure how the edit filters work, so I'll take your word for it on the capitalization stuff. Their primary MO seems to be (1) "punctuative" in the edit summary and (2) the curly quotes.
I'll look in some other IP ranges under 68.98.*.* now that I learned how to get the user contribs to show IP ranges. Cheers! EvergreenFir (talk) Please {{re}} 16:32, 6 April 2015 (UTC)[reply]
Found another from April - December 2013: 68.98.154.182 (talk · contribs · deleted contribs · logs · filter log · block user · block log) EvergreenFir (talk) Please {{re}} 16:58, 6 April 2015 (UTC)[reply]
And some more:
@Samwalton9: Thank you very much! Is there anything I should do when the filter catches them other than revert (and possible warn)? EvergreenFir (talk) Please {{re}} 20:06, 7 April 2015 (UTC)[reply]
@EvergreenFir: Not for now. I'll be keeping an eye on the filter too and after a while of having the filter in log only mode we'll see if its useful and how to proceed (perhaps disallowing the edits). Sam Walton (talk) 20:34, 7 April 2015 (UTC)[reply]
@Samwalton9: Sounds great! Thank you! EvergreenFir (talk) Please {{re}} 20:36, 7 April 2015 (UTC)[reply]

Draft being moved to mainspace before review

  • Task: This filter should be able to detect whether an editor attempts to move a draft before review (technically, before the draft has been sumbitted for review (not counting declined reviews) or if the draft has been moved by the draft creator) and warn the moving editor with a friendly notice (since people who do that tend to not know how drafts work and are doing it in good faith).
  • Reason: People who create drafts tend to move their drafts to the mainspace before sumbitting them for review, which can be bad because the article can have some problems that could normally be detected by a draft reviewer, and can lead it to a(n) PROD/CSD/AFD tag on the article.

- TL22 (talk) 18:20, 27 March 2015 (UTC)[reply]

Since I'm not receiving any response; User:Samwalton9? --TL22 (talk) 23:53, 13 April 2015 (UTC)[reply]
@ToonLucas22: I've been going back and forth on whether this is suited to an edit filter. Though I agree it would be useful for everyone, edit filters are more for edits which are necessary to track or stop, and we're quickly running out of room for new filters. Sam Walton (talk) 12:04, 25 May 2015 (UTC)[reply]
An additional flag might be the inclusion of 'dead link' in the edit summary. User Tomofm2 wrote it in one and the newly noticed user Divine4778 also wrote it in some of their edit summaries. Sam Walton (talk) 17:58, 9 December 2014 (UTC)[reply]
Would it also be possible to track changes to the url that precedes a {{dead link}} template? If we start monitoring the use of this template the SEO people might just leave it in the articles and replace the broken links anyway. So a possible marker should also detect changes to links that have been flagged with [dead link]. De728631 (talk) 01:00, 10 December 2014 (UTC)[reply]
If we detect "dead link" in removed_lines and completely ignore whether it occurs in added_lines, it will accomplish that. Jackmcbarn (talk) 05:45, 28 December 2014 (UTC)[reply]
See the recently created WP:DEADLINKSPAM for an overview. Sam Walton (talk) 16:52, 24 December 2014 (UTC)[reply]

Redirects to badimages

To deal with socks of 1abacada (talk · contribs) who have discovered that they can redirect pages to files on the bad images list, a filter is needed to prevent this, at least for autoconfirmed users (though I can't see a reason for anyone short of an admin to do this). Acroterion (talk) 02:51, 4 May 2014 (UTC)[reply]

@Acroterion: I've created a filter definition that will stop this behavior and emailed it to you. To enable it here, first add yourself to the abusefilter group at Special:UserRights/Acroterion. Then, go to Special:AbuseFilter/import, paste it, click "Import data", then click "Save filter". Jackmcbarn (talk) 00:50, 13 May 2014 (UTC)[reply]
Acroterion did this filer get set up? Sam Walton (talk) 21:57, 21 January 2015 (UTC)[reply]
Not that I'm aware of: I don't have the technical expertise to review it, and since I'd b e responsible for its effects, I didn't make myself an abuse filter editor, etc. Acroterion (talk) 21:54, 24 January 2015 (UTC)[reply]

Replacing text with repeating characters vandalism

  • Task: To disallow editors that not only add repeating characters to an article but also remove content.
  • Reason: I did a check on the recent changes page that all of the most recent edits that remove content under the "repeating characters" tag filter are vandalism.

- Minima© (talk) 22:06, 17 February 2014 (UTC)[reply]

John Galea

  • Task: Prevent creation of pages using the text of the deleted page John Galea. I'd suggest filtering on anything containing the name "John Galea" (although he's not been averse to mis-spelling his own name in order to get around create protection).
  • Reason: The page has been repeatedly recreated under multiple alternate titles by a veritable army of socks. Salting and rangeblocking are not effective measures against this user.

- Yunshui 雲‍水 15:19, 17 February 2014 (UTC)[reply]

Yunshui, could you list some of the other article titles? And do you know if this is still an issue? Sam Walton (talk) 00:03, 17 January 2015 (UTC)[reply]
@Samwalton9: Sure - some of the other titles were: John Galea (Singer), John Galea (singer), John- Paul Galea, John Galea!, John Galea (musican), John_Galeaa, John Galeea, John Galea (Paul), Do it my way, (Singer) John Galea, John Galea (performer), John Galea ( the musican ), J.Galea, John Galea (artist)... Not an exhaustive list by any means, but it gives you an idea of the problem. He was still apparently socking as recently as last week (albeit after a hiatus of several months), so I'd say the problem is still ongoing. Thanks for taking a look. Yunshui 雲水 10:12, 5 February 2015 (UTC)[reply]

G-Zay

@Sjones23: I imagine the reason for this filter having not been made yet is the lack of anything obvious to search for. Can you suggest anything we could search for in edits that would pinpoint this user? Perhaps something relatively unique they tend to write in edit summaries or edits. Sam Walton (talk) 20:36, 8 April 2015 (UTC)[reply]
Well, I'll see what can be done about it. Lord Sjones23 (talk - contributions) 20:53, 8 April 2015 (UTC)[reply]

Joshua Hawkins fakearticling

@The Bushranger: If this is still happening, remove the stale tag. PhantomTech (talk) 06:06, 20 March 2015 (UTC)[reply]

Page deleted, then identically-named page created

  • Task: Say a page has been deleted, then a non-sysop with an editcount under 5,000 later creates another page with the same title. If this happens, please silently tag the page-creating edit with a tag. The tag should say that a page with that name has already been deleted X number of times.
  • Reason: This will help Wikipedians to more easily notice when a formerly-deleted page has been recreated. This will help make it more obvious to them when they should nominate "new" pages for deletion. It will also make it more obvious when a page title should be SALTed.

Thank you, —Unforgettableid (talk) 08:38, 30 August 2013 (UTC)[reply]

Unless I'm mistaken, the edit filter can't see the deletion log for a page being edited, so this isn't technically possible. A bot would be better for this. Jackmcbarn (talk) 13:51, 30 August 2013 (UTC)[reply]
Thank you for your reply. AnomieBOT (talk · contribs) runs a task called NewArticleAFDTagger, which tags recreated AfD-deleted pages with {{old AfD multi}}. AFAIK there is no bot which tags recreated PROD-deleted pages or recreated speedily-deleted pages. Where is the best place for me to request that someone provide that functionality? (In vaguely-related news: bugzilla:10331, which requests a page-creation log, is still unfixed despite five years and one patch.) Cheers, —Unforgettableid (talk) 06:03, 2 September 2013 (UTC)[reply]
That would be Wikipedia:Bot requests. Jackmcbarn (talk) 16:37, 3 September 2013 (UTC)[reply]
OK, thank you. So it's best that I go there instead of contacting User:Anomie directly? —Unforgettableid (talk) 19:14, 4 September 2013 (UTC)[reply]
Can do both. Wifione Message 19:34, 4 September 2013 (UTC)[reply]
I requested a {{old prod full}} tagger bot at Wikipedia:Bot requests/Archive 57#Bot to tag "PROD Survivors" and "Recreated Articles", and put a {{talkback}} template at User talk:AnomieBOT, but (despite one empty promise I got) nobody coded anything. Many have thought about writing such a bot, but nobody has ever written a practical one; I explained the matter further in my request there. I didn't create a bot request to tag recreations of CSD-deleted pages, but someone else is welcome to do so if they like. —Unforgettableid (talk) 07:53, 27 November 2013 (UTC)[reply]
Hmm, I just had an idea for how this can be implemented:

(action = "delete") | (action = "edit" & old_size = 0)

with a per-page throttle.
However whether this might be worth it or better suited for a bot is another question. Triplestop (talk) 03:53, 21 September 2013 (UTC)[reply]
That's an interesting solution. I hadn't thought of using the rate limiter like that. It looks like it would work, but it seems a bit hacky and would have a few FPs and oddities, such as tagging the deletion of a newly-created page, and tagging undoing of page blanking. Jackmcbarn (talk) 12:45, 21 September 2013 (UTC)[reply]
User:ThePhantomBot is currently logging this type of behavior and I plan to make it do more than that after it has been approved. PhantomTech (talk) 06:47, 25 March 2015 (UTC)[reply]

Plot summary copyvios

  • I had a look and found nothing. However, you've set the size threshold too high -- my experience is that the copyvio plot summaries are usually only one paragraph (3-7 lines) long. I've changed it to 400 bytes. You are also missing the very important scenario when a plot section is added without there being an existing section and an edit summary -- you'll want to search for the addition of a level 2 header like [Pp]lot|[Ss]ynopsis as the trigger. MER-C 12:04, 23 January 2015 (UTC)[reply]
MusikAnimal has improved the filter a bit so we'll see how it goes. Narrowed to adding a new section rather than checking for any largeish plot/summary edit to existing ones. Sam Walton (talk) 19:20, 7 April 2015 (UTC)[reply]
@MER-C and MusikAnimal: Could you check if the filter is worth continuing? Sam Walton (talk) 12:08, 25 May 2015 (UTC)[reply]

Lions at Cat Creek

@The Bushranger: Are they still active? PhantomTech (talk) 06:06, 20 March 2015 (UTC)[reply]
@PhantomTech: I've seen a few socks of this user recently, but not as many as in the past. Ping me if you need me to reply EvergreenFir (talk) Please {{re}} 21:26, 23 April 2015 (UTC)[reply]

Language speaker data

  • Task: Filter the 'speakers' parameter of {{Infobox language}} for changes, similarly to how changes to height and weight in bio boxes are tagged. (If possible, filter 'date' and 'ref' under the same tag: these are all elements of the population figure.) Should apply to editors without advanced permissions.
  • Reason: Population inflation is a chronic problem with our language articles, and isn't easy to detect if you don't see it happen. Although this wouldn't catch changes to the text, vandals and POV warriors normally change the info box as well. This isn't just a problem with IPs, but often with signed-in POV editors. The date may be changed to make the data look recent. If 'date' and 'ref' can be covered without increasing server load significantly, please include them; otherwise 'speakers' is the main problem.

— kwami (talk) 00:28, 2 November 2012 (UTC)[reply]

XSS Filter detect

  • Task: Block Internet Explorer 8+ users from saving edits that triggered the XSS Filter (Bug 32013). The XSS Filter transforms periods, bracket, and parentheses to the Number sign. One simple implementation might be ##[^{|}<\n>[\];:*]{5,255}?##
  • Reason: Because IE is extremely sneaky doing this after show change displays everything correctly and devs refuse to send the header to stop the filter. We've had many complaints about this.

— Dispenser 06:11, 27 August 2012 (UTC)[reply]

Dispenser, do you know if this is still an issue? Sam Walton (talk) 09:07, 19 January 2015 (UTC)[reply]
Yes, its still an issue 29 months later. This will continue as computers running Internet Explorer 7/8 (End of Life 12 Jan 2016) and have commendable market share. Windows XP has an 18% market share or 1 out of every 6 computers and the highest version of Internet Explorer is 8. Of course WMF could increase interoperability, but they're too distracted by the Web Design Hipsters. — Dispenser 19:52, 19 January 2015 (UTC)[reply]
You could match lines removed with \[\[([a-z ]{1,50}\|)?[a-z ]{1,25}]] and lines added with ##([a-z ]{1,50}\|)?[a-z ]{1,25}## to identify the changing links. PhantomTech (talk) 07:33, 21 March 2015 (UTC)[reply]

This is not a talk page. Please add new requests at the top of this section, not here at the bottom. Thank you!

Completed requests

Emoji

  • Task: Filter should detect anonymous edits which add emoji and prevent the edits. Filter should add a tag and/or warn for similar edits from logged in users. (They could be legitimately used on talk pages.)
  • Reason: Anonymous editing on mobile devices has been turned on,[34] and we are now seeing vandalism that uses emoji.[35]

- Kaldari (talk) 00:01, 8 April 2015 (UTC)[reply]

Emoji has long lists of unicode blocks. Should all of those be checked or only some of them? (A specific list of code ranges would be useful here.) Do any of these have legitimate uses? For my part, my browser fails to render most of those symbols. Dragons flight (talk) 00:55, 8 April 2015 (UTC)[reply]
Five main ranges from emoji: U+1F300..U+1F5FF, U+1F600..U+1F64F, U+1F680..U+1F6FF, U+2600..U+26FF, U+2700..U+27BF Dragons flight (talk) 17:02, 8 April 2015 (UTC)[reply]
Testing as Special:AbuseFilter/680. Dragons flight (talk) 17:48, 8 April 2015 (UTC)[reply]
What a cute regex: [🌀-🙏🚀-🛳☀-➿] :) Kaldari (talk) 17:53, 8 April 2015 (UTC)[reply]
For me that displays as "unprintable square"-"unprintable square", "unprintable square"-"unprintable square", "sun icon"-"unprintable square". You don't happen to know what fonts actually include these things? I'm a little impressed to see 50 hits in the log during only the first 8 hours. Dragons flight (talk) 02:42, 9 April 2015 (UTC)[reply]
I've added a warning now. So far, the only edit I've noticed that might be considered legitimate was [36] where ★ was used in an album name. Dragons flight (talk) 18:55, 9 April 2015 (UTC)[reply]
I've removed ★ and ☆ which seems to be getting a number of false positives, mostly from stylized music titles in Asian languages (e.g. Japanese), but don't seem to be used for vandalism in the current sample set. Dragons flight (talk) 18:00, 13 April 2015 (UTC)[reply]
 Done Seems to be working well. Sam Walton (talk) 11:57, 25 May 2015 (UTC)[reply]

Ankur Mazumder filters

  • Task: This filter should be able to catch the prolific sockpuppeteer Ankur Mazumder (talk · contribs) who persistently creates autobiographies about himself (example of such a page) in both userpages of his socks (usually) and in the article space (not as often; deleted examples Ankur Mazumder, Www.Ankurmazumder.com, and Computer Field Of Ankur). In addition, another filter should be added to detect additions of his mention into articles like Hooghly Collegiate School by IP socks of this sockmaster.
  • Reason: Adding these filters and setting them up to just log such actions would greatly assist us in finding and blocking socks of this user. Currently, one needs to do a lot of searching to come across socks, and in some cases it may take long before the sock is discovered.

- Gparyani (talk) 05:07, 18 March 2015 (UTC)[reply]

@Gparyani: Are all pages exactly the same? If not could you provide a few more samples so I can find similarities, I'm not an admin so I can't see the deleted pages. PhantomTech (talk) 22:49, 20 March 2015 (UTC)[reply]
@PhantomTech: Unfortunately, most of the pages he's created have been deleted. However, there may be another one laying around in old revisions of the userpages of the sockpuppets. It would be great if an admin could see this request. Gparyani (talk) 23:17, 20 March 2015 (UTC)[reply]
I've pasted some deleted userpages here: User:Anna Frodesiak/Blue sandbox. They are the last few socks and the original master userpage. Anna Frodesiak (talk) 23:33, 20 March 2015 (UTC)[reply]
Thanks, ThePhantomBot is being setup to detect the page creations and will begin logging them in the next few days, whenever it is running. PhantomTech (talk) 00:07, 21 March 2015 (UTC)[reply]
What about the other part? Did you read my last sentence in the task section? Gparyani (talk) 00:13, 21 March 2015 (UTC)[reply]
I missed the last part but I'll add that too. PhantomTech (talk) 00:40, 21 March 2015 (UTC)[reply]
@PhantomTech: As an edit filter, or to your bot? Gparyani (talk) 00:46, 21 March 2015 (UTC)[reply]
@Gparyani: To the bot, I don't have the permission for edit filters and I'm not sure that this is frequent enough to have an edit filter made. PhantomTech (talk) 00:58, 21 March 2015 (UTC)[reply]
Created at Special:AbuseFilter/673 as log only for now. Sam Walton (talk) 21:46, 29 March 2015 (UTC)[reply]
@Samwalton9: You may also want it to check for userpage creations, which is more important. Gparyani (talk) 02:50, 3 April 2015 (UTC)[reply]
Wasn't aware, thanks, now checking user space too. Sam Walton (talk) 09:06, 4 April 2015 (UTC)[reply]
 Done Seems to be working fine so I've switched it to disallow the edits. Editors who attempt to make the edits should still be double checked. Sam Walton (talk) 22:20, 18 May 2015 (UTC)[reply]

"<User X> IS A PROSTITUTE" sockpuppets

- TL22 (talk) 18:31, 17 May 2015 (UTC)[reply]

Make that 9, 6 pointed at me, 1 at User:Geraldo Perez, another one at User:EvergreenFir and another one at User:Joseph2302. --TL22 (talk) 22:54, 17 May 2015 (UTC)[reply]
 Already done by Callanecc. See the SPI I linked above. Self-closing request. --TL22 (talk) 10:41, 18 May 2015 (UTC)[reply]

Denied requests

  • Task: Check whether {{navbox}} was directly added in article namespace like [37], [38], [39].
  • Reason: To replace with the correct template. It messes code for good by adding wrong categories, noinclude tags, etc.

- Magioladitis (talk) 07:52, 13 April 2015 (UTC)[reply]

@Magioladitis: I don't think this is adequate for an edit filter. If you want to track this happening you could search for "How to manage this template's initial visibility", which appears to show where this has happened. Sam Walton (talk) 11:57, 25 May 2015 (UTC)[reply]

Ref desk troll

  • Task: Prevent IP addresses and new accounts (not yet autoconfirmed) from blanking sections of the Ref Desk. Recommend "disallow and report" so we can quickly block accounts.
  • Reason: See Wikipedia:Sockpuppet investigations/CapriSon33333 for some background. A few weeks ago, a regular ref desk troll, who had been gone for some time, was blocked again for his trolling questions. He took exception to this, and started massive DDOS-like attacks against the Ref Desks by rapidly and repeatedly blanking large sections of them as soon as they aren't semi-protected. He has a massive sock puppet farm and apparently nothing better to do, because within minutes of being unprotected, he starts his rapid attacks which are impossible to control. We don't want the ref-desks to be indefinitely semiprotected, but that's all we've been able to do to contain him. A report at ANI caught the attention of a checkuser, who said he had a rangeblock he thought would work. Apparently, it didn't. The next step seems to be to try an edit filter. I'm not so good on the technical aspects of writing an edit filter, but this doesn't seem like too hard of a filter to code for, just prevent any removal of text from the reference desks from non-autoconfirmed accounts and IPs, and report any accounts or IP addresses that repeatedly attempt to do so. Thanks again for your prompt attention to this!

- Jayron32 21:03, 17 April 2015 (UTC)[reply]

Actually, there is also the action "block", though it is very rarely used. --TL22 (talk) 01:20, 18 April 2015 (UTC)[reply]
The abuse filter extension supports a "block" action, but that action is not presently enabled on enwiki. Dragons flight (talk) 23:14, 18 April 2015 (UTC)[reply]
I'm fine with manually blocking when it comes up, human eyes would cut down on false-positives anyways. But we need some means of shutting this guy down so we can re-open the ref-desks to IPs with legitimate questions. --Jayron32 01:24, 19 April 2015 (UTC)[reply]
Testing as filter 683. עוד מישהו Od Mishehu 18:07, 23 April 2015 (UTC)[reply]
@Od Mishehu and Jayron32: Disabled as none of the hits from the past month were removal of a section. Sam Walton (talk) 11:33, 25 May 2015 (UTC)[reply]
Thanks. It seems to have died down. Let's keep this on in our pockets in case he returns. --Jayron32 00:26, 26 May 2015 (UTC)[reply]

Persistant spam/ sockpuppetry username

@Joseph2302: Is this a long term issue? I'm only seeing the two usernames. Sam Walton (talk) 13:46, 25 May 2015 (UTC)[reply]
there's been 3 over the space of a week or so. Is this only for long-term issues (like 20+ users) then? If so, my mistake and feel free to decline thid. Joseph2302 (talk) 13:51, 25 May 2015 (UTC)[reply]
Yeah, wait for it to become a long term issue before requesting an edit filter, it's entirely possibly this will stop in the next week. Sam Walton (talk) 13:59, 25 May 2015 (UTC)[reply]

Kenny Loggins hoaxes

  • Task: Disallow a range of IPs from editing any article containing the surname Loggins, and disallow the IPs from adding Loggins to an article.
  • Reason: See Wikipedia:Long-term abuse/Kenny Loggins vandal. For a couple of weeks now, IP6s have been brought to bear on various Kenny Loggins-related pages. All of the IP6s start with 2602:306:BD7E:CAA0, so this string could be filtered along with the surname Loggins. Or a rangeblock can be set, but there will be collateral damage to others who are innocent.

- Binksternet (talk) 16:32, 10 May 2015 (UTC)[reply]

 Not done It seems a rangeblock has been placed. Please re-request if this doesn't solve the problem. Sam Walton (talk) 11:24, 25 May 2015 (UTC)[reply]

David Beals

  • Task: Block all edits with summaries that start with "goo.gl/", maybe report the user, too?
  • Reason: Sockpuppets of LTA David Beals has been using this tactic to spam videos of ceiling fans.

- Ian.thomson (talk) 23:29, 16 April 2015 (UTC)[reply]

Still going. This is the sort of thing edit filters were made for, right? Ian.thomson (talk) 23:00, 28 April 2015 (UTC)[reply]
 Not done David Beals regularly evades the edit filter. Reaper Eternal (talk) 19:17, 29 April 2015 (UTC)[reply]

Russian Beatles copyvio

  • Task:A highly dynamic IP and an occasional named sock account (FreedomRome (talk · contribs) is the latest sock of the master Crazy1980 (talk · contribs)) has been plaguing the OTRS noticeboard and other pages with spurious statements of permission to upload Beatles songs or to link to a Russian copyright-violating host. Typically includes a link to britishcouncil.org, which I'd rather not have blacklisted the usual way, and statements like "Team of the volunteers of the British Council gives the permission to use these materials", "The Beatles for Cultural Diversity" and all IPs are from Russian ranges. See [40], [41], [42], [43], [44] and so on.
  • Reason: Obviously Apple Records, Paul McCartney, Ringo Starr and the estates of George Harrison and John Lennon will disagree with the idea that somebody at the British Council has given permission to Russians to upload Beatles songs to WP or to link to Russian copyvio sites, and given the incredible persistence of this user, a filter might be helpful - it's been going on for months. No actual uploads that I know of, but the spurious permissions are becoming tiresome. I believe the copyvio host has long since been blacklisted.

- Acroterion (talk) 01:58, 31 May 2012 (UTC)[reply]

@Acroterion: Is this still needed? — This, that and the other (talk) 12:58, 18 July 2013 (UTC)[reply]
I haven't seen any activity of this kind in the past few months, so I'd let it drop unless they start up again. Acroterion (talk) 13:35, 18 July 2013 (UTC)[reply]
Update: they're back, Russian IPs on Jimbo's talkpage, so I'd like to pursue this again. See [45], [46] and [47] and User:Music1245's contributions. Acroterion (talk) 00:43, 15 August 2013 (UTC)[reply]
@Acroterion: Marking stale, if the vandal is still active remove the tag. PhantomTech (talk) 06:06, 20 March 2015 (UTC)[reply]
 Not done per inactivity. Sam Walton (talk) 12:06, 25 May 2015 (UTC)[reply]

Retard

  • Task: Filter redirects to pages like Mental retardation.
  • Reason: Very likely an attack page.

- FrankDev (talk) 02:41, 16 January 2013 (UTC)[reply]

Hmm, I think this could be a good filter, but could cover more than just this term. I'd propose writing up a list of pages that would be offensive if redirected to and go from there. Sam Walton (talk) 19:23, 22 January 2015 (UTC)[reply]
Not sure how useful this is going to be but trialling at Special:AbuseFilter/679. Sam Walton (talk) 22:47, 7 April 2015 (UTC)[reply]
Not very useful apparently, so  Not done. Sam Walton (talk) 21:57, 18 May 2015 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Edit_filter/Requested&oldid=666063554"