Jump to content

Stewards' noticeboard/Archives/2019-02

From Meta, a Wikimedia project coordination wiki
Latest comment: 5 years ago by MarcoAurelio in topic 2019 Steward elections results

Suggesting modification to practice of global abuse filters not blocking

Dear Stewards. I am writing to address the practice of not setting global abuse filters to have a blocking action.

There are a number of xwiki vandals who are purposely exploiting the weaknesses in defences to undertake their vandalism. I would like to propose a change to practice where a temporary (short 2 hour) block can be placed by a global abuse filter where the filter has a demonstrated history of functioning with an acceptable rate of low false positives and is reviewed by a second abuse filter writer as meeting criteria. I would suggest that our undercoding to manage risk continue.

The two hour window allows for a review by users with administrator permissions to check the edits, and escalate the matter to stewards where a global block is required, or undertake other actions as deemed appropriate. This process has been utilised at meta in recent times in certain cases and appears to be worked well.

As part of such a change, I would propose that global sysops are allocated the right abusefilter-view-private to their global group set—a right that I think that we should be looking to allocate irrespective of whether we change practice or not. This right is available to local administrators, and the global sysops go through a similar rigorous process to demonstrate suitable knowledge and skill to utilise this information in their cross-wiki role.

Add that we will need log view too abusefilter-log-private

Thanks for your consideration.  — billinghurst sDrewth 10:39, 2 February 2019 (UTC)

While I generally do not oppose, I am a bit afraid of what you said above about acceptable rate of low false positives. I think that blocking an advanced user in eg. huwiki because they used in their Wikipedia article some phrases that are abusive/promotional in English, but neutral in some other language is not acceptable. I think that non-zero rate of false positives would be not acceptable for automated blocking. Especially, that in case of just disallowing user's action, the user might simply abandon their action because they found it not to be worth investigating why it was considered wrong, so we may even not know that this case is a real false positive. And blocking a user here would be abusive. So creating a blocking global AF needs much more care than eg. placing a phrase on the title blacklist, IMO (even if the reported false positive rate is extremely low). Ankry (talk) 11:52, 2 February 2019 (UTC)
Ankry, for what its worth filters can avoid 'advanced users' (which I'm taking as you meaning an established editor on a project) by skipping users in certain groups (such as autoconfirmed) or with more than x project edits. — xaosflux Talk 13:08, 2 February 2019 (UTC)
Gday Ankry. The global filters that I am specifically addressing
  • are designed to not be catching advanced users, primarily our filters are aimed at spambots, and xwiki LTAs
  • "acceptable rate of low false positives" is whatever the community specifies, and you will note that I am purposefully wanting more than one person's opinion on the filter
  • active global filters that do things here usually progress through a step-up process; and endeavoured to have the minimal possible action, eg. our emoji filter simply tags
  • these mentioned spam filters are clearly designed to be actively used against a highly problematic set of abuse, and actively reviewed, not a set and forget, and to be turned off when not required.
Your points are valid. I am definitely not dismissing your concerns, and why I have raised this issue publicly, rather than directly with stewards. I think we can work within your concerns as that has been our process with global abuse filters for a long period. This is clearly "hasten slowly" stuff.  — billinghurst sDrewth 13:20, 2 February 2019 (UTC)
  • @Billinghurst: I'm fully in support of adding abusefilter-view-private to the GS permission set - this seems more of a procedural oversight when the GS group was made (as GS's have abusefilter-modify access within the GS set already. — xaosflux Talk 13:15, 2 February 2019 (UTC)
    Likely due to its name, and that it gets easily confused with the other non-used access related to 'private information' (where this is just administratively hidden information). — xaosflux Talk 13:17, 2 February 2019 (UTC)
    The global abuse filters came well after the creation of the group, and with the guarded relaxation of the global abuse filters (both to edit and to view), I think that it is more that it wasn't clear that was the next progression step. I truly think it is time.  — billinghurst sDrewth 13:23, 2 February 2019 (UTC)
    That makes sense too - this is an easy add, can't think of any reason to hold it back (and most GS's could qualify for global abuse-filter helpers - but no need to make them go though hoops). — xaosflux Talk 13:28, 2 February 2019 (UTC)
    Yep, and there is one global sysop (and I won't mention her name <stare>) who hasn't got around to applying for AFH and this is my "step back" approach to actually doing this better.  — billinghurst sDrewth 13:34, 2 February 2019 (UTC)
    (ec) Noting the rights abusefilter-view-private allows the filter to be seen and abusefilter-log-private allows the detail in the log to be seen.  — billinghurst sDrewth 13:32, 2 February 2019 (UTC)

Advanced permissions on io.wikipedia and sysops on ie.wikipedia

Hello stewards,

I received a notice two weeks ago about removal of sysop privileges on io.wikipedia (the planned language Ido) where I became one in 2005, if I remember correctly. Around 2010 I switched my main auxlang activity to the language Occidental / Interlingue, which is at ie.wikipedia.org. Also, from 2011 to 2018 I did not have much Wikipedia activity in general as I was working on finding a way to move back to Korea, where I live again, and as a result I use Wikipedia in the way I used to before.

When the notice went up on io.wikipedia about removal of permissions Joao Xavier contacted me on Facebook and expressed his desire that I remain a sysop for the growing Ido Wikipedia, which I am happy to do. I have indeed been watching the recent changes there but Joao is so quick to revert spam that there has been nothing for me to do. I can ask him to repost it here but here's the text in the meantime:

"Hi, Dave! How are you? Today, a steward (a kind of "super-administrator") of Wikimedia posted the text below on your discussion page in Ido Wikipedia. (text of message) That means: if you do not modify anything (I don't know if only correcting a text, for example, will be considered valid) in Ido Wikipedia, they will remove your status as administrator. Until now, I didn't know that they could do that, but as I've read the guy's message, I decided to advice you. Please make a visit ad Ido Wikipedia, and see how much it has developed! The amount of reasonably long articles (longer than 10 kbytes) surpassed 350! Gonçalo Neves helped me a lot by correcting many of them!"

Now on a related note, on ie.wikipedia we still do not have a single sysop which means nobody has been able to delete this page for example. Over there I could take on that role, or maybe someone like RWMuc given his activity though I'm not sure who is interested. Alternatively if there is a task force of people here that can patrol ie.wikipedia to do the same, that would work just as well. Occidental-Interlingue is quite easy to read so it's not hard to navigate through it and patrol for spam.

Any thoughts on the above? I do not mind losing advanced permissions on io.wikipedia if that is the policy but all things considered it is probably better to retain them.

Mithridates (talk) 04:05, 16 February 2019 (UTC)

Note by GS. Identified page at ieWP has been deleted, it had not been tagged for deletion.  — billinghurst sDrewth 06:03, 16 February 2019 (UTC)
Hello Mithridates, thanks for this post. As you can see in the community approved admin activity review policy, all admins/bureaucrats that have been continuously inactive for a period of two years (no edits and log actions for said period) do receive a message and, if we do not receive any reply in aprox. one month, we proceed to remove the permissions. In your case, given that you've replied to the notice, and stated that you wish to keep your permissions and made some edits on the wiki where you were notified, then you'll keep your permissions. Hope that this answers your questions but in any case feel free to ask any further questions you may have. Best regards, —MarcoAurelio (talk) 13:57, 20 February 2019 (UTC)
Hello MarcoAurelio, thank you very much for the quick resolution. Best, Mithridates (talk) 10:14, 21 February 2019 (UTC)
This section was archived on a request by: —MarcoAurelio (talk) 18:52, 12 March 2019 (UTC)

Wikimedia Forum#Partial blocks on Meta-Wiki

Stewards to note this post by S Poore (WMF) about rollout of partial blocks to meta. It has the scope to otherwise manage components currently being undertaken by protection and abuse filters.  — billinghurst sDrewth 01:29, 20 February 2019 (UTC)

This section was archived on a request by: —MarcoAurelio (talk) 18:52, 12 March 2019 (UTC)

Admin Review on Wikivoyage/de

User 4omni, who was reviewed by the stewards due to recent inactivity, is back and still supported by the active community. He should keep his admin rights. See Discussion. -- DerFussi 06:16, 25 February 2019 (UTC)

Please make a request at Meta:RFP. Leaderboard (talk) 10:46, 25 February 2019 (UTC)
Why? The removals were expected to happen by tomorrow, so the user still has their permissions. There's nothing to restore. I've marked this user as kept accordingly and informed the local community that 4omni is certainly keeping the permissions after their reply and community endorsement. Best regards, —MarcoAurelio (talk) 10:51, 25 February 2019 (UTC)
Sorry, misread "reviewed" as "removed". Leaderboard (talk) 11:07, 25 February 2019 (UTC)
This section was archived on a request by: —MarcoAurelio (talk) 18:52, 12 March 2019 (UTC)

2019 Steward elections results

The 2019 Steward elections have ended at 14:00 UTC.

The Election Committee, after verifying the votes, has announced the final results. The following 5 candidates were elected:

  1. Base (talk · contribs)
  2. Einsbor (talk · contribs)
  3. Jon Kolbert (talk · contribs)
  4. Schniggendiller (talk · contribs)
  5. Wim b (talk · contribs)

The Election Committee wishes to thank all of the candidates for their time and interest and the voters for the time spent reviewing the candidates and taking part in this important global process. Also the ElectCom wishes to thank all the other volunteers who helped in the coordination of the election process.

The results of the 2019 Stewards confirmation will be determined in the upcoming days.

For the Election Committee,

This section was archived on a request by: —MarcoAurelio (talk) 18:53, 12 March 2019 (UTC)