Talk:Spam blacklist/Archives/2017-02

Add topic
Active discussions

Proposed additions

  This section is for completed requests that a website be blacklisted|words)

We are getting spam attempts, and maybe success (uncertain) of with urls containing and I cannot think that we are going to want to widely use these if ever use them. In case others can see their valued use, I am tempted to block them to get around some of these other spamming attempts.  — billinghurst sDrewth 23:56, 25 January 2017 (UTC)

(added tracking so we know where we discussed parts)
I don't see much proper use for this that could not be handled with a bit of local whitelisting. Comes with the /url and /cse .. in a way, there is not much reason to link to google in mainspace at all except for the dedicated articles ... --Dirk Beetstra T C (en: U, T) 08:03, 26 January 2017 (UTC)
@Billinghurst:   Added to Spam blacklist. -- — billinghurst sDrewth 12:09, 11 February 2017 (UTC)

Spambots. MER-C (talk) 12:14, 21 February 2017 (UTC)

@MER-C:   Added to Spam blacklist. --Dirk Beetstra T C (en: U, T) 12:42, 21 February 2017 (UTC)

URL shortening service, spammed on --Versageek (talk) 18:09, 23 February 2017 (UTC)

@Versageek:   Added to Spam blacklist. -- — billinghurst sDrewth 03:26, 24 February 2017 (UTC)


Contains false information about Kepompong (a TV series which was actually aired on SCTV). Ref: 1 ArdiPras95 (talk) 23:51, 23 February 2017 (UTC)

@ArdiPras95:   Declined perceived inaccuracy is not a criteria for the global blacklisting; ideally it would be managed locally if problematic.  — billinghurst sDrewth 09:55, 27 February 2017 (UTC)

url shorteners 1, Feb 27

all in spam attempts  — billinghurst sDrewth 09:56, 27 February 2017 (UTC)

@Billinghurst:   Added to Spam blacklist. -- — billinghurst sDrewth 09:57, 27 February 2017 (UTC)

url shorteners 2, Feb 27

 — billinghurst sDrewth 20:19, 27 February 2017 (UTC)

  Added to Spam blacklist. -- — billinghurst sDrewth 20:20, 27 February 2017 (UTC)

url shorteners 2, Feb 27

 — billinghurst sDrewth 20:19, 27 February 2017 (UTC)

  Added to Spam blacklist. -- — billinghurst sDrewth 20:20, 27 February 2017 (UTC)

url shorteners 2, Feb 27

 — billinghurst sDrewth 20:25, 27 February 2017 (UTC)

  Added to Spam blacklist. -- — billinghurst sDrewth 20:25, 27 February 2017 (UTC)

url shorteners 2, Feb 27

 — billinghurst sDrewth 20:25, 27 February 2017 (UTC)

  Added to Spam blacklist. -- — billinghurst sDrewth 20:25, 27 February 2017 (UTC)

url shorteners

Defender (talk) 08:47, 28 February 2017 (UTC)

  Added to Spam blacklistDefender (talk) 08:48, 28 February 2017 (UTC)

Proposed removals

  This section is for archiving proposals that a website be unlisted.

Block site

I was wondering if the following link may be unblocked. We wrote a pages about education site

You can check our site it is normal and help people to teach how to type. We have mention about our site in Wiki

Ratatype is FREE online typing tutor! And it helps people to learn touch typing and save their time on typing. I think people need now about useful free resources like Ratatype. As for links about Ratatype:,, — in this article (in section Sources and Citations) they use our research and page http://www.ratatype/learn/ — use Ratatype as one of the sources — use our research and article http://www.ratatype/learn/average-typing-speed/ — they use our research and article http://www.ratatype/learn/average-typing-speed/ — they write about Ratatype as one of the touch typing resources Also, you can see links to on Wikipedia in different languages: Ru



  Declined subject of heavy spam e.g.. Vituzzu (talk) 14:26, 13 February 2017 (UTC)
  Declined Dirk Beetstra T C (en: U, T) 03:42, 14 February 2017 (UTC)

*remove* 18:32, 26 February 2017 (UTC)

  Declined It is an advertisement, why would we want those tens of links that you did to the Wikipedias? That is not the purpose of an encyclopaedia.  — billinghurst sDrewth 01:47, 27 February 2017 (UTC)

This is a website about a village/winery, but the page at has useful information which I hoped to add as a reference to, which is currently being discussed at I've commented out the ref for now, to be able to save the rest of my edit, but would like to include it. PamD (talk) 21:58, 27 February 2017 (UTC)

@PamD: You can find it was blocked from this request. Usually we say request a local whitelist for problematic domains, and from there we can check whether abused and then look to a universal removal. As it is Beetstra's addition, worthwhile checking with him.  — billinghurst sDrewth 11:23, 28 February 2017 (UTC)
@PamD: Since you only want it for one page on one Wiki, I would suggest indeed to request whitelisting (I tend to be conservative through the experience that spammers hardly ever stop). --Dirk Beetstra T C (en: U, T) 11:42, 28 February 2017 (UTC)
@Beetstra: OK, thanks - but where do I go, apart from this page, to request Whitelisting? I've had no dealings with the spam listing setup before. PamD (talk) 12:14, 28 February 2017 (UTC)
@PamD: sorry, forgot to mention: w:en:WT:SWL (it is linked from the message that you encounter when you save a page with a blacklisted link). --Dirk Beetstra T C (en: U, T) 12:20, 28 February 2017 (UTC)
@Beetstra: Thanks. It's not a very editor-friendly system - I didn't keep a copy of the message I got when I tried to save the page, just went back and commented out the problematic link so I could save the rest of the edit. And now it's taken me trips to three different pages to leave a request in the right place - and even there, the documentation isn't wonderful (no indication that I need to make the first line a level 3 heading, till I looked and realised it didn't match the rest; had to open two copies of the page to be able to see the instructions and the place where I was editing ...) ah well, I hope my request will be granted after all this! PamD (talk) 12:58, 28 February 2017 (UTC)
  Done. --Dirk Beetstra T C (en: U, T) 13:08, 28 February 2017 (UTC)

Troubleshooting and problems

  This section is for archiving Troubleshooting and problems.


  This section is for archiving Discussions.

Unblocking YouTube's redirection and nocookie domains

This section was archived on a request by:  — billinghurst sDrewth 12:03, 11 February 2017 (UTC)

  • --- Note past closure - a vandal kept posting pictures and linking to videos of ceiling fans - that's why the bottom two are blocked. Kernosky (talk) 13:58, 24 April 2016 (UTC)

Apparently youtu(dot)be and youtube-nocookie(dot)com, both of which are official YouTube domains owned by Google, are on this blacklist. For over ten years, the SpamBlacklist MediaWiki extension has loaded this blacklist on third-party wikis, big and small. This is quite an issue for third-party sites such as ShoutWiki, a wiki farm, since SpamBlacklist doesn't currently have the concept of "shared" whitelists — blacklists can be shared (loaded from a remote wiki), whitelists cannot. Given that the main YouTube domain isn't blocked, and also that YouTube itself hands out youtu(dot)be links, I don't think that "but it's a redirecting service" is a valid argument against it, and therefore I'd like to propose removing these two entries from the blacklist. --Jack Phoenix (Contact) 23:17, 29 August 2015 (UTC)

There are several links on youtube blacklisted here on Meta, as well many, many on local wikis. Youtube has videos that get spammed, there are videos that should simply not be linked to. Leaving open the redirects then makes the issue that not only the link needs to be blacklisted, but also all redirect to those links. That gives either extra work to the blacklisting editors, or leaves the easy back-door open. On wikis it leaves more material to check. That in combination with that redirect services are simply never needed, there is an alternative. Additionally, Wikipedia has their built-in redirect service which also works (I mean templates, like {{youtube}}).
That there is no meta-analogue of the whitelist is a good argument to push that request of years ago to re-vamp the spam-blacklist system through and have the developers focus on features that the community wants, and certainly not an argument for me to consider not to blacklist something. Moreover, I do not think that the argument that it hampers third-party wikis is an argument either - they choose to use this blacklist, they could alternatively set up their own 'meta blacklist' that they use, copy-pasting this blacklist and removing what they do not want/need.
The problem exists internally as well, certain of our Wikifarms do allow for certain spam, which is however inappropriate on the rest of the wikifarms, and on the majority by far (in wiki-volume) of the wikis. That also needs a rewriting of the spam-blacklist system, which is crude, too difficult. A light-weight edit-filter variety, specialised on this would be way more suitable. --Dirk Beetstra T C (en: U, T) 04:05, 30 August 2015 (UTC)
  • Oppose unblocking for the reasons given above. Stifle (talk) 08:32, 21 October 2015 (UTC)
  Declined  — billinghurst sDrewth 06:22, 24 January 2016 (UTC) can only be used for, so it's no redirecting service, so remove it from the black list. If you need to block certain yt video (what I consider btw as a little stupid) just update that system and include as well as that's it.
Djamana (talk) 20:08, 2 February 2016 (UTC)
@Djamana: Why do you consider that blocking of a specific YouTube video a little stupid? --Dirk Beetstra T C (en: U, T) 04:11, 3 February 2016 (UTC)
@Djamana: I did not see the above list earlier - out of the 5 on the meta spam blacklist there are still 3 active videos. Those 5 were abused and for the one of the cases where I was involved in (still active), that was pretty persistent promotion. I doubt that these need to be removed. The two that are specifically not there anymore could indeed be removed (or maybe they need to be corrected ..), still leaving 3. Moreover, these are not the only rules blocking youtube, also the many individual wikis have specific youtube videos blacklisted (and youtube can be used to earn money (and those are known to circumvent the blacklist; even regulars do!), and there is information there that simply should NEVER be linked to ..). Again   Declined. --Dirk Beetstra T C (en: U, T) 06:28, 14 April 2016 (UTC)

Partial matches: <> blocks <>

This section was archived on a request by:  — billinghurst sDrewth 12:04, 11 February 2017 (UTC)

I tried to add a link to <>, and was told that I couldn't add the link, as <> was blacklisted. Is this partial-match blacklisting (based, I guess, on an incorrect interpretation of URL specifications) a known bug? Cheers. --YodinT 15:46, 21 October 2015 (UTC)

This is more of a limitation to the regex, we tend to blacklist '\bchange\.org\b', but a '-' is also a 'word-end' (the \b). I'll see if I can adapt the rule. --Dirk Beetstra T C (en: U, T) 07:46, 22 October 2015 (UTC) is not here, it is on en.wikipedia. That needs to be requested locally and then resolved there. --Dirk Beetstra T C (en: U, T) 07:48, 22 October 2015 (UTC)
Thanks for looking into this; is it worth replacing the regexes globally to fit URL specs? I'm sure I'm not the only one who will ever be/have been affected. --YodinT 11:27, 22 October 2015 (UTC)
@Yodin: Sorry, but there are no global regexes to replace, is only blacklisted on en.wikipedia. You'll have to request a change on en:MediaWiki talk:Spam-blacklist (so there is a local request to do the change, then I or another en.wikipedia admin will implement it there). --Dirk Beetstra T C (en: U, T) 11:38, 22 October 2015 (UTC)
Thanks Dirk; just read this (sorry for the repeat on regexes there!). Isn't the main blacklist here also using '\bexample\.com\b'? I can come up with the general case regex if you like! --YodinT 11:44, 22 October 2015 (UTC)
You mean for every rule to exclude the '<prefix>-'-rule (i.e. put '(?<!-)' before every rule in the list - well, some of them are meant to catch all '<blah>' sites, so that is difficult. And then there are other combinations which sometimes catch as well. It is practically impossible to rule out every false positive. --Dirk Beetstra T C (en: U, T) 12:01, 22 October 2015 (UTC)
I see... much more complicated in practice than I thought. My idea was to apply it to a wider class of false positives, including the '<prefix>-' rule and more, by replacing "\b" with a regex rule which covers all and only the unreserved URI characters (upper & lowercase letters, decimal digits, hyphen, underscore, and tilde; with "dots" used in practice as delimiters). But this wouldn't cover the '<blah>' examples you gave, and having read some of the maintenance thread below which covers false positives, I won't try to press the issue! Maybe one day? Until then, I hope this goes well! Cheers for your work! --YodinT 12:26, 22 October 2015 (UTC)
@Yodin: If the foundation finally decides that it is time to solve some old bugzilla requests (over other developments which sometimes find fierce opposition), and among those the ones regarding overhaul of the spam-blacklist system, then this would be nice 'feature requests' of that overhaul. In a way, stripping down the edit-filter to pure regex matching 'per rule', with some other options added (having a regex being applied to one page or set of pages; having the regex being excluded on one page only, having the whitelist requests being added to the blacklist rule they affect, whitelisting on one page or set of pages, etc. etc.) would be a great improvement to this system. --Dirk Beetstra T C (en: U, T) 14:25, 23 October 2015 (UTC)
  Closed nothing to do, a block at enWP, nothing global.  — billinghurst sDrewth 09:45, 22 November 2015 (UTC)

Free domain names

This section was archived on a request by:  — billinghurst sDrewth 12:06, 11 February 2017 (UTC)

.ml, .ga, .cf, and .gq offer free domain names [1]. I'm sick of playing whack-a-mole with the TV show spam; is there anything else we can do? MER-C (talk) 13:38, 8 April 2016 (UTC)

@MER-C: could easily be blacklisted, provided that there is not too much regular material that needs to be linked on those sites. What countries do these belong to? --Dirk Beetstra T C (en: U, T) 06:38, 20 April 2016 (UTC)
For .ml we have 1581 links on en.wikipedia. It looks the majority of that are and and similar (many used as references). --Dirk Beetstra T C (en: U, T) 06:41, 20 April 2016 (UTC)
.gq looks like low hanging fruit: 21 links on en.wp, only half of which are in mainspace. MER-C (talk) 12:29, 2 May 2016 (UTC)
Less than half I would say, however some of those are 'official' (I see the registrar itself, and a university). Moreover, this blocks more than only en.wikipedia (though a quick check on other wikis does not enlarge the set of genuine links too much). If we write the rules so that the (large majority of the) currently used 'good' subdomains (on, say, the 5 major wikis) are exluded, I'll pull the trigger. --Dirk Beetstra T C (en: U, T) 13:11, 2 May 2016 (UTC)
@MER-C: are we getting much spam outside of enWP? If the spam is centred on enWP, can we try the local blacklist there initially?  — billinghurst sDrewth 05:11, 23 May 2016 (UTC)
The situation is mostly under control on enWP -- we now have a private abuse filter in front of the spam blacklist which works fairly well. They've now turned to spamming via facebook, which is something I struggle to care about. Blocking these domains isn't necessary at this moment, but one sees parallels with the .tk and situation. MER-C (talk) 07:10, 27 May 2016 (UTC)

google and springer together

This section was archived on a request by:  — billinghurst sDrewth 12:01, 11 February 2017 (UTC)

I do not understand regular expressions at all. On the Finnish Wikipedia, the following link was blacklisted:

Any ideas why? The IP in question has reported the problem on our local admin noticeboard, but I cannot help them. --Pxos (talk) 20:17, 13 April 2016 (UTC)

Pxos (talk · contribs), they should link directly to the Springer website. Google redirect links include link tracking. John Vandenberg (talk) 10:11, 19 April 2016 (UTC)
@Pxos: - use - this link is copied from the search-result page of google, it is not the actual link to the document. --Dirk Beetstra T C (en: U, T) 06:37, 20 April 2016 (UTC)
  Closed direct links usable, redirecting links blocked.  — billinghurst sDrewth 15:08, 23 May 2016 (UTC)
@Pxos: at dewiki in the blacklisting message de:MediaWiki:Spamprotectiontext we included a link to which simply converts such google-url to the original urls. Maybe this could be an option for you, too? -- seth (talk) 19:23, 14 September 2016 (UTC)

Escaping dot in regex

This section was archived on a request by:  — billinghurst sDrewth 12:00, 11 February 2017 (UTC)

  • [a-z]\b

Should the . at the beginning of .com be escaped? – JonathanCross (talk) 15:01, 25 April 2016 (UTC)

@Billinghurst: looks like you added the regex in this revision. It was based on User:COIBot/XWiki/ which suggests escaping the dot. – JonathanCross (talk) 14:49, 14 May 2016 (UTC)

The regex has been removed. Not sure there is any ongoing issue unless we get spam again.  — billinghurst sDrewth 03:45, 15 May 2016 (UTC)
Ah, great, thanks! – JonathanCross (talk) 15:31, 15 May 2016 (UTC)


This section was archived on a request by:  — billinghurst sDrewth 11:58, 11 February 2017 (UTC)

Why is the list not alphabetical, so I can look up whether a certain site is listed and then also look up when it was added? --Corriebertus (talk) 08:55, 21 October 2015 (UTC)

there are advantages and disadvantages of a alphabetical list. for example it would be very helpful to group all url shorteners at one place (see discussion thread above). sometimes it's better to have a chronological list. additionally to that regexp can't be really sorted domain-alphabetically.
if you want search the blacklist, you can use a tool like -- seth (talk) 17:16, 30 October 2015 (UTC)
Because no one has done it. It is not something that I will spend my time doing.  — billinghurst sDrewth 10:59, 3 December 2016 (UTC)

This section was archived on a request by:  — billinghurst sDrewth 03:39, 24 February 2017 (UTC)

Whitelist the website - Jonathan629 (talk) 13:10, 30 January 2017 (UTC)Jonathan629=== Hello,

I'm very new in editing. And i've drafted my first page about Gokul Shrinivas, 3 National award holder in the field of Inventions. There are lot of media profiles was available, so I thought of collecting it all and write a page about him and thus I'm the first editor to write about him

A month before, REJI KUMAR PILLAI, President of India Smart Grid forum wrote an article about Gokul Shrinivas, I tried to include that link in the page, but there's some problem reflecting as blacklisted. Could you please unblock this link which would be really helpful for me

I really don't have much idea that posting here will unblock or not. If someone sees this article, please unblock it.

Jonathan629 (talk) 13:10, 30 January 2017 (UTC)jonathan629

@Jonathan629: Firstly it is not globally blocked, it is locally blocked at English Wikipedia, so any request will need to be there. Secondly, I believe that it has been blacklisted as an abused or abusable source. From the linksummary template "find entry" link you can check the status by adding "en" into the local searchbox. These links are the pertinent discussions that I could find at enWP
  no global block  — billinghurst sDrewth 20:58, 31 January 2017 (UTC)

External repositories

I recently found's notCredible list. I think this, along with similar projects, could be a good addition to the list.

Also, it would be great if the updates could be automated.--Strainu (talk) 16:05, 10 February 2017 (UTC)

It hasn't been our approach to proactively seek out lists of problematic sites, it has instead been reactive to problematic editing, typical examples are the redirect urls where we wait until they are spammed. To do that I would be prefer to see an RFC at meta and something notified to the wikis. Such a change needs to have a demonstrated value. Do we have an indication that we are being abused? Maybe we are better talking to @Beetstra: on whether we can utiise COIBot to simply add each domain to the bot's monitor list, at least as part of the proof of problem needing to be managed.

*If* were going to change to pro-active generated lists, I would like to see a re-architecture of how the blacklist is generated. It would be more useful to keep all such lists as their own entities and have a means to concatenate them at list generation time. I look at the indication of time that it takes for spamblacklist to roll out and anything that makes that longer and longer with little clear gain.  — billinghurst sDrewth 11:36, 11 February 2017 (UTC)

To note that I have picked 10 clickbait sites and having COIBot run some checks to see what may be there.  — billinghurst sDrewth 11:40, 11 February 2017 (UTC)

Mentioning the sites here, in a LinkSummary template, makes a record in COIBot that they were reported here, and a report is generated. As anything can be spammed, I see little gain in having everything being monitored, even pre-emptively. We know that some spammers will create domains as they need them, and they cannot be pre-emptively monitored. Having a stronger detection mechanism may be nice, but since the linkwatchers are having a pain following everything, I don't think I will have resources to make a stronger mechanism (nor do I have time).

What sDrewth is suggesting is of what we want already for a long, long time: A proper rewrite of the spam blacklist. Developers unfortunately are busy with what WMF thinks is more important.</frustration> What I would like to have is a spam blacklist based on the EditFilter - take the current EditFilter, strip it completely of its interpretation function, and replace it with a simple field with regexes (formatted like in our blacklist). The only test that the system needs to do is to test, for each 'SpamFilter', whether the regexes match against the added external links. You then have the options (already available from EditFilter) to log only, warn, throttle, (warn and block), and block, add a custom custom ('hej <expletive>, you are using a redirect site, please adapt your edit and use the real site'). You can then fancy it up with more options if you like (add whitelisting, wikifamily-selection, namespace-selection, per-page exclusion, etc. etc.). It would give much more flexibility and control to the current spam-blacklist, and pre-emptive monitoring through log-only (or even warning) would be a non-disruptive option. Working a lot here and on en.wikipedia, I am sometimes surprised how much 'good use' there is of some links that nonetheless need to be blacklisted due to uncontrollable abuse. --Dirk Beetstra T C (en: U, T) 03:48, 13 February 2017 (UTC)

@Hoo man: is there value transferring some of this into phabricator request for spam blacklist v. extrapolation of abuse filter? I remember there was some discussion about abuse filter needing some sort of morph, so not sure how we best address this to the developer community.  — billinghurst sDrewth 03:53, 24 February 2017 (UTC)
@Billinghurst: Sorry for the late reply, but yes please, bring this up on Phabricator. We have been talking about this change since at least 2013, but sadly nothing happened since (except for some discussions here and there). Cheers, Hoo man (talk) 10:40, 30 March 2017 (UTC)
This section was archived on a request by: —MarcoAurelio (talk) 11:13, 26 February 2018 (UTC)
Return to "Spam blacklist/Archives/2017-02" page.