Talk:Global sysops/2008 proposal: Difference between revisions

From Meta, a Wikimedia project coordination wiki
Content deleted Content added
Monobi (talk | contribs)
Line 790: Line 790:
:I tend to think that the reality of applying the policy of immediately removing privileges will be a bit slower than immediately; but the real point is simply the fact that the incorrect use of these powers will not be allowed and will be dealt with swiftly. [[User:Dbiel|Dbiel]] 00:58, 15 June 2008 (UTC)
:I tend to think that the reality of applying the policy of immediately removing privileges will be a bit slower than immediately; but the real point is simply the fact that the incorrect use of these powers will not be allowed and will be dealt with swiftly. [[User:Dbiel|Dbiel]] 00:58, 15 June 2008 (UTC)
::Furthermore, local policy may mean that differing degrees of severity are used relating to blocking the accounts of users who use their tools, from accidental use to malicious use. [[User:Daniel|Daniel]] ([[User talk:Daniel|talk]]) 01:36, 15 June 2008 (UTC)
::Furthermore, local policy may mean that differing degrees of severity are used relating to blocking the accounts of users who use their tools, from accidental use to malicious use. [[User:Daniel|Daniel]] ([[User talk:Daniel|talk]]) 01:36, 15 June 2008 (UTC)

== What happened to [[Global rollback]]? ==

Was Global Rollback absorbed with the global sysop? Personally I'd like to have it over the entire sysop bit... at the moment deleting, etc w/ global sysop is excessive and would go unused for most users, while being able to rollback instead of undo (which is epicly slow) would be great. '''[[User:Monobi|Monobi]]''' (''[[User talk:Monobi|talk]]'') 03:50, 15 June 2008 (UTC)

Revision as of 03:50, 15 June 2008

Name

The following discussion is closed: Name is 'global sysop'.

Just to say that I didn't find a better name than a descriptive one. So, if anyone has a better name, let they say it here :) --Millosh 07:22, 31 May 2008 (UTC)[reply]

Double negative may be different in English than in Norwegian, but to me an "Anti-vandal fighter" sounds like someone fighting anti-vandals. And that's not good, is it? Jon Harald Søby 11:39, 31 May 2008 (UTC)[reply]
Actually, I think that we should find one word for this role. Something like keeper or watcher or whatever. I don't prefer aggressive or military terminology, but I am not a native English speaker and it is not so easy to me to find a better word. --Millosh 12:10, 31 May 2008 (UTC)[reply]
And, as you may see in this article, double negative is usual in Slavic languages :) --Millosh 12:10, 31 May 2008 (UTC)[reply]
Certainly not a fan of "Anti-vandal" as a name. As with the word "spam" is leads to pointless arguments & misconceptions. For some time now I have considered I am involved with dealing with disruptive editing. I stress I'm not looking for "group of people who deal with disruptive editing across more than x wikis" as a title :) However it should be something that encompasses the various forms of disruptive editing - I agree with Millosh - hard to find a good word --Herby talk thyme 12:40, 31 May 2008 (UTC)[reply]

We should find the right name for this role. As I said above, I don't like this name. Please, add your suggestions here in alphabetic order. Suggestions will be gathered for ~ one week (until June, 7th) and we will vote from June 8th, 2008 to June 14th, 2008. Name with the most supporting votes will be the name for the role. So, vote from June 8th only if you are in favor of one or more names (so, no opposing votes). without sense now --Millosh 12:29, 31 May 2008 (UTC)[reply]

I'm not a fan of "anti-vandal fighter" either. "Global maintenance" or "Cross-wiki maintenance" perhaps?  – Mike.lifeguard | @en.wb 12:48, 31 May 2008 (UTC)[reply]
Definitely a step in the right direction. "Global rollback" are the rights (?) so something "Global ***" seems good --Herby talk thyme 12:52, 31 May 2008 (UTC)[reply]
I would like to have one word name for them :) As other roles have that. --Millosh 12:56, 31 May 2008 (UTC)[reply]
We call this role "janitor" on Wikia. It fits well with the idea that adminship is no big deal and that these people are here to help clean up, not to rule the wiki or make policy decisions. Angela
OK. One word, so I agree :) --Millosh 13:48, 31 May 2008 (UTC)[reply]
"Global janitor"? It emphasizes that this is a global role, which is new for WMF projects.  – Mike.lifeguard | @en.wb 16:26, 31 May 2008 (UTC)[reply]
"globetrotter"? :) I really prefer one word. --Millosh 22:35, 31 May 2008 (UTC)[reply]
Or even just "traveler" or something similar? --Millosh 22:50, 31 May 2008 (UTC)[reply]
Oppose Oppose "janitor" is aready in use in the Wikiversity. Its a bad idea to label two different jobs with one common name. --Purodha Blissenbach 13:09, 4 June 2008 (UTC)[reply]

I'm not too sussed about the name, although I agree it probably should be addressed before the permission itself goes live. "vandalism cleaner" ?? ++Lar: t/c 18:46, 31 May 2008 (UTC)[reply]

Janitor seems fine to me. Majorly talk 22:39, 31 May 2008 (UTC)[reply]
  • global helper; global assistant; because I prefer something positive, a fighter is something aggressive, but users in this group should just help out, also vandal is sometimes not the right word, not everything is vandalism, even if it looks like it sometimes. (global dirt sweepers ;) ) --birdy geimfyglið (:> )=| 22:54, 31 May 2008 (UTC)[reply]
    • I am asking for one word because it would be much more easy to refer to those contributors much more easily by saying "we need help from janitors|globetrotters|travelers|custodians|keepers..." than "we need help from someone from the group of global dirt sweepers" :) Also, if we don't give to them a short name, they will be called colloquially different; something like "globallers" or so. Some abbreviation or something similar may be good, too. --Millosh 23:09, 31 May 2008 (UTC)[reply]
  • Janitor and Helper is already taken by wikia sadly, and they do exactly the same thing but I think we need a unique name for the WMF wikis and right now, I have only 1 idea:
  • Nomads - "nomads, are communities of people that move from one place to another, rather than settling down in one location." which is what these group of people will be doing (can be used as Global nomads to)..--Cometstyles 23:02, 31 May 2008 (UTC)[reply]
Actually, I really like "nomads" :D  – Mike.lifeguard | @en.wb 23:10, 31 May 2008 (UTC)[reply]
Nomads is meaningless to me (and possibly most people). The name should at least be obvious as to what the job entails. Majorly talk 23:13, 31 May 2008 (UTC)[reply]
I like nomads, too :) --Millosh 23:32, 31 May 2008 (UTC)[reply]
"Taken by Wikia"? There is no ownership of a single name. And, that Wikia uses that name for the same purpose, that is to me a very good argument that we should do it too. Why purposefully add another name to the same function when we can keep things consistent over the wikiworld? Jon Harald Søby 11:18, 1 June 2008 (UTC)[reply]

How about monitor? Since the members of this group will be "monitoring" small wikis, it makes sense. Majorly talk 23:16, 31 May 2008 (UTC)[reply]

Sounds good to me. --Millosh 23:29, 31 May 2008 (UTC)[reply]
I like it. Cbrown1023 talk 00:47, 1 June 2008 (UTC)[reply]
Not bad, just add a Global in front and it won't sound cabalish ;) ...--Cometstyles 00:55, 1 June 2008 (UTC)[reply]
I love nomad or traveller & the fact that no one understand it would be a big plus - they won't come trying to get the rights themselves.
More seriously "global monitor" sounds pretty good (I see no harm in two words one of which is global) --Herby talk thyme 12:55, 1 June 2008 (UTC)[reply]

These are users with global sysop access, so why not call them "global sysop" or "backup sysop"? This is unambiguous and draws a parallel to the local group (so users don't need to read technical pages to know what sort of access they have), and removes the artificial limits imposed by names like "anti-vandal" (are they prohibited from dealing with other tasks, like updating the interface on a wiki with no administrators?). —{admin} Pathoschild 19:36:34, 01 June 2008 (UTC)

I like "backup sysop" and "global janitor" - how about "auxilliary sysop"? To me, "monitor" has a slight big brother is watching you feel to it... WjBscribe 09:44, 2 June 2008 (UTC)[reply]
Not sounding to "Militarish", but "Reserves" also sounds like a good name, reserves are backups and are called in when the the main team can't handle the pressure, probably sounds funny, but its quite possibly a better name to choose and probably less intrusive..--Cometstyles 10:04, 2 June 2008 (UTC)[reply]
how about wiki Knights ? joking aside i think global janitor is good enough --Mardetanha talk 10:07, 2 June 2008 (UTC)[reply]
I don't think we should use obscure names. The proposal is for global sysops, so we should call them 'global sysops' or 'backup sysops'— not knights, reserves, monitors, nomads, janitors, or apples. The name should reflect their purpose, be as intuitive as possible, and be easily translated. It's not a noble military title, they're not intended to back up Wikimedia's standing army, they're not tasked with passively watching, they're not simply travelers, we're not looking to hire cleaners and repairmen, and they're not fruit of the Malus sylvestris tree. —{admin} Pathoschild 21:20:47, 02 June 2008 (UTC)
To be precise, they are not sysops. While they have similar permissions to sysops, they don't have similar rights. For example, they wouldn't be allowed to block anyone for, for example, trolling. If such [controversial] action is needed on a wiki without admins, it's a stewards' job, not their. The most precise description is the current name: they are anti-vandal fighters (yes, they will fight against vandals and this will be their first and last job; while it is not good to have an aggressive terminology, this will be the fact). However, we need some other name... --Millosh 22:34, 2 June 2008 (UTC)[reply]
I don't think that the name is a big deal. So, I think that we should vote soon and that the name with the most of votes in the first round should win (everyone should be able to vote for as many options as they like). Whatever proposal wins, I would be happy :) --Millosh 22:34, 2 June 2008 (UTC)[reply]
this name is misleading. Sysop is a term generally applied to administrators. Anti-vandal fighter is not a sysop (administrator) but has a subset of sysop powers. To call them global sysops would be deceptive and a gross missuse of the term. Dbiel 02:25, 4 June 2008 (UTC)[reply]
Sysop does not equal administrator. example ([move=sysop]) vadlal fighter would be able to move these pages as can serveral other classes of users Dbiel 20:57, 6 June 2008 (UTC)[reply]
oppose, they should not act as sysop, how to call stewards for their technical access then --birdy geimfyglið (:> )=| 13:17, 4 June 2008 (UTC)[reply]
vandal fighters** Cbrown1023 talk 23:40, 2 June 2008 (UTC)[reply]

As I think that there are not a lot more to be said about names, I think that it is a time to make a poll. I'll announce it foundation-l, too. If I missed some proposal, pleas add it in alphabetical order. --Millosh 18:32, 3 June 2008 (UTC)[reply]

Janitor Oppose Oppose People may get confused by the "Janitor" right on Wikiversity--penubag 15:53, 5 June 2008 (UTC)[reply]

Note: Wikiversity uses the term "custodian", which is sometimes synonymous to "janitor".Hillgentleman 01:27, 7 June 2008 (UTC)[reply]
the preceding two entries (starting with Janitor Oppose Oppose were move out of the poll list which is not to include any negative votes Dbiel 20:05, 9 June 2008 (UTC) [reply]

Poll

Poll is closed. The name of the proposed role is global sysop.

Please, vote for one or more proposals. Negative votes won't be counted. Voting begins at the time of announcing this page (around 18:30 UTC at June 3rd) and lasts 7 days, up to June 10th, 2008 at 18:30 UTC.

Proposal with the most of votes will win in the first round.

Anti-vandal fighter

  1. Up to discussion seems to be ok by me--Mardetanha talk 18:35, 3 June 2008 (UTC)[reply]
  2.  – Mike.lifeguard | @en.wb
  3. I think this one should be used one step lower, e.g. people who do a lot of cross-wiki cleanup, but do not need access to e.g. deletion (but with e.g. rollback etc.). --Dirk Beetstra T C (en: U, T) 09:11, 4 June 2008 (UTC)[reply]
  4. --penubag 15:49, 5 June 2008 (UTC)[reply]

Apple

  1. ...

Backup sysop

  1.  – Mike.lifeguard | @en.wb
  2. WjBscribe 08:10, 4 June 2008 (UTC)[reply]
  3. --Dirk Beetstra T C (en: U, T) 11:32, 10 June 2008 (UTC)[reply]

Cross-wiki maintenance

  1.  – Mike.lifeguard | @en.wb
  2. WjBscribe 08:10, 4 June 2008 (UTC)[reply]
  3. Seems OK .. though calling deletion of rubbish 'maintenance' does not give a true reflection. --Dirk Beetstra T C (en: U, T) 09:12, 4 June 2008 (UTC)[reply]
  4. --birdy geimfyglið (:> )=| 14:57, 4 June 2008 (UTC)[reply]

Global assistant

  1. ...

Global dirt sweeper

  1. ...

Global helper

  1. ...

Global janitor

  1. And this sounds good for me also --Mardetanha talk 18:43, 3 June 2008 (UTC)[reply]
  2. Either this or simply "janitor" as below. Adambro 19:14, 3 June 2008 (UTC)[reply]
  3. WjBscribe 08:10, 4 June 2008 (UTC)[reply]

Global maintenance

  1.  – Mike.lifeguard | @en.wb
  2. WjBscribe 08:10, 4 June 2008 (UTC)[reply]
  3. Dbiel 23:16, 4 June 2008 (UTC)[reply]

Global monitor

  1. --Yaroslav Blanter 13:23, 4 June 2008 (UTC)[reply]

Global rollback

  1. Different function. Not real sysops. --Dirk Beetstra T C (en: U, T) 09:13, 4 June 2008 (UTC)[reply]

Global sysop

  1. The proposal is for users with global sysop access, so they should be called "global sysop" or "backup sysop". The name should reflect their technical access and not their community role (while their role can change, their name is permanent). This is unambiguous and draws a parallel to the local group (so users don't need to read technical pages to know what sort of access they have), and removes the artificial limits imposed by some of the other proposals (which don't reflect the actual proposal or allow future change). —{admin} Pathoschild 20:37:01, 03 June 2008 (UTC)
  2. Second choice. Nakon 20:38, 3 June 2008 (UTC)[reply]
  3. --Nemo 21:15, 3 June 2008 (UTC)[reply]
  4. I will agree with anything but Janitor as this will cause more tension between ppl who believe wikia is wikimedia..--Cometstyles 22:53, 3 June 2008 (UTC)[reply]
  5. Per Pathoschild. — H92 (t · c · no) 00:52, 4 June 2008 (UTC)[reply]
  6.  – Mike.lifeguard | @en.wb
  7. Per Pathos. --Rory096 01:24, 4 June 2008 (UTC)[reply]
  8. Janitor is too ambiguous. giggy (:O) 01:38, 4 June 2008 (UTC)[reply]
  9. Janitor is too much like wikia Legoktm 03:38, 4 June 2008 (UTC)[reply]
  10. Mr.Z-man 04:00, 4 June 2008 (UTC)[reply]
  11. per Pathoschild. Werdna 05:32, 4 June 2008 (UTC)[reply]
  12. --.snoopy. 05:36, 4 June 2008 (UTC)[reply]
  13. WjBscribe 08:10, 4 June 2008 (UTC)[reply]
  14. --Nick1915 - all you want 09:00, 4 June 2008 (UTC)[reply]
  15. Sounds good to me. (or 'Global Administrator'?)--Dirk Beetstra T C (en: U, T) 09:09, 4 June 2008 (UTC)[reply]
  16. Sounds informative — VasilievV 2 09:15, 4 June 2008 (UTC)[reply]
  17. Per VasilievVV --Erwin(85) 14:55, 4 June 2008 (UTC)[reply]
  18. per vasilievVV. --Ahonc 20:45, 4 June 2008 (UTC)[reply]
  19. DerHexer (Talk) 20:58, 4 June 2008 (UTC)[reply]
  20. Both informative and pleasant. Global administrator too. Cenarium (talk) 02:43, 5 June 2008 (UTC)[reply]
  21. Is good. --Anonymous DissidentTalk 06:11, 5 June 2008 (UTC)[reply]
  22. --MF-W 14:45, 5 June 2008 (UTC)[reply]
  23. --penubag 15:49, 5 June 2008 (UTC)[reply]
  24. Support Support sysop does not = administrator - example ([move=sysop]) vadlal fighter would be above to move these pages Dbiel 20:52, 6 June 2008 (UTC)[reply]
  25. AVF is a misnomer, since they would be involved in far more than vandalism fighting – this name more accurately reflects this. haz (talk) 10:27, 9 June 2008 (UTC)[reply]
  26. Second choice. - Mailer Diablo 14:25, 9 June 2008 (UTC)[reply]
  27. SPQRobin 15:42, 9 June 2008 (UTC)[reply]
  28. Platonides 21:16, 10 June 2008 (UTC)[reply]

Common Sysop

Cross-wiki Sysop

  1. --Purodha Blissenbach 23:29, 4 June 2008 (UTC)[reply]

Cross-wiki Admin

  1. --Purodha Blissenbach 23:30, 4 June 2008 (UTC)[reply]

Globetrotter

  1. --Millosh 09:15, 4 June 2008 (UTC)[reply]

Helper

  1. pro, --birdy geimfyglið (:> )=| 00:46, 4 June 2008 (UTC)[reply]
  2. WjBscribe 08:10, 4 June 2008 (UTC)[reply]

Janitor

  1. Support Support no need for redundancy of adding "Global", this role isn't available locally so Janitor alone would suffice.--Alnokta 19:08, 3 June 2008 (UTC)[reply]
  2. Support Support I like this term, I think it describes the role pretty well. Not too fussed whether we prefix it with "Global", this might be appropriate. Adambro 19:14, 3 June 2008 (UTC)[reply]
  3. Support Support The name corresponds well with the name steward and it seems to describe the role well. I think we should try to avoid the words "vandalism" and "fighter" (and also "spam") here, those words can be counter productive. --Jorunn 20:12, 3 June 2008 (UTC)[reply]
  4. Majorly talk 20:17, 3 June 2008 (UTC)[reply]
  5. Nice one. Also used at Wikia — VasilievV 2 20:18, 3 June 2008 (UTC)[reply]
  6. Seems OK. Nakon 20:29, 3 June 2008 (UTC)[reply]
  7. Thunderhead 20:41, 3 June 2008 (UTC)[reply]
  8. Support Support Cbrown1023 talk 21:02, 3 June 2008 (UTC)[reply]
  9. This name is already in use on other sites and working well. Angela 22:09, 3 June 2008 (UTC)[reply]
  10. Support Support --Kiensvay 00:35, 4 June 2008 (UTC)[reply]
  11.  – Mike.lifeguard | @en.wb
  12. Support Support Jkasd 04:05, 4 June 2008 (UTC)[reply]
  13. WjBscribe 08:10, 4 June 2008 (UTC)[reply]
  14. Support Support "Global Sysop" implies too much power.--Werdan7T @ 16:05, 4 June 2008 (UTC)[reply]
  15. While this it is unfortunate this is the same name than Wikia, this is the one that describes the position the best (global sysop implies that they have more power than local sysops). -- lucasbfr talk 12:24, 5 June 2008 (UTC)[reply]
  16. Very fortunate that this is the same name as on Wikia. It is a very big plus to keep consistency around the wiki-world. It's pointless to create another (new) name for the exact same position, and create ambiguity. Jon Harald Søby 01:08, 7 June 2008 (UTC)[reply]
  17. Mailer Diablo 14:25, 9 June 2008 (UTC)[reply]

Knight

  1. ...

Keeper

  1. ...

Monitor

  1. --Purodha Blissenbach 23:31, 4 June 2008 (UTC)[reply]

Nomad

  1.  – Mike.lifeguard | @en.wb
  2. ..--Cometstyles 04:13, 4 June 2008 (UTC)[reply]
  3. --.snoopy. 05:37, 4 June 2008 (UTC)[reply]
  4. --Millosh 09:14, 4 June 2008 (UTC)[reply]
  5. --Cradel 21:17, 5 June 2008 (UTC)[reply]

Reserve

  1. my second choice, Probably a less intrusive name :) ..--Cometstyles 23:00, 3 June 2008 (UTC)[reply]
  2.  – Mike.lifeguard | @en.wb

Traveler

  1. ...

Vandal fighter

  1.  – Mike.lifeguard | @en.wb
  2. Zouavman Le Zouave

Vandalism cleaner

  1.  – Mike.lifeguard | @en.wb

Cross-Wiki Vandal Fighter

  1.  – Mike.lifeguard | @en.wb

Watcher

  1. ...

SWMT

  1. --Purodha Blissenbach 23:30, 4 June 2008 (UTC)[reply]
  2.  – Mike.lifeguard | @en.wb
  3. Daniel (talk) 00:13, 7 June 2008 (UTC)[reply]

No no no

The following discussion is closed: Avoid battleground mentality

WikiMedia is about providing useful free of cost copy-left educational information to all mankind. Our openness to vandals is a problem, not a reason to transform us into a battleground. No no no. Better to stop allowing random edits than to go this route. (WAS 4.250) 4.251.121.134 08:17, 31 May 2008 (UTC)[reply]

After some point it is useless to find a constructive way for dealing with vandals. Battleground exists for a long time and stewards are fighting there. BTW, how to "stop allowing random edits"? --Millosh 08:29, 31 May 2008 (UTC)[reply]
This is a needful task that occupies a goodly fraction of steward and small wiki task force volunteer time. Providing better tools to address the issue (I avoid "fight" or "combat" because I take your point) seems a good idea to me. ++Lar: t/c 18:47, 31 May 2008 (UTC)[reply]
The following discussion is closed: Tweaked requirements

This section seems rather unclear to me. That sounds like exactly the same criteria as Meta:Requests_for_adminship#Regular_adminship. "The candidate will be named sysop here only if..." -- but we are not talking about sysop here, we are talking about this global antivandal role. I'm not sure whether the requirements should be changed, but perhaps clarifying would be a good idea.  – Mike.lifeguard | @en.wb 11:14, 31 May 2008 (UTC)[reply]

Yes, I copied conditions from that page :) I didn't change that a lot, so, yes, if it is needed, this part should be adapted. I am sure that you understand the sense (anti-vandal fighters should be elected and there are some requirements for them), so, please, rephrase (or whatever) that as you think that it is the best. --Millosh 11:21, 31 May 2008 (UTC)[reply]
OK, I updated the language. I think that the requirements need tweaking though. I don't think the first point ("Having been a participant for at least 2 months on at least two projects other than Meta" is what we really want - we want people to be active on many projects and fighting spam/vandalism on them. If they have made small numbers of edits but spread out over 700 wikis that is still fine. I'd change it to something like "Having been a noted cross-wiki anti-spam/vandal fighter for a period of at least 2 months." That of course raises the question of the meaning of "cross-wiki" (how many?) and of "anti-spam/vandal fighter" (how active?). These are topics for discussion.  – Mike.lifeguard | @en.wb 12:11, 31 May 2008 (UTC)[reply]
I think there is a sense in which if you are active cross wiki it will be on quite a few projects. I think it is unlikely that someone will show up asking for the rights unless they already have cross wiki interests (we are quite a rare breed!). However I guess it would be an idea to make it quite a high bar to prevent the "ooo nice shiny new tool & I want it" approach from anyone. --Herby talk thyme 12:17, 31 May 2008 (UTC)[reply]

Or, maybe, being a member of SWMT for at least two months and actively participate in its work or so? This means that we will have the exact date of becoming a member of SWMT (the date when they added themselves at the SWMT members list) and we would be able to see user's actions from that time. Also, it should be noted that a candidate should give relevant links (contributions on various projects) for their involvement into SWMT housekeeping. --Millosh 12:20, 31 May 2008 (UTC)[reply]

Yes, but the problem is that SWMT is mostly-self managed. You put yourself on the list and nobody checks if you're actually doing it. On the other hand, those who are active do know each other, so when you ask for the bit, we will say "but you don't do anything".  – Mike.lifeguard | @en.wb 12:47, 31 May 2008 (UTC)[reply]
Maybe to give a better introduction at the SWMT page? Something like: If you want to start immediately, go to #wikimedia-stewards channel and ask how can you help. Also, some more documentation at SWMT page would be good, too. Etc. etc. Also, note that this role is discussed at the stewards' list exactly because of the members of SWMT (there are some users who are not stewards and who are willing to help, so, we are able now to give them the right tools). Also, as a lot of SWMT members are stewards, too -- stewards would be able to monitor actions of other members of SWMT. --Millosh 12:53, 31 May 2008 (UTC)[reply]
Agree with Mike and Herby above, most ppl are driven by "ooh shiny new tools. me can have it?" rule, and believe me most don't even contribute, its just a trophy, Global rollback was a start, and you have managed to take it 4 steps forward, great, but this will be problematic in years to come, I have a good name for this, "Stewards'..better let the professionals handle it for now :P ..--Cometstyles 13:09, 31 May 2008 (UTC)[reply]
Maybe this looks like without a context, but, again, we were discussing at stewards' list about what is needed for housekeeping for a week or so. Giving to a SWMT member just a right to revert change is similarly to asking someone to clean some dirt with towel, but without water and without detergent. Yes, it is possible, but it is a harder way for doing so. Also, it is not so hard to become a "professional". I know a number of contributors who have a lot of knowledge and who are helping a lot, but who didn't pass voting for stewardship. --Millosh 13:57, 31 May 2008 (UTC)[reply]
I'm with Comets here. It is sad that "rights" are seen as something get rather than something that is for using. It would be a pity if this merely attracted people who thought it would be another badge for them - there are enough of those around as it is. I am not sure I would warrant such rights for example even though I do deal with excessive linkage anywhere I have the time to do so. People should be really active to warrant this not simply to want it --Herby talk thyme 11:23, 1 June 2008 (UTC)[reply]
It may be solved partially by confirmation of the rights; with participation of global maintainers in voting (stewards, meta bureaucrats and anti-vandal fighters themselves). (BTW, I would like to see the same for stewards and Meta bureaucrats: not to be confirmed only by themselves.) --Millosh 11:57, 1 June 2008 (UTC)[reply]
"Confirmation" I would certainly be in favour of. I know I keep saying it but tools are for using, if people don't use them any more the are removed & handed back if needed again --Herby talk thyme 12:53, 1 June 2008 (UTC)[reply]
BTW, it is really good to build a policy together :) I missed this part, even it exists for stewards, too. --Millosh 18:48, 1 June 2008 (UTC)[reply]

I added "Policy for removing rights" from Meta:Administrators. --Millosh 12:04, 1 June 2008 (UTC)[reply]

Mark as patrolled

Mark as Patrolled is not necessarily used for quality control, at least on nl.wp it's used for first vetting of contributions for vandalism, makes-sense, ... Be carefull when assuming global permissions :) Henna 11:31, 31 May 2008 (UTC)[reply]

Considering that these users may well be working on projects in languages they do not understand, I'd be wary about allowing patrolling. OTOH, much/most vandalism is easy to spot regardless of the language. Still, I think patrolling is not required for this set of tasks.  – Mike.lifeguard | @en.wb 12:01, 31 May 2008 (UTC)[reply]
it's a fairly usefull way to revert something and mark it as patrolled so that it doesn't show up in recent changes anymore :) Henna 12:39, 1 June 2008 (UTC)[reply]

unnecessary

there's no need for two different names for the same job. to me it seems, we're trying to solve a social problem in a technical way: technically, cleanup-admins, anti-vandal-fighters or whatever you call them need to have nearly the same permissions as regular admins: deleting pages, reverting edits, maybe blocking a vandal. the important difference is a social one: a cleanup-admin should deal with vandals only and not meddle in the affairs of the community. this cannot be controlled by software. -- 21:41, 31 May 2008 (UTC)[reply]

This is different from regular admins. Regular admins have rights locally, on one wiki. This proposed "Anti-vandal fighter" would have admin rights glboally, on every Wikimedia wiki. That is why there are two different names. Cbrown1023 talk 22:25, 31 May 2008 (UTC)[reply]

Exempting projects

Perhaps it wouldn’t be such a bad idea to integrate a section permitting individual communities to opt-out from being subject to actions by this user class. This would also be handy in avoiding any misunderstandings as to whether an individual project is in sufficient `stock` with admins or not - each community can decide if they feel up to the task of keeping their project tidy on their own, and would help avoid global sysops earning some sort of a vigilante rep (which I’m not saying they would, but it doesn’t hurt to be on the safe side of things when it comes to such a potentially intrusive tool). Don’t get me wrong, I think it’s a great idea to supply volunteers of the SWMT with an efficient tool for the work they do, but it would be nice to assure that the integrity of individual projects remains intact not just through expecting AVfs to be responsible. --Кале 20:15, 1 June 2008 (UTC)[reply]

Permissions are global, so it is not possible to make opt-out. If an anti-vandal fighter makes an action on some project with active admins, they will lose permissions immediately. I think this is a very good protection for abusing the rights. --Millosh 21:09, 1 June 2008 (UTC)[reply]
The opt-out doesn't necessarily have to yield a change in actual permissions (although, this would be the ideal scenario), but rather an implicit one. My issue (perhaps I'm alone on this one) is with the ambiguity of the `project with active admins` definition. I am thus suggesting to allow the communities themselves to appraise their situation and decide whether they need the help or not, and then notify Meta of such a decision. Should they decide they're fine on their own, their project would become a no-go for global admins. Otherwise, unsolicited actions could result in less than desirable outcomes, staining what is generally a very good concept for keeping small projects junk-free, allowing them the time needed to build an able community. --Кале 21:37, 1 June 2008 (UTC)[reply]
Kale, this is about projects with at most 5-10 admins or so. Such communities are not able to take care about themselves 24 hours. Sr.wp is definitely not one of such projects and SWMT doesn't care about sr.wp; which means that if help is needed for sr.wp, someone needs to explicitly ask (now) stewards to help to the project (like, for example, we had similar situation with it.wp, where admin who was awake was not sure how to make a range block). Also, when you need help, you will not take care if someone is steward or anti-vandal fighter, you will come to ask someone to help, no matter who is that person. And if you really don't want any help, just don't go to #wikimedia-stewards channel. So, for a small project it is very stupid not to be there (if they have a relevant community at all), and for a bigger (like sr.wp is), just don't ask for help and by default nobody of anti-vandal fighters will go there. --Millosh 22:21, 1 June 2008 (UTC)[reply]
My comments relate in no way to sr.wp (otherwise I'd be hunting you down at gtalk, not spellchecking my entries at Meta :)), but rather me, expressing concern over what I believe is a far-reaching policy in need of refinement to assure no community feels threatened by the implications it has, which in turn would secure its successful implementation. If your mentioning of sr.wp is fueled by inability to otherwise explain my spurt of commenting here, I stumbled upon the local RC and this caught my eye. I am sorry if I am in violation of some unwritten rules on degrees and dynamics of Meta involvement :) So, if the issue is not whether communities should be allowed to opt-out or not, but merely how, I suggest notice of such an option should be an integral part of the policy. --Кале 22:45, 1 June 2008 (UTC)[reply]
Things are very simple: If a community is able to take care about their project, nobody will go there. If a community is not able to take care about itself, then AVfs and stewards will take care about that project. If a community flipped out (like ru.wb was because of one bureaucrat), AVfs definitely will not make any controversial action; it is stewards' job. This group is designed as a group which should support stewards' actions. And, now, stewards are doing exactly the same job with help of other volunteers who are not able to react independently. So, nothing special would be changed. --Millosh 23:06, 1 June 2008 (UTC)[reply]
If you are asking if a community which is not able to take care about itself would be able to opt-out: the answer is no; as they are not able to opt-out from stewards' actions. Keeping spam, nonsenses and vandalisms at Wikimedia projects is not acceptable; as well as obstruction of maintaining work by demanding that only stewards may do maintaining on a specific project is not acceptable. --Millosh 23:06, 1 June 2008 (UTC)[reply]
And, again, for any project which community is able to take care about it -- the rule is simple: AVfs wouldn't be able to use even a "rollback" permission without explicitly asking the community. --Millosh 23:06, 1 June 2008 (UTC)[reply]
Also, keep in mind that the goal is to have mature communities which don't need stewards' or AVfs' help. So, everybody will be happy to see the community mature enough to go out of the SWMT's patronizing. --Millosh 23:06, 1 June 2008 (UTC)[reply]

Righty, since opting-out doesn't seem to be an option, shouldn't the policy then at least define:

  • a loose number of sysops deemed sufficient;
  • that global sysop tools must not be used in an arbitrary fashion, not in content disputes, and not for overriding local communities' or admins' decisions;
  • that blocking actions, with the exemption of cross-wiki vandalism, must conform to local blocking policy, if there is any;
  • that these tools must not be used on one's home project, unless one has locally been granted sysop status.

I don't think it is appropriate to leave such things to interpretation of vague sentences or perusing of talk pages. The goal is definitely not creating a bureaucratic policy, but one that will unambiguously postulate the essence of this user group through listing what it is, and what it is not intended for. Cheers --Кале 12:22, 2 June 2008 (UTC)[reply]

But, your additions are mostly bureaucratic: --Millosh 14:30, 2 June 2008 (UTC)[reply]
This is your opinion, which I respect, but disagree with. The fact is that while you may have a clear idea in your head of what the policy should be, that has not been articulated into a clear, non vague set of rules and rationales. IMHO, anyway --Кале 15:44, 2 June 2008 (UTC)[reply]
  • Number of admins is a very problematic issue. There are well organized projects with 10 admins, while there are not so good organized projects with 15 admins. So, something like "between 0 and 30 active administrators" may pass, but it is a really bureaucratic addition. The other thing is that it is definitely up to a team which deals with vandalisms to realize when some project becomes a mature one. --Millosh 14:30, 2 June 2008 (UTC)[reply]
And this can remain the team's prerogative, bearing in mind that no one finds particular pleasure in monitoring more projects than absolutely needed and that the merit of a community's capability to deal with day-to-day vandalism is certainly measured by more than just their admin count, but a ball-park figure in this aspect or, even better, some form of guidelines (this would only help the team optimize its scope of activity thus procuring the greatest degree of efficiency) on the extent of the team's involvement would do miracles in showing this policy as more than just a heap of text serving to introduce a new global user class with little regard to the implications it bears. --Кале 15:44, 2 June 2008 (UTC)[reply]
  • "... not to work on home projects..." -- Preferably not, in the sense of the rules for stewards (I'll add it). --Millosh 14:30, 2 June 2008 (UTC)[reply]
    It is good to known we can agree on something in the very least. --Кале 15:44, 2 June 2008 (UTC)[reply]
    Added. --Millosh 15:19, 3 June 2008 (UTC)[reply]
  • Other of your asks are related to the mature projects: --Millosh 14:30, 2 June 2008 (UTC)[reply]
    • A couple of times inside of the text is clearly stated that this group deals just with vandalisms, not with any other kind of issues. If any administrator is present at some project in some time frame, AVf mustn't act if not called explicitely. It may be added somewhere that they shouldn't use their permissions in other ways than to fight against vandalism, but I don't see that as necessary because it is very clearly stated that they will get permissions only for anti-vandalism purpose. Also, all of them need to pass confirmation process every year. --Millosh 14:30, 2 June 2008 (UTC)[reply]
    Could you please quote one of such instances where this policy clearly states ”that this group deals just with vandalisms”. I believe not. Hence my suggestion to introduce all that I've stated in the second bullet above. You have chosen not to add any of the items there presented, which frankly I don't understand, seeing as how not even you are claiming they would be allowed to participate in content disputes or take arbitrary actions. Or are you? --Кале 15:44, 2 June 2008 (UTC)[reply]
    Something like "Anti-vandal fighters" or, more explitcly: --Millosh 16:20, 2 June 2008 (UTC)[reply]

    While anti-vandal fighters should have similar permissions to administrators, their role is not to maintain wikis at the regular basis, like wiki administrators are doing. Their role is to fight against vandals and to allow to the community at a particular project to take care about their own wiki. Because of that, while anti-vandal fighters need and have permissions like deleting pages and blocking users, they don't need and don't have permissions like "Mark others' edits as patrolled", which is generally a feature which every administrator should have.

--Millosh 16:20, 2 June 2008 (UTC)[reply]

  • I really don't see what is not clear there. --Millosh 16:20, 2 June 2008 (UTC)[reply]
    • "... must conform to local blocking policy..." -- no way! Because of:
      You are completely correct when you say it is utterly unreasonable to expect potential AVfs to be acquainted with all blocking policies in existence. No one expects them to, if they deal with cross-wiki vandals or clear issues of vandalism that bear roughly the same sanctions on any project. This is merely another point which would assure no arbitrary actions are taken. --Кале 15:44, 2 June 2008 (UTC)[reply]
      Again, they are defined as volunteers which are dealing with vandals; as well as it is explicitly said that they are not regular administrators. --Millosh 16:29, 2 June 2008 (UTC)[reply]
      • It is totally unreasonable to demand from one person to be introduced in ~500 policies (do they exist? if exist, what may they do? what number of languages should know one AVf? (policies are not transcribed into English) etc.). If communities around smaller projects are willing to have visible policies, they should write it in English and unify them. --Millosh 14:30, 2 June 2008 (UTC)[reply]
      • Again, projects with developed policies possibly have enough of admins. No one of those AVfs will go to en.wp or de.wp or even to much smaller projects like sr.wp is -- if not explicitly asked. (And in the cases like en.wp and de.wp are, they will not go there even explicitly asked.) --Millosh 14:30, 2 June 2008 (UTC)[reply]
      • And, the most important, which this your addition makes as a clear example of unnecessary bureaucratization, is that any local admin may revert that block (made during, let's say, night hours) into one according to the local rules. --Millosh 14:30, 2 June 2008 (UTC)[reply]
        To care so seemingly little for the potentially harmful consequences unsolicited actions might provoke doesn't appear to be very prudent. --Кале 15:44, 2 June 2008 (UTC)[reply]

Finally, since it is becoming painfully obvious this policy is not destined to change, not even in the miniscule of ways, I'm afraid further attempts to do so would be less than fruitful. The best of luck with it --Кале 15:44, 2 June 2008 (UTC)[reply]

Of course it may be changed during the discussion, as it was already changed signficantly. However, it can't address paranoia. Also, this is the policy, not the full set of documents around the policy. Nothing was created at once. Even the documents around the Board are far from fairly complete. --Millosh 16:10, 2 June 2008 (UTC)[reply]

Alternative idea

Instead of the above, which contemplates projects wishing to actively refuse the services of people with this right, I wonder if there are projects which (though they technically have enough sysops) would be willing for such people to do routine cleanup/anti-vandal work. It might be worth considering having a list of projects that are willing to trust users with this global right regardless of their local supply of sysops - though this may lead to extra confusion for those with this right as to where they are allowed to use it. WjBscribe 09:57, 2 June 2008 (UTC)[reply]

Hana mentioned that in relation to the "mark as patrolled" permission, but this means adding a new permission to this group of users (which is not big deal, so it may be added, too). I agree with that possibility. And I think that it is up to a particular community to decide. We should make a separate page (after the policy adoption) with a list of projects which decided so. I expect that we will have more of AVfs than stewards and that they will be able to have in depth communication with all communities. So, this should be one of the first tasks for this group to do. --Millosh 10:36, 2 June 2008 (UTC)[reply]

Requirements

I did not quite get from the text of the proposal: does a candidate need to have more than 100 edits in each of at least two projects, or 100 combined? I mean, does the candidate have to be seriously involved in both? --Yaroslav Blanter 07:52, 2 June 2008 (UTC)[reply]

Combined, of course :) --Millosh 09:37, 2 June 2008 (UTC)[reply]
I would hope that in practice substantially more experience would been seen as necessary --Herby talk thyme 14:32, 2 June 2008 (UTC)[reply]
Yes, I was thinking about that, too. Maybe, something like 6 months, 1000 edits at all and 100 edits during the last month? --Millosh 14:33, 2 June 2008 (UTC)[reply]
Maybe something about "over x wikis" too? There are times I would love these rights (have been playing with an Xrumer bot earlier) however I'm not sure I warrant them & I do wonder about "local" reaction to such people unless it is quite tightly controlled.
In the case of this one I probably would have used the rights but within about an hour all but one of the pages were deleted. However only one block has been placed as a result of my "bot on OP" message so others may see things differently I guess. --Herby talk thyme 14:42, 2 June 2008 (UTC)[reply]
Hm. This is the first global role thanks to global permissions (of course, stewards are the first at all) and this shouldn't be the last. During the next couple of months we may build some other roles, like "bot owner" may be. I asked developers for roles like "user is a bot" (different than a "bot flag") and "bot owner", as well as "delete if I am the only contributor". In coordination of those permissions, bot owner would be able to delete pages created by their bot. So, I hope that it would solve your need. Getting AVf permissions for working with a bot is not a good path. --Millosh 14:52, 2 June 2008 (UTC)[reply]
Nope I was tracking a vandal bot not me as the bot - apologies for the lack of clarity. On teh speedy del requests I put the narrative "cross wiki spam bot on OP" hope folk would block the IP as well as delete the page - mostly I was wrong :( --Herby talk thyme 14:54, 2 June 2008 (UTC)[reply]
Oh, OK :) "Over x wikis"... Let's try to analyze: (1) We need an active contributor. (2) We need an active Meta participant. (3) We need a person who is working on at least 2 content projects. So, I'll try to formulate the rule: "... at least 1000 edits at all of which at least 100 edits at all on Meta and 100 edits at all on at least two content projects (per project); as well as at least 50 edits during the last month (if you are applying at August 12th, this means that the last month is July, not the period July 12th - August 11th) at Meta and the same amount of edits at at least 2 content projects (per project)." --Millosh 15:06, 2 June 2008 (UTC)[reply]

I added requirements. --Millosh 14:36, 3 June 2008 (UTC)[reply]

Knowing if one has global or local rights

Is there anything to indicate to the person with this permission whether their rights on a given project are global or local? I presume they just get the delete and protect tabs as if they were local sysops... Those with rights on multiple projects who are given this permission may need to take care to remember where they actually have local rights and to double check what wiki they are on before they delete/protect/block if the consequences of use on the wrong project will be fairly automatic removal... WjBscribe 09:53, 2 June 2008 (UTC)[reply]

Hehe... I woke up one day, went to the English Wikipedia Main Page to see what's new and realized that I have "edit" and "delete" options there. I really don't think that it would be a big deal if someone misused some of those buttons by mistake. Maybe different skins may help: Only projects on which admin rights may be used should have some other-than-default skin or so. I don't know for any other possibility. --Millosh 10:41, 2 June 2008 (UTC)[reply]
Sure. They should have a JavaScript hack — VasilievV 2 14:45, 2 June 2008 (UTC)[reply]
To answer WjBscribe's question : yes, there is a way; for example, stewards currently have "global sysop" permissions; they do have sysop tools e.g. on en.wikipedia, but they don't appear as local sysops in en:special:listusers/sysop. guillom 13:20, 3 June 2008 (UTC)[reply]

Rights removal

It says the rights would be removed immediately..stewards don't remove rights immediately. may be admins should be able to remove the rights of them?--Alnokta 13:39, 3 June 2008 (UTC)[reply]

Stewards do remove them immediately in case of vandalism. And in case of abuse for AVfs — VasilievV 2 13:42, 3 June 2008 (UTC)[reply]


Policy on who can request this rights

Number of edits

Looking through the request policy, it seems that its a bit to easy to get this right..such as

  • Sum of edits
  "* at least 1000 edits total (all over the Wikimedia projects)
   * at least 100 edits total at Meta
   * at least 100 edits total on at least two projects (per project)"

that in my point is really really low to request a right which allows you to delete/undelete pages and also Block users. Sometimes private information are deleted instead of and it can be seen by people with this global right. I agree with the 100 edits to Meta but that 100 edits should mainly include mainspace and not userspace but the last one ("at least 100 edits total on at least two projects (per project)") really doesn't make any sense to me, so my proposed change to that will be:

  • Sum of edits
  "* at least 5000 edits total (all over the Wikimedia projects)
   * at least 100 edits total at Meta (which doesn't include userspace)
   

and regarding other changes:

  1. Having unified account.
  2. Being a Administrator, Bureaucrat or CheckUser on a content project"
  1. Having a fully merged unified account.
  2. Being an active Administrator, Bureaucrat or CheckUser on a content project"

These are just some of the minor changes I will propose for now :) ..--Cometstyles 00:52, 4 June 2008 (UTC)[reply]

OK. I think that 5000 edits are not too much for experienced Wikimedians. However, the sense of 2x100 edits (note, not 100 edits at one project) at content projects is to confirm that user is fairly active in cross-wiki matters. --Millosh 01:02, 4 June 2008 (UTC)[reply]
Why not 5000 edits at their home wiki(s), 100 at Meta, another 100 across the rest of the projects (ie projects where they do only anti-vandal/spam) and have some additional access somewhere? Here, define a home wiki as anywhere they contribute content and/or have additional access - by this definition I (for example) would have 4 home wikis which don't count towards the "100 across the rest of the projects" count. This ensures that they are active somewhere (their home wikis), active at meta, and active cross-wiki (that "everywhere else" count). Perhaps increase the home wiki edit count and the other-projects edit count. On the other hand, we will be doing RF?s for them, so it's not the end of the world if the number isn't perfect - people will (I hope) be judged more on their merits than their edit counts (though particularly for this edit counts will be an important metric).  – Mike.lifeguard | @en.wb 01:21, 4 June 2008 (UTC)[reply]
Users with this right need to be supremely trustworthy, given they could go and block an established user on enwp, something which always results in drama and lost contributors. Unless this right is in some way technically limited to exclude certain wikis, I would support a required edit count in the thousands, experience on meta, as well as require that no objections be raised in X days if they aren't a sysop on a major project (eg any of the >300,000 article ones [More than 300,000 articles: Deutsch · English · Español · Français · Italiano · Nederlands · 日本語 · Polski · Português], or Meta). They should also be made known that they sacrifice their local sysop access if they abuse their global access, as both will be removed with prejudice. Daniel (talk) 01:23, 4 June 2008 (UTC)[reply]
It wouldn't be fair to demand from a user be an admin of a specific project. But, I support the most of other requirements. --Millosh 09:26, 4 June 2008 (UTC)[reply]
Well if they don't have adminship anywhere, then they don't really know how to use the tools, do they?..nah it should be a strict requirement to avoid it being abused even by mistake..--Cometstyles 09:35, 4 June 2008 (UTC)[reply]
You didn't understand me... I said that it wouldn't be fair to demand from a contributor to be an admin of German, English..., or Portuguese Wikipedias. --Millosh 09:52, 4 June 2008 (UTC)[reply]
sorry but 1000 is really tooooo low, anyone can accumulate that much edits with the right tool, Huggles comes to mind..and as Daniel mentioned, they need to be supremely trustworthy and someone with 1000 edits is either inactive, new or a sock account..neither worthy enough to demand the tool, Millosh there are people with over 40,000 edits on enwiki who are not admins and but whose accounts are very much less than 9 months old ;)...btw, this is a global right and only those involved cross-wiki really have the need for it and basically more than 90% of wikimedians don't even know how many wikis there are and mainly emphasise on their homewiki, voting is probably not the right way to go either with all the canvassing that might happen but its a start, I'll agree with Mike here, 5000 on home wiki is a better idea and atleast 1000 edits wikimedia wide sounds way reasonable compared to the original one...--Cometstyles 03:58, 4 June 2008 (UTC)[reply]
5000 + 1000 sounds a good start to me. I'd throw in a preference (did I miss it?) for some reasonable number here on Meta - it would indicate a cross wiki interest? Cheers --Herby talk thyme 06:41, 4 June 2008 (UTC)[reply]
1000 at one content wiki and at least 100 at another (and 5000 total, 100 at Meta)? It would be good to have people who are active at more than one wiki. --Millosh 09:30, 4 June 2008 (UTC)[reply]
I have been looking at that '100 edits on two content wikis' - I think quite some of us here who are fairly active cross wiki do have only a main account on one content wiki and here on meta. I feel that it needs a bit of tweaking there. What about: "5000 edits in total, admin here and on a content wiki, considerable work done on the content wiki and meta concerning 'vandalism', a unified account, and a fair number of edits accross a fair number of wikis concerning cross wiki cleaning and no large conflict situations on the cross-wiki part". --Dirk Beetstra T C (en: U, T) 09:21, 4 June 2008 (UTC)[reply]
That sounds good to me, too. --Millosh 09:33, 4 June 2008 (UTC)[reply]
5000 edits (home) + 1000 edits (across wikimedia) is very reasonable and yes, those with no vandal fighting background would really have no need for the rights anyways....but it should apply only to wikimedia wikis, and the requirement for the user to be active on Meta per herby is also a really good idea...--Cometstyles 09:35, 4 June 2008 (UTC)[reply]
I strongly oppose 5000 edits at home (because even don't have it :). 5000 edits (globally) + 1000 edits (home) + common sense seems to be fine — VasilievV 2 09:38, 4 June 2008 (UTC)[reply]
The number should not be carved in stone, it should be met with some reasonability, and if someone has 2 home wikis with each approx 3000 edits, and is admin on both (or other combinations etc.), that should be discussable then. But as a starting point 5000 should be good. --Dirk Beetstra T C (en: U, T) 09:42, 4 June 2008 (UTC)[reply]
{ec) There has to be "wriggle room" here.
There are those who I think most would agree should be allowed to request these rights & Beetstra is certainly one of those in my mind. However we do need to consider the fact that there will be a rush of people who suddenly decide a new badge is the in thing to have (sorry getting old & cynical :)). Getting 5K edits with even popups is easy - it seems some of the newer stuff could get it done very quickly.
I think we must consider the qualitative contribution as well as the quantitative here. They must be actively (currently) doing things that need to tools on multiple wikis to me - I'm not certain I qualify (& feel free to look at luxo). There really are a very small number of Wikimedians that could be considered for this role in my view --Herby talk thyme 09:43, 4 June 2008 (UTC)[reply]
And as a ps I think you have to have 5K edits somewhere --Herby talk thyme 09:43, 4 June 2008 (UTC)[reply]

I think that 5k edits shouldn't be a local, but a global requirement. 1k edits may be a local demand. Note that we are talking now about a lot of edits. --Millosh 10:11, 4 June 2008 (UTC)[reply]

(meh & I said I wouldn't edit :)) A point was made up a bit about 3K+3K being ok & that seems an idea BUT there must be quality & quantity across wikis currently (all important words IMO).
5k is no biggy - I have close to 10k on two (today maybe 10k on both :)) and close to 5k on another & I'm not sure I am experienced enough....--Herby talk thyme 10:17, 4 June 2008 (UTC)[reply]
The limit on the home wiki is a difficult one. 5k on en is not a strange number, 5k on a small wiki is something else. --Dirk Beetstra T C (en: U, T) 10:19, 4 June 2008 (UTC)[reply]
Yes, when there are no a lot of articles, you can't make minor edits for edit count; but you have to write/translate articles. It is very easy to me to get 100+ edits per day at en.wp; but, much more difficult to sr.wp (while I have 15k+ edits there, but I am there from the first half of 2004). --Millosh 10:22, 4 June 2008 (UTC)[reply]
(If I keep editing this thread enough I'll have 5k on Meta :)). I still say - quantity & quality & currently active - there has to be room for discretion. Reflecting I actually have three 5k wikis (including my original username) & one close to but if I'm not actively working across wikis on something (spamming included) I would not need or justify the rights - I know I have strong feelings about using rights but there it is - if you need them, get them, if you are not using them hand them back. Sadly this is not a commons view among those who actually have the rights --Herby talk thyme 10:47, 4 June 2008 (UTC)[reply]
Well, if you are performing 5k of edits giving sensible thought to this thread, then you certainly earn to be a global admin ;-). I indeed to believe that, like being an admin here, it might be good to have a semi-annual review of the global admins (do they still use the tools, did they not get into major trouble etc.). --Dirk Beetstra T C (en: U, T) 11:06, 4 June 2008 (UTC)[reply]
Annual confirm is already added. --Millosh 11:09, 4 June 2008 (UTC)[reply]

Conclusion?

May we try to conclude something? My proposal is: --Millosh 11:02, 4 June 2008 (UTC)[reply]

I think that this is fair. We are searching for cross-wiki editors and the most important number is total at all wikis. It is not reasonable to demand the same amount of edits at the home project and the second one if numbers are high. --Millosh 11:02, 4 June 2008 (UTC)[reply]

Added (after one day without comments). --Millosh 11:01, 5 June 2008 (UTC)[reply]
I disagree with the edits to Meta requirement, for a number reasons. (a) It's easy to run up a couple of hundred edits here doing nothing terribly useful (voting in polls, discussing inter-wiki links and such), (b) those edits are often not indicative of the level of knowledge we should be looking for (c) it's an additional obstacle that could prevent a few interested users from qualifying and (d) there's no indication what use the edit requirement here, or indeed, elsewhere will actually achieve. They look like completely random, arbitrary numbers. Is someone with those exact number of edits better suited to the AVF than someone who doesn't ? Nick 11:24, 7 June 2008 (UTC)[reply]
Do you have some formula for calculating those numbers? If so, please share it with us. As I see, you want to remove number of edits at Meta. OK. Anything else? --Millosh 13:53, 7 June 2008 (UTC)[reply]
Removal of an edit count requirement at Meta would be one thing, and the potential to remove an edit count requirement completely, so how about 5000 edits OR 6 months experience on the home wiki, and 100 edits or 1 months experience on another wiki, all of which could potentially be waived if the user has permanent sysop on those wikis (unlikely, but plausible). Edits don't equal experience and I feel it's easy to make the necessary number of edits to qualify for the AVF position without learning anything about the local projects policies and the way things are done around the WMF sites. Nick 00:45, 8 June 2008 (UTC)[reply]
That's ok because we have an RFP before they get tools. If they don't know what they're doing they won't pass muster. The edit requirements are just so we can set a lower bound on who may apply to avoid having our time wasted by people who shouldn't even be applying.  – Mike.lifeguard | @en.wb 03:23, 8 June 2008 (UTC)[reply]

Levels?

Might I suggest we focus more on a criteria that reflects the type of activity we would expect someone applying for this right to be doing, rather than at pure edit count. What we really want is someone with a history of cleaning up vandalism on small wikis - you don't need sysop rights to start doing that (some of it just needs reverting). Isn't what we're looking for someone who has spent say a month or two actively reverting vandalism on small projects? If the Global rollback permission comes into effect, could we perhaps require that people spend some time using that lesser right on global wikis without incident before they can have this right? Use global rollback as a training ground, and as an indicator that the user is truly willing to spend a reasonable amount of time on small wiki cleanup? WjBscribe 09:45, 4 June 2008 (UTC)[reply]

Yes, it is possible. We may make agenda of introducing new permissions to the group. --Millosh 09:54, 4 June 2008 (UTC)[reply]

Just because you don't have 5000 edits, doesn't mean its a bad idea ;)...i think people haven't got this idea clearly in their head , we are talking about "Global sysops" (or whatever name is chosen), the ability to be admins on 730 wikis on wikimedia, the ability to block ips and delete pages, this is not something small like Global rollback, this is a very dangerous user right and in the wrong hands can do more damage than good, and having such minute edit requirement is really very silly, this will really create problems on bigger wikis like enwiki, dewiki, frwiki and plwiki and most of these wikis will definitely disagree with this idea, so setting such a "low" requirement is just calling for trouble, I will also like to add a new proposition to the policy i,e,

  • It can only be given to those that have been on any wikimedia wiki for over a year and only to
  • those users who have never been banned on any wiki.
    these requirements will make sure that it doesn't go into the hands of socks, meatpuppets cross-wiki trolls i.e users banned on one wiki for trolling/vandalism/socking etc....--Cometstyles 09:56, 4 June 2008 (UTC)[reply]
I don't think I said anywhere that there shouldn't be a requirement in terms of edits/times served/being a sysop somewhere. What I was proposing was that people should spend some time with the lesser right (global rollback) to demonstrate they can use a cross-project right without incident, before being given the greater level of global access. WjBscribe 10:05, 4 June 2008 (UTC)[reply]
We may make some levels: (1) during the first month user has global rollback and something more; (2) during the second, user has something more; (3) during the third user has full global rights (as described at the policy page). --Millosh 10:08, 4 June 2008 (UTC)[reply]
(stopping editing here now toooo many ecs!) Comets speaks wisely & to WJBS I did use the word "qualitative" above & I think there must be that. Not totally happy with the "never been blocked" one - if you use Luxo with blocks you will see I've been blocked on two wikis - but it wasn't me honest :) I had to get some imposter accounts cleared for SUL --Herby talk thyme 10:10, 4 June 2008 (UTC)[reply]
Cometstyles did say "banned" not "blocked". WjBscribe 10:12, 4 June 2008 (UTC)[reply]
So what's the difference between the two? - Andre Engels 08:51, 12 June 2008 (UTC)[reply]
Ok I stand corrected --Herby talk thyme 10:18, 4 June 2008 (UTC)[reply]
Thanks for pointing that out, banned users pose more of a risk but if a user was blocked for small violations like 3RR or minor civilty issues, then its not really a problem :).. --Cometstyles 10:25, 4 June 2008 (UTC)[reply]
Personally I would not be very happy with anyone with a block log entry getting cross wiki rights (I'm not saying "don't let them have them" - I'm saying I would not be that happy about it --Herby talk thyme 10:49, 4 June 2008 (UTC)[reply]
As I said, we may make some "transitional period" with introducing less controversial permissions firstly. And about global rollback... We (stewards) don't have a consensus about introducing anything without asking community. So, we are asking the community once (this proposal) instead of a number of times (per permission). --Millosh 10:03, 4 June 2008 (UTC)[reply]
Getting blocked when doing cross-wiki clearing is very easy, I have seen someone being blocked after doing cross-wiki cleanup .. if it is banned then indeed that should disqualify for this tool, and large conflicts (outside the home wiki(s)) should be brought to discussion. --Dirk Beetstra T C (en: U, T) 10:23, 4 June 2008 (UTC)[reply]
Guidelines should be written for those purposes. If some global sysop is blocked because of some crazy admin(s) (for example, admins who are treating real nonsenses [so, not articles like "2 is a number bigger than 1 and lesser thatn 3"] as valid articles), global sysop shouldn't lose their privileges, but if blocked because of making troubles, they should. In the first case, they should leave to the stewards to solve the problem. --Millosh 10:31, 4 June 2008 (UTC)[reply]

I am certainly a fan of WJBscribe's idea of requiring potential AVFs to have a period with lesser permissions globally (ie global rollback). This has been sort-of done at my home wiki of en.wikibooks, where we have just given patroller and rollback status to someone who is on their way to admin but isn't quite ready. This lets us watch their work with those tools while they gain the necessary experience. Similarly here, we can require that people get global rollback first. Then, provided they behave themselves and do good work and (etc...), they may request more global tools.  – Mike.lifeguard | @en.wb 15:11, 4 June 2008 (UTC)[reply]

Same here. I think being a global rollbacker is a good start before becoming an AVF. On a side note, what is the difference between a sysop and an AVF? Besides the latter having the rights globally. --Erwin(85) 15:47, 4 June 2008 (UTC)[reply]
A local sysop voted by the local community can decide what to delete, some things without needing to have requests for it. An AVF should not decide what to delete on a local wiki but only intervene if there is ongoing vandalism and no local available. An AVF should leave the controversial stuff (like stubs) etc. alone. Best regards, --birdy geimfyglið (:> )=| 15:58, 4 June 2008 (UTC)[reply]
Fully agree with birdy. Cross wiki work is a sensitive area & needs handling with sensitivity. When I stray onto other language wikis I am very conscious that the links I see as "spam" may not be seen the same way by locals & that I need to tread carefully --Herby talk thyme 16:06, 4 June 2008 (UTC)[reply]
Indeed, but I meant the technical difference. AFAIK an AVF can do all things that a sysop can do, correct? It speaks for itself though that AVF's should only handle spam and vandalism. --Erwin(85) 08:00, 5 June 2008 (UTC)[reply]

Conclusion?

May we say something like: "For the period of the first month AVf will get only 'Editing permissions'. After that time, people will be able to vote to oppose a AVf. If there is no opposition for AVf will get full rights. If opposition is voiced, then the AVf may lose AVf status if support falls below 75%. No quorum is required. It is not a vote to gain support status, but a poll to express disagreement with the current situation. The point is not to bug everyone to vote to support the AVf again (if there is no opposition, there is no point in voting your support again. If a AVf is not really strongly infringing rules, but is creating work for the community because of a lack of trust, then it is best that people have the possibility of expressing their opposition." (Mostly copied from the part related to confirmation.) --Millosh 17:09, 4 June 2008 (UTC)[reply]

I don't understand what you mean by "editing persmissions" - anyone can edit the wikis already. Do you mean global rollback? If so, I think there will be a process to request that too, so... this will not work as suggested.  – Mike.lifeguard | @en.wb 00:30, 5 June 2008 (UTC)[reply]
See Anti-vandal fighter#Editing permissions :) --Millosh 06:21, 5 June 2008 (UTC)[reply]
So, you mean that someone should request for both permissions? I think that we should make the process easier. User is asking once for permissions, but they have period to prove themselves (it may be one moth, but it may be three months, too; whatever we decide). And AVf has a confirmation process after that period. Note that stewards don't have such conditions for their rights. We are here around a step too far. --Millosh 06:27, 5 June 2008 (UTC)[reply]

May we conclude here to add my proposal or someone has a better idea? --Millosh 13:46, 6 June 2008 (UTC)[reply]

I think splitting these permission up is a tad silly, and increases bureaucracy unnecessarily. Keep in mind that we will have (hopefully) sane people discussing the potential AVF in an RFP, so there is no need to add layer upon layer of useless bureaucracy. If people wish to have a trial period, then a while with global rollback should be sufficient. With this suggestion, once would get then global rollback, then global "editing permissions", then full "AVF permissions"?!
As a side comment, Millosh, you may wish to stop pushing this proposal so hard. Adding things to the proposal "after 24h with no comments" is a tad pushy. This needs to have a very high level of consensus, which you are not helping by being so single-minded. Please chill out a bit.  – Mike.lifeguard | @en.wb 15:00, 6 June 2008 (UTC)[reply]
No one said in the previous section anything about the proposal for a day, after very frequent posting. No one said nothing above, until now. As well as everything may be changed if someone disagrees again. There are a lot of potential issues regarding the policy to be solved, so I am trying to make things more efficient. --Millosh 20:56, 6 June 2008 (UTC)[reply]
About "levels": I have to say that I didn't understand you what do you want. I don't think that any kind of testing period is needed. However, you said that you want some kind of testing period -- assuming that those persons should be admins before their candidacy, as well as raising requirements for their candidacy. Global rollback is a lesser permission than being a local admin. --Millosh 20:56, 6 June 2008 (UTC)[reply]
So, the question is: Do you think that it is really necessary to have "rollback" level before or not? Or, what is your proposal? (Again, I am fine without any change.) --Millosh 20:56, 6 June 2008 (UTC)[reply]

General requirements

Separating the new issue... So, we need a set of "general requirements" like "user is not disruptive", "user wasn't banned [not blocked]" etc. May someone make some wording for that for the policy? --Millosh 10:36, 4 June 2008 (UTC)[reply]

My proposal for general requirements is: --Millosh 15:32, 5 June 2008 (UTC)[reply]
  • Candidate should be technically skilled. They should show that understands issues like: (1) General knowledge of Internet: IP addresses and ranges, open proxies, manipulating with web based and host based tools like "nslookup/host", "whois", RIPE site etc. are. (2) Good (but not perfect) knowledge of wiki syntax. (3) General knowledge in wiki concept and Wikimedia projects. (4) Manipulating with bots would be plus. --Millosh 15:32, 5 June 2008 (UTC)[reply]
  • Candidate should show that they are trustworthy: (1) Candidate didn't show disruptive behavior. (2) Candidate is following Five pillars in their Wikipedian work or equivalent in their work at other projects (which should be defined). --Millosh 15:32, 5 June 2008 (UTC)[reply]

For the technical skills, candidate may show their work or may wait to be asked. Examples for the questions may be: (a) Describe with your own words what would you do if you see that someone made vandalisms from IP addresses 70.70.70.5, 70.70.64.3 and 82.82.90.1. (b) Write a wiki code for table which foreground color is green and background color is blue in such manner that I may see your code clearly. --Millosh 15:32, 5 June 2008 (UTC)[reply]

For the trustworthy issues, voters should be able to show examples which are proving that a candidate was showing a disruptive behavior or that candidate doesn't follow Five pillars or similar in their work. --Millosh 15:32, 5 June 2008 (UTC)[reply]

Sorry, some of this seems silly to me. Since these users will be doing only simple "antivandalism" work, what justification is there for a requirement about being able to make a table? Equally, some may well be able to serve in this role without knowing detailed info about rangeblocks. And what's the deal with making then write a little essay? No, I think the normal model of granting permissions is fine; this would be rather restrictive.  – Mike.lifeguard | @en.wb 20:21, 5 June 2008 (UTC)[reply]
The other their role is to help to the communities, including dealing with wiki syntax. By the role, they will be maybe the most important contact between the global community and some small communities. We shouldn't send to some small local community with not a lot of touch with the rest of the community a contributor who is not able to show reasonable level of knowing wiki syntax. Making table is not so hard (it may be copied). --Millosh 20:29, 5 June 2008 (UTC)[reply]

So, is it possible to conclude some set of general requirements (not necessarily related to questions) or not? If so, please, propose your version. --Millosh 13:45, 6 June 2008 (UTC)[reply]

No. Each person will have their own (hopefully rational) set of ideals against which they will test the candidate. The normal RFX process is fine; we need no further constraints.  – Mike.lifeguard | @en.wb 15:02, 6 June 2008 (UTC)[reply]
So, [your] conclusion is not to change anything? (I am fine with it.) --Millosh 20:57, 6 June 2008 (UTC)[reply]

Voting for candidates

I think that voting for candidates shouldn't be simple "yes/no", but a list of explained reasons why some candidate should or shouldn't be an AVf. Votes without explanation and links shouldn't be counted. Votes expressing only that a voter like or not a candidate shouldn't be counted. Voter has to give all answers: --Millosh 15:32, 5 June 2008 (UTC)[reply]

For example, one vote should look like: --Millosh 15:32, 5 June 2008 (UTC)[reply]

  • Against --~~~~
    • Technical skills:
      • General knowledge of Internet: Candidate showed enough of knowledge.
      • Good knowledge of wiki syntax: Candidate showed enough of knowledge.
      • General knowledge in wiki concept and Wikimedia projects: Candidate showed enough of knowledge.
      • Manipulating with bots: Candidate manipulates with bots very well.
    • Social skills:
      • Disruptive behavior: Candidate showed disruptive behavior at a number of situation [1][2][3][4][5].
      • Following Five pillars: Candidate partially follows Five pillars: Problems with NPOV [6][7][8] and problems with civility [9][10][11].
    • Rationale: While candidate is technically skilled, candidate showed a high level of disruptive behavior and because of that permissions and rights shouldn't be granted to them.

Voting yes/no is a really bad way for giving trust to someone. At the other hand, expressing full position toward a candidate will give to voters much higher responsibility. If they are using lies or obstructive methods for or against a candidate -- it will be prove that they are obstructive, so they wouldn't be able to count on the same permissions in the future. --Millosh 15:32, 5 June 2008 (UTC)[reply]

Of course, I added some of the "general requirements". You may add other requirements or make better concept or wording for requirements classification. --Millosh 15:32, 5 June 2008 (UTC)[reply]

Similarly here... requesting permissions should never be akin to checking boxes off on a form. Rather we should be discussing things as we do normally for RFXs. This idea has me baffled :(  – Mike.lifeguard | @en.wb 20:21, 5 June 2008 (UTC)[reply]
Such formalized way of voting should be of some help in passing simple support/oppose options. At least, voters will have to express their position toward a couple of issues. We may model those questions for voters during the time. --Millosh 20:34, 5 June 2008 (UTC)[reply]
I agree with Mike, this is getting overly bureaucratic IMO... we don't need this many specific guidelines. RfX's have worked fine for Wikimedia so far. Cbrown1023 talk 23:46, 5 June 2008 (UTC)[reply]
Whatever... In 100% of polls I see 80% support/oppose votes without any explanation or even with explanations without a lot of sense. Check, for example, last steward elections. People are voting for their friends and against contributors which they don't like. At some point it should be stopped, but it is obvious that this is not the right time for that. --Millosh 21:04, 6 June 2008 (UTC)[reply]

Guidelines

If policy successfully adopted, we need to write guidelines for this role. So, let's start. --Millosh 10:41, 4 June 2008 (UTC)[reply]

  • AVf should have a typical template at their user page where users should be introduced where to complain about AVf's behavior. --Millosh 10:41, 4 June 2008 (UTC)[reply]
  • We should have such page here. --Millosh 10:41, 4 June 2008 (UTC)[reply]
  • Some more description around what should be treated as abusing rights (making troubles to a community) should be added there. Also, there should be written what is not an AVf's mistake and how to react in such circumstances (generally, to leave some steward to deal with it). --Millosh 10:41, 4 June 2008 (UTC)[reply]
  • Don't delete stubs. --Millosh 17:15, 4 June 2008 (UTC)[reply]
    • Against. "Don't delete stubs _just because they are stubs_." might be ok (but even so, deleting a one-word article on a project without any users doesn't seem like a bad thing to do). But a rule that would allow them to delete a page consisting of the word "assholes" 200 times, but not one consisting of the word "assholes" 3 times is just plain stupid. - Andre Engels 08:47, 12 June 2008 (UTC)[reply]
  • Be careful about treating something as spam. Some cultures are much smaller (than English or German) and something which may be treated as a spam inside of bigger cultures may be completely valid information in smaller ones. --Millosh 17:15, 4 June 2008 (UTC)[reply]
    • Disagree. Spam is spam. - Andre Engels 08:47, 12 June 2008 (UTC)[reply]
        • spam is not spam a quick reading of en:Wikipedia:Spam will make it clear that there are cases where what looks like spam is allowed. Now put it in other languages and cultures and it becomes even less clear, so the warning by Millosh is very valid. Vandalism is much easier to identify than spam. Dbiel 14:34, 13 June 2008 (UTC)[reply]
          • In practice it is frequently the behaviour that makes some editing "spam". I far prefer "excessive external link placement". Where an IP (or a user) edits a number of wiki (or just one) only placing links which are removed by users whether local of global in perspective then that is "spam" as we speak of it here. If I ever were a global sysop then I certainly would be using the rights on that basis - not lightly but with some experience --Herby talk thyme 08:10, 13 June 2008 (UTC)[reply]
            • I agree that "excessive external link placement", which is only one portion of Wikipedia spam, is an area that GS members can be very effective without the need for much care, but other types of Wikipedia spam do require care and in many cases should simply be ignored by GS users unless they have a high degree of understanding in the specific wiki they are editing. Dbiel 14:14, 13 June 2008 (UTC)[reply]
              • Sure - care comes with experience. I currently remove spam on a number of non en wikis etc. I look carefully before I do it. I try & find a local sysop for advice if I am unsure. I've been reverted once out of quite a lot of edits. The role we are looking at here is most certainly not for the inexperienced or unthinking folk. Thanks --Herby talk thyme 14:26, 13 June 2008 (UTC)[reply]
      • I said just "be careful". Some cultures really have market as an inherent part of their society (Kenya is the example). So, it is possible that giving links to the commercial sites with short description may look like a spam, but it may be not a spam. Maybe to reword this? --Millosh 09:57, 12 June 2008 (UTC)[reply]
  • If you want to use your permissions on larger projects, you may ask the community for yourself or for a general approval for all AVfs. --Millosh 11:03, 5 June 2008 (UTC)[reply]
  • Content disputes are not your business. You should leave it to the local community. --Millosh 16:29, 5 June 2008 (UTC)[reply]
  • Trolling is not your business. If some troll disrupts a project without admins, report it to stewards. If a troll is marked as such by the community and community without sysops (or the only sysop is a troll) clearly asks for your help, you may help them. However, in the later case, you have to consult others. Note, also, that the goal in such situations is to have more local admins who are able to deal with such problems. --Millosh 16:29, 5 June 2008 (UTC)[reply]
  • If you abused your rights your rights will be removed. However, it is very likely that you will get indefinite block at the local project, too. --Millosh 00:55, 7 June 2008 (UTC)[reply]
  • ...

From the English Wikipedia, and information of wikis

There is a discussion on en: at Wikipedia:Wikipedia_talk:Administrators#New_global_userright. Users agree there that there are enough rollbackers and administrators to deal with vandalism and that timezone is not an issue. So asking the community on en.wikipedia for usage of AVF tools would mean passing the mandatory processes (w:wp:RFR for rollback rights and w:wp:RFA for admin tools). Some are talking of a possibility to opt out too. En: is a special case, but I think that other major wikis will be concerned by this as well. I suppose that at least all major wikis should be informed (I just informed fr:). Cenarium (talk) 02:49, 5 June 2008 (UTC)[reply]

Thanks for your comment on enwiki just to inform you"Local projects must be respected. Even though their permissions are global, anti-vandal fighters are not allowed to use their rights on any project with a substantial community of active administrators (e.g. the English Wikipedia) without explicitly asking the community, even for such minor things as using rollback. Anti-vandal fighters not respecting this rule will lose their privileges immediately."see Anti-vandal fighter --Mardetanha talk 02:55, 5 June 2008 (UTC)[reply]
I am aware of this, as you can see from my comment on en:, "This is pretty clear in the proposal, en.wp is given as an example where the AVF tools should not be used without consent of the local community". I am merely stating facts, I commented here in the hope to discuss the issue further. Cenarium (talk) 03:04, 5 June 2008 (UTC)[reply]
I think that it's important to inform the local communities, so that we can have a wider view and bring them in the discussion. I'm new at meta so I don't really know how things work. Cenarium (talk) 03:08, 5 June 2008 (UTC)[reply]
Hmm .. I have a bit of a problem with this part of the discussion. Spammers (and I expect vandals) tend to hit the top wikis (mainly because the interwiki links are better for those). So we do get global rights (lets say rollback), and then we can't use it for the biggest part of vandalism/spam because we have to let it be done by locals, as they have enough rollbackers, until we also go through the whole procedure on all those wikis? So then it is just as easy for me to go through all the wikis and ask for rollback on them, and when I have rollback on the top 40 (which probably I will not get on number 35, as I only have 4-5 revert edits there), I ask here for global rollback? --Dirk Beetstra T C (en: U, T) 10:22, 5 June 2008 (UTC)[reply]
Some people at the mentioned page on en.wp expressed positive statement toward giving to this group a rollback permission. So, AVfs will have it, but they should ask such communities for using rights. And I am almost sure that it will pass because it is not a big deal. This policy has to respect developed local communities. Because of that, Meta is not the right place for talking about permission usage on larger projects. --Millosh 10:40, 5 June 2008 (UTC)[reply]
In that case we need a multilingual guide here explaining global right, with a boilerplate explanation of why this user has global right (in each language). When acquiring global rollback here, that user has to go to the rollback-request-page on all these wikis, with a formal message (which can then point back to the boilerplate message here on meta) why the user has global rollback, with formal request for admins to allow them to use it (I expect some admins on the e.g. Polish wikipedia to speak English, but the majority may not speak it (sufficiently), nor can I understand replies).
Do we also have to ask all wikis whether they want to be asked on a user-by-user case, and compile lists of where we have to acknowledge and where not? Also, do we have to excempt the wikis where we have to acknowledge from the global rights, as I am afraid that we will make mistakes in the heat of vandalism-revert actions (rollbacking on wikis where we should not, but if I have to revert all these (did not check if I have to!) I am sure I will forget that I should not have used it on one or two ..)? --Dirk Beetstra T C (en: U, T) 12:06, 5 June 2008 (UTC)[reply]
I see this as one of the first issues for AVfs group. Contributors elected as AVfs should organize themselves at the best possible way to avoid endless asks for permissions by every AVf. --Millosh 14:16, 5 June 2008 (UTC)[reply]
Some issues should be solved at the beginning of their work as SWMT members: --Millosh 14:16, 5 June 2008 (UTC)[reply]
  1. To define what precise which project is not active, which needs maintenance during some part of the day and which doesn't need any kind of AVfs activity. Of course, it should be organized transparently, at Meta. --Millosh 14:16, 5 June 2008 (UTC)[reply]
  2. Finding a way to monitor and measure progress of the small projects. Even finding what some project needs to become a mature one (ok, a number of active admins -- or, better, hours of activity of admins; but how to reach needed numbers?). --Millosh 14:16, 5 June 2008 (UTC)[reply]
  3. To make all necessary documents related to their work. One of them is a template "I am AVf and you may complain about my work here." which should be put at every user page of one AVf. The other is a document which describes an ask to large projects: which permissions we may use? Etc. --Millosh 14:16, 5 June 2008 (UTC)[reply]
It might be a good idea to get input from as many communities as possible regarding the proposed policy, no matter how small or large (maybe run a notification bot, or something like that). It is reasonable to expect not all will approve, but addressing such concerns should help ease the potential introduction of this user group. It would also assure an adequate representation of the global community in voting to adopt this policy (which is only fair) and allow everyone to get familiar with the work these volunteers would do, before they actually get to it. Just a thought, anyway... --Кале 14:48, 5 June 2008 (UTC)[reply]
Write such bot for me and ask for permissions to use it at all projects :) I think that this idea is a good one, but not so possible. I was willing to make such tool a couple of months ago, but a number of contributors was against such idea because it would be "dangerous", "spamming" or whatever. There is no reasonable method even for sending messages at all projects. List of Wikimedian pubs (the page started, also, by me) has only 25 listed pubs/village pumps of 700. Plus some of them (like ru.wp and da.wp) have weird methods for adding a comment. Unfortunately, according to present technical possibilities, community may be informed only through foundation-l mailing list. --Millosh 15:39, 5 June 2008 (UTC)[reply]
Of course, it doesn't mean that you are not free to inform as many projects as you are able to inform; as well as to organize such effort with other contributors. --Millosh 15:39, 5 June 2008 (UTC)[reply]
I think this discussion is unneeded because the vandal fighters should/will work mostly on small wikis with only some and/or inactive admins and not on big projects like en:, de: or fr: because there are really enough admins that can deal with vandalism themselves very easily. --MF-W 18:38, 6 June 2008 (UTC)[reply]

Hmm, no, I don't think this is the same, and I think this needs discussion. Yes, vandalism on any wiki should be dealt with by the editors/admins on en. But what AVF should be concerned with is the vandalism on wikis where there are no(t enough) administrators, and with (large/wide scale) cross-wiki vandalism/POV-pushing/spamming (which may seem small on the individual wikis). When possible, the latter should be handled by the local administrators, but when the activities are ongoing, disrupting, and clearly cross-wiki the AVF should have the possibility to perform his actions also on the big wikis, to minimize disruption, and succes of the vandal/spammer/POV-pusher. I therefor hope that the individual wikis are not going to say "no, we don't want to see AVFs here at all", but I hope they are going to say "They should not be using their abilities here without first asking if they can use it, and then only when it is to stop and revert blatant cross-wiki disruption." And then still they should try to let the local community do most of the work (note: even on en there sometimes are posts on en:WP:AN that there is a large backlog on en:WP:AIV, and bots can do a lot of damage in a short time). --Dirk Beetstra T C (en: U, T) 17:33, 7 June 2008 (UTC)[reply]

Technical restrictions?

It's completely possible that I'm simply overlooking something in the proposal, but is there any way to restrict use of the tools by AVFs on a technical level? While I know that only trusted users and current sysops would be selected as AVFs, it seems that the whole issue could be solved (and a lot of feathers unruffled) if there was a way to allow individual wikis to opt out from even allowing AVFs access to the tools. --jonny-mt en me! 01:24, 6 June 2008 (UTC)[reply]

Permissions are global and, AFAIK, there is no possibility to exclude it technically, except by some hardcoding MW or database. However, all large projects are by default out of this. Usage of the permissions at those projects without community approval will lead to permissions removal. --Millosh 11:34, 6 June 2008 (UTC)[reply]
It will need to be made clear to people with this access that any usage on major Wikis, even accidential, will result in indefinite blocks unless it was self-reverted immediately and did not cause any harm (eg. not a block of an established contributor, which always results in drama even if immediately self-reverted). If it was not malicious or deliberate, they could obviously be unblocked when the tools are removed and they promise never to do such again, should they get the tools back at a later date. Most likely, if it can be proved the act was deliberate and made knowing that the wiki was excluded from AVF scope, the user will be considered banned from that project for breaching trust and abusing global tools. I will certainly advocate for such a position on the wikis I am active on, in the event that someone deliberately and knowingly uses AVF tools on the likes of enwp, enwn, and Meta. Daniel (talk) 23:42, 6 June 2008 (UTC)[reply]
But, it is related to the local policy. As nobody really thought about policy related to the global blocks (only spammers and open proxies were discussed), it is not possible to add that somebody will be blocked globally. The only action which may be done globally is the rights removal. However, particular projects may adopt some policies (or they don't need because similar policies exist; e.g. admin rights abuse regulation may be good enough for this purpose, too) and it may be added inside of AVf policy that abusing rights at a particular project may result not only with removal rights, but, also, with a local block. But, I don't think that it is a matter of a global policy. If someone abused rights, stewards will remove their rights immediately and local community may block such user before or after rights removal (and abusing rights with removing a block may result urgent action by stewards). --Millosh 23:57, 6 June 2008 (UTC)[reply]
Sorry for the confusion; yes, I meant blocked locally, of course. Daniel (talk) 00:06, 7 June 2008 (UTC)[reply]
I think that such information should stay inside of the guidelines. Policy doesn't deal with well developed local communities and it is up to the local communities what to do with abusive users in the sense of blocks. So, something like "Note that you may be indefinitely blocked at the local project if you abuse your global rights." should stay inside of the guidelines (and I'll add it now at the section "Guidelines" at this talk page). It may stay inside of the policy, but I don't see a lot of sense: definitely, they won't be blocked by a steward, but by a local admin. --Millosh 00:21, 7 June 2008 (UTC)[reply]

And about projects lists: This should be one of the first jobs after the policy adoption. I mentioned above some of the first jobs for AVfs. This should be discussed by the interested communities, AVfs and stewards. --Millosh 00:03, 7 June 2008 (UTC)[reply]

Anti-vandal fighter/Wikis. Thoughts? Daniel (talk) 00:20, 7 June 2008 (UTC)[reply]
I rearranged your page. Projects without enough of admins are not able to decide. Consensus should be reached about which particular projects are small and which are not. --Millosh 00:26, 7 June 2008 (UTC)[reply]
If its Global and if there is no opt-out option, then I believe we would have to make the requirements a bit more" harder" since big wikis like enwiki and dewiki have strong RfA policies and I don't think they would want someone with 1000 edits on enwiki have global rights, since I believe enwiki's policy says 3000...--Cometstyles 00:31, 7 June 2008 (UTC)[reply]
I already added stronger requirements. See the section at the policy page. --Millosh 00:35, 7 June 2008 (UTC)[reply]
As well as the bottom of the section #Number of edits. --Millosh 00:37, 7 June 2008 (UTC)[reply]
Oh great,sorry for not seeing that...:) ..--Cometstyles 01:44, 7 June 2008 (UTC)[reply]

Difference between Admin and AVF?

Can someone list the differences in tools between Admin and AVF? Why not just give the AVF the full Admin privileges? Why bother restricting edits to full-protect pages when the AVF can simply change the protection level and then edit? Why bother restricting anything when the AVF can edit user JS files, the most sensitive admin function, with which they can impostor login screens to obtain user passwords? —Centrxtalk • 02:05, 7 June 2008 (UTC)[reply]

AVFs are not administrators of local wikis. If a local community decided to protect some page and AVF unprotects it, AVF will lose their privileges immediately. So, editing protected pages is forbidden for AVFs. --Millosh 14:01, 7 June 2008 (UTC)[reply]
That should be outside of their scope, anyway. They should be concerned with large-scale active cross-wiki vandalism, which can not happen on a page which is fully protected (except when an admin/AVF would be the culprit, in which case the situation should also not be handled by the AVF, but on 'a higher level'). For all other edits (the good faith ones) they should restrain themselves. --Dirk Beetstra T C (en: U, T) 17:41, 7 June 2008 (UTC)[reply]

If they be focused on vandalism, why should they be allowed to edit user JS files? —Centrxtalk • 19:18, 7 June 2008 (UTC)[reply]

Because vandals may use scripts to disrupt the wiki. Also, I think this is not a separate permission, so it cannot be removed without also removing other permissions we clearly want them to have.  – Mike.lifeguard | @en.wb 20:03, 7 June 2008 (UTC)[reply]
So, then, is the role of AVF solely to deal with matrix vandalism and disruption, or is it to assist in administration on wikis with few or no admins? From reading the proposal, I thought the latter was the case; either way, there are issues.
Cross-wiki vandalism is incredibly hard to identify: how are AVFs supposed to identify such cases? There is no Special:GlobalSearch or Special:GlobalRecentChanges. Vandalism on individual wikis is identified and dealt with, but cross-wiki vandalism can't be dealt with by AVFs unless it's identified, and expecting every single vandalism block to be cross-referenced across all WM wikis is ludicrous. In the case of persistant vandalism which isn't being contained by the application of blocks, which an editor identifies (through whatever means) as cross-wiki vandalism, then a steward can be requested code 3 to apply emergency rights changes in order to combat the threat; indeed, this emergency insurance is one of the main reasons for the existence of the steward class, besides SRP actions on non-crat wikis.
Note: Crosswiki vandalism is not hard to identify... see SWMT, we have an IRC channel that all Recent changes for the small wikis are reported by a few bots, see #cvn-sw. Cbrown1023 talk 00:08, 10 June 2008 (UTC)[reply]
In the second case, where AVFs are designed to take on administrative roles in small projects with too few admins, we will be plunging people into projects which they are possibly unfamiliar with, in languages most of which they will not even speak, and expect them to "lend a hand" in performing controversial tasks like blocking, page protection and (mass) deletion. There are going to be huge barriers in communication between local administrators and AVFs, and the "locals" are inevitably going to end up clearing up any unwitting mistakes made in unfamiliar circumstances – after all, it's a lot easier to identify blatant vandalism in English than in Serbo-Croat – which could further add to their workload.
Let's cut the weasel wording – this is global adminship by any other name. What's more, instead of the consensus procedure used in project RfAs, RfAVF will be qualified by a majority voting system like WM:RFA; considering that adminship requests on large wikis can fail with a majority of 80% with enough serious issues raised, and the fact that AVF is a far greater responsibility and has far greater potential to be misused (not to mention the controversy over the usergroup existing in the first place), this has serious potential to go horrifically wrong in many cases. Who's going to keep tabs on the AVFs to ensure that they're not using their rights on wikis where they are forbidden to use them by policy? More importantly, who do they answer to? haz (talk) 10:27, 9 June 2008 (UTC)[reply]

Mass delete of pages

"As far as the proposer of this policy understands, the number deleted pages is limited. However, if a vandal already made a very large number of pages, anti-vandal fighter should be able to delete them with or without help of a program." - What do you mean by this? As far as I know, if you have the delete right, you can delete as many pages as you want. Do you maybe mean the bigdelete permission which allows you to delete pages with more revisions than $wgDeleteRevisionsLimit? --MF-W 14:28, 7 June 2008 (UTC)[reply]

I guess he's suggesting there's a throttle, so you can't delete more than a certain number of pages per unit time. The same is true for rollback on enwiki. You can't rollback more than a given number of edits per hour. --Erwin(85) 14:40, 7 June 2008 (UTC)[reply]
OK, but I think there's no throttle for deleting. And what is proposed? A throttle or no throttle? I guess it's the latter. --MF-W 15:05, 7 June 2008 (UTC)[reply]
I don't see the sense of this permission except regular delete is time limited. We should find the answer from developers... --Millosh 16:23, 7 June 2008 (UTC)[reply]
I think this is referring to Extension:Nuke. I don't think delete has any rate limits, or that avfs need bigdelete--Werdan7T @ 17:40, 7 June 2008 (UTC)[reply]
May you, please, research this and be sure about the conclusion? If you are sure, that permissions should be removed. --Millosh 18:17, 7 June 2008 (UTC)[reply]
There is no throttle on deleting pages. And the bigdelete permission is absolutely not required. However, I think access to Special:Nuke would be a good idea (ie it would have to be installed on all wikis then, as it is an extension). However, I'm not sure that we can decide here to install an extension on all other wikis without a bigger discussion on that. (Though it should be uncontroversial, IMO. It doesn't give any ability sysops do not already have.)  – Mike.lifeguard | @en.wb 20:06, 7 June 2008 (UTC)[reply]

I removed mass delete from the list of permissions. --Millosh 18:47, 8 June 2008 (UTC)[reply]

Why? Having Special:Nuke on all wikis should be done.  – Mike.lifeguard | @en.wb 19:01, 8 June 2008 (UTC)[reply]
:) I realized that it was the conclusion. I reverted my edit. --Millosh 20:14, 8 June 2008 (UTC)[reply]

Hm. I think that this is not Special:Nuke. --Millosh 20:17, 8 June 2008 (UTC)[reply]

Sorry, I don't understand you. I'm saying that we should install mw:Extension:Nuke on all WMF wikis, and these folks should have access to it.  – Mike.lifeguard | @en.wb 21:09, 8 June 2008 (UTC)[reply]
May you add/change [into the policy page] whatever you think that it should be added? If you think something should be added/changed... --Millosh 03:33, 11 June 2008 (UTC)[reply]

Small and large wikis

Please, take a look at Anti-vandal fighter/Small and large wikis. Maybe the best criteria for classifying wikis into small and large is existence of checkusers. Please, talk here about the criteria and edit page there. --Millosh 15:11, 7 June 2008 (UTC)[reply]

There are some fairly large wikis that doesn't seem to have their own checkusers, like fr.wiktionary and zh. and no.wikipdia for instance. --Jorunn 15:56, 7 June 2008 (UTC)[reply]
Yes, I am aware of this and it is mentioned that border cases should be discussed. --Millosh 16:16, 7 June 2008 (UTC)[reply]
I've modified criterias and generated a list. Please, check if there are wikis which are in the small wikis list but should be in large — VasilievV 2 12:26, 8 June 2008 (UTC)[reply]
Shouldent the size criteria (number of articles) be different for different projects, e.g. the albanian wikibooks is the 10th largest (by number of articles) but it hasnt even got 3000 pages yet, while the albanian wikipedia has more than 20,000 articles and is 50th-60th ?--Cradel 15:33, 8 June 2008 (UTC)[reply]

I don't think that the number of articles is a good definite criteria, like having CUs are. A project may have a lot of articles, but it may be not self-sufficient (Volapuk case). Having CUs assume at least around 30 active contributors interested in community issues, which means that they have more than 30 very active contributors. --Millosh 18:44, 8 June 2008 (UTC)[reply]

And the first example which I found there is Yiddish Wiktionary, which is very far from a mature project. --Millosh 18:52, 8 June 2008 (UTC)[reply]

Hm. I realized now that Yiddish Wiktionary has 170 articles. What is the criteria for them to be a large wiki? --Millosh 18:56, 8 June 2008 (UTC)[reply]

yi.wiktionary is generous with the sysop rights. I am under the impression that anyone who comes by and revert vandalism or behaves like a friend of the project gets sysops rights, and there is now 51 admins. If there is/was a criteria that says more than 10 sysops=large wiki, that is probably the reason it ended up on the large wiki list. --Jorunn 23:09, 8 June 2008 (UTC)[reply]
Wikinews: As a Wikinewsian, I know that out of en, de and pl, there is no any "large wiki" at Wikinews. --Millosh 18:58, 8 June 2008 (UTC)[reply]

So, it seems that 10 active admins rule is not good enough, too. I removed "50.000" articles rule, but I will not change anything more there until I hear the rest of you who participated in the page creation. --Millosh 19:05, 8 June 2008 (UTC)[reply]

I realized that a very good principle is 50.000 articles and more than 10 admins. However, projects like en.wn don't have 50.000 articles, so such projects should be excepted from the list. --Millosh 10:15, 10 June 2008 (UTC)[reply]
Hm. But, en.wn has CUs :) So, the first principle applies. --Millosh 10:18, 10 June 2008 (UTC)[reply]
What's the point in arguing over semantics? No one's going to say: "it doesn't matter that you don't want global sysops to work on your project, you don't meet our criteria for being a 'large wiki' so we're going to 'help' you anyway". If a project doesn't want the assistance of these users, then their opt-out is guarranteed (although I doubt any wiki, large or small, would turn down assistance in fighting vandalism). Unless they have opted out, acceptance of global sysops is implied, surely? If a project is not large enough to want to exclude global sysops, they are not large enough to not need extra pairs of hands. Happymelon 16:20, 10 June 2008 (UTC)[reply]
The point is to find as better as possible rules which may be automatically applied and save our time in analyzing performances of the communities. If some wiki has more than 10 admins or more than 50,000 articles (but, not both) and doesn't have CUs, it has to be analyzed initially (if it is not obvious case, like yi.wikt [large number of admins, very small number of articles] or Volapuk Wikipedia [large number of articles, very small number of admins] are): we need to talk with those communities and realize are they self-sustainable. In all other arguments, I agree with you. --Millosh 17:36, 10 June 2008 (UTC)[reply]

Access to global blocking

Since global sysops are created to fight vandalism on small wikis, they need access to Special:GlobalBlock (IP blocks, not installed yet) and Special:CentralAuth (SUL account blocks, now only lock/hide, but there'll be also blocks in future) — VasilievV 2 11:06, 8 June 2008 (UTC)[reply]

Not that I disagree, but what is the difference with stewards then?  – Mike.lifeguard | @en.wb 17:30, 8 June 2008 (UTC)[reply]
Stewards have access to checkuser, oversight and granting/removing of user rights. Nakon 17:37, 8 June 2008 (UTC)[reply]
Well now I just feel silly :)  – Mike.lifeguard | @en.wb 17:45, 8 June 2008 (UTC)[reply]
I think that I said inside of the proposal that AVFs will work at global block with stewards when this feature become reality. --Millosh 18:45, 8 June 2008 (UTC)[reply]

Only concern

The only thing I can see where problems would arise would be with this:

Edits for the last month (so, if you apply for permissions at August 12th, this is your sum of edits for July, not for the period July 12th -- August 11):

  • at least 50 edits at Meta during the last month
  • at least 50 edits on at least two projects (per project) during the last month

Mainly the part about Meta and the "last month" standard. All this will do is cause a rush of users request the permission at the end of the month when they have racked up the needed number of edits. Also, I don't think requiring 50 meta edits is a great idea either; I think we need something to show the user is watching meta, but edits are not a great indication of activity level. To solve the first, I say we stick with 50 edits in the last 2 months for Meta, and for the other two projects, the currently system is fine. Thoughts? Monobi (talk) 22:37, 8 June 2008 (UTC)[reply]

I also felt this requirement was not so good but I think this would be a good solution. --MF-W 14:02, 9 June 2008 (UTC)[reply]
Agree with Monobi. If I was looking which requirements I had / I had not, the one about Meta was the only one I had not. Though, I am not really active in vandalism fighting so we should add a requirement to check if you're active cross-wiki, in fighting vandalism. SPQRobin 15:50, 9 June 2008 (UTC)[reply]
I also agree with Monobi. I'm going to change the proposal. OhanaUnitedTalk page 01:49, 14 June 2008 (UTC)[reply]
And Cometstyles reverted it. So, is there anyone who does not want that it will be changed? --MF-W 12:28, 14 June 2008 (UTC)[reply]
I think that edits on Meta are important because they show the meta activity of the contributor. And keeping small projects means that a contributor should be active at Meta. --Millosh
I am asking Cometstyles to come over and comment his revert. OhanaUnitedTalk page 02:28, 15 June 2008 (UTC)[reply]
I am sure it was just a minor oversight, there is definitely strong consensus here. I have undone his reversion and Comets will surely comment here. :-) Cbrown1023 talk 03:13, 15 June 2008 (UTC)[reply]

Generally a bad idea

Asking people to not do something, and then giving them the capability to do it, and then providing them with the easy excuse that they didn't realize what they were doing, is not going to work. Nearly every "admin abuse" complaint comes from the idea that admins CAN do something, but shouldn't. We are not creating a small wiki patrol team, we are creating global admins, regardless of the name you give them, something that the large projects will detest, and the small projects will be overwhelmed by. Don't make this group, please... It just invites admining on projects where people don't know what they are doing, and I am sure this right will be harder to remove than advertised, they always are. Please don't make this group! Prodego talk 04:45, 9 June 2008 (UTC)[reply]

I totally agree with you, when I first heard about this..I was really against it because I knew abuse is imminent, it will be just a matter of time,but slowly realized that this idea will benefit wikimedia more in the long run, though looking from the POV of bigger wikis like enwiki or dewiki, I would generally find this too be mildly absurd :P ...anyways this will benefit the small wikis more than it will to the bigger wikis, though it does create a sort of a cabal, its generally more beneficial to the smaller wikis in terms of vandalism and spambot attacks which generally gets unnoticed by most people...--Cometstyles 06:23, 9 June 2008 (UTC)[reply]
Well, then again, adminship on an existing content project is a requirement for the proposed rights allocation, and potentials would still have to apply via (pseudo-)RfA. However, having a bunch of administrators with what is essentially nothing less than global adminship, and giving them the ability to perform controversial administrative tasks on possibly unfamiliar projects and, in most cases, in completely alien languages, seems crazy. If they have experience in the project and know the language, then why not file a request at SRP like everyone else? haz (talk) 10:27, 9 June 2008 (UTC)[reply]
Stewards don't know all Wikimedian languages and they are doing the same job. --Millosh 10:38, 9 June 2008 (UTC)[reply]
Stewards however, are some of the most respected users, and have been elected by all the projects, not simply a small group on meta, as these global sysops would be. Prodego talk 17:25, 9 June 2008 (UTC)[reply]
I understand Prodego's concerns (and expect that there will certainly be 'global admin abuse'), but also agree with cometsyles and haza-w that the net benifit to smaller projects will be substantial. A couple of non-wikimedial projects I fool about with do occasionally have 'lack of admin' type issues. So, I support the proposal with a clear 'deadmin' procedure. --Rocksanddirt 17:36, 9 June 2008 (UTC)[reply]

Could it not be made such that projects can opt out on a technical level? That would doubtless provide the advantages, without the major disadvatages. I know this is not currently possible, but I don't see why it couldn't be done. Prodego talk 19:56, 9 June 2008 (UTC)[reply]

Over at en:Wikipedia:Global rights usage we are working up a model policy to address some of Prodego's concerns. Ideally I would like a technical feature prohibiting AVFs from adminning wikis that opt out, especially because they will have unlogged permissions such as viewing deleted contribs, etc. But if this policy does pass, I'd like a prominent link to the en.wp policy to remind AVFs who might not be familiar with local customs. MBisanz 01:01, 10 June 2008 (UTC)[reply]
Yes. en.wp policy should be a good example for other projects how to make a policy for AVFs. So, it should go into the policy, as well as at a separate page which lists all existing policies. --Millosh 10:11, 10 June 2008 (UTC)[reply]
Agree, wikis with a custom policy regarding theses users should show them a notice (easy to target with js checking wgUserGroups), preferably with some kind of standarised icon showing what are they allowed to use and what not. Although these would mostly be big wikis where they shouldn't do any action, a mid-size wiki opting out could be problematic. And having to start looking in a foreign langage where do they have the janitor policy (if they have), just to revert an edit, is completely wrong. Platonides 11:15, 11 June 2008 (UTC)[reply]

I thoroughly object to technical "opting out" of global groups. enwiki is as much a part of Wikimedia as other wikis. I discussed with VasilievVV the possibility of defining "sets" of wikis on which specific global groups apply — but it is absurd to allow wikis to disallow global groups altogether. Werdna 08:09, 12 June 2008 (UTC)[reply]

Generally a good idea

I like this idea and heartily support it! Bstone 02:41, 11 June 2008 (UTC)[reply]

Tho perhaps the name "global sysop" is not the best? Can be a bit misleading. Bstone 04:15, 11 June 2008 (UTC)[reply]
I don't like that name, too. However, it is the decision for the name. --Millosh 04:23, 11 June 2008 (UTC)[reply]
"Global sysop" is better than "anti-vandal fighter", though, as that could be interpreted as someone who fights anti-vandals or is against vandal fighters. :) Scott5114 06:31, 11 June 2008 (UTC)[reply]
Yes, better, but I preferred other names, like "nomad" is, for example :) --Millosh 09:34, 11 June 2008 (UTC)[reply]

SUL

Why should a global sysop have a unified account without unattached accounts? Is this the same for all global groups? --Erwin(85) 07:05, 11 June 2008 (UTC)[reply]

Global sysop with username X which at yy.wiki has username Y wouldn't have global sysop rights at yy.wiki. It is, also, applied for stewards, but the main steward's job is at meta (granting and revoking permissions). --Millosh 09:38, 11 June 2008 (UTC)[reply]
To Erwin, since I made that change, If is user is not totally Global meaning not all his account is his, then him having a global right will mean he won't be able to use it on those wikis, and it also helps to prevent abuse by users with that name, so if you are fully unified, you can be judged easily for all your actions...--Cometstyles 10:05, 11 June 2008 (UTC)[reply]

Opt-out and removing rights

There has been some significant concern about this proposal voiced on, at least, the English Wikipedia. I think it would be reasonable to post a bug request asking that developers enable an opt-out for this rights group (and related "levels"). I see no harm in allowing local project opt-out, since the point of the proposal isn't to force admins on projects. The policy of immediately removing rights from users that violate local policies is there - but as others have said, you aren't expecting global sysops to be familiar with all local project admin action policies. How many should they be familiar with? Just the large ones? The large ones, ones with CU, ones with an active community including admins and 'crats even if they are relatively small and no CU?

I'm also concerned that the stewards will decide not to enforce the "immediate" nature of desysopping, particularly in the case of rollback or other minor use in contravention of local policy. Have we had an expression from them of their opinion on this issue, and will this policy (once accepted, if it is) result in a description of this sort of event being added to the steward policy?

I'm prepared to submit the bug request citing the WT:ADMIN page on the English Wikipedia as evidence of consensus that en.wp does not request the services of global sysops/anti-vandal fighters on that project. I'm a little concerned that I don't think objections and problems pointed out on this page have received the right sort of attention - with all due respect to Millosh, his defense of this proposal and enthusiasm in pushing it forward seems to be resulting in less than due regard for local project concerns. Avruch 00:27, 12 June 2008 (UTC)[reply]

I think that the need to immediately remove rights is not that big of an issue. Any local admin can block any Goblal Sysop for violations on the local Wiki. Actually local Wikis can technically opt out. It is not a clean as one would like, but it would work. Simply block all Global Sysops locally and require them to have a separate local account if that wanted to work in that wiki. Something like a bot account. The local wiki would need to maintain a separate log clearly indicating that the blocks were for technical reasons of blocking global rights rather than for any violation of those rights. Dbiel 04:21, 12 June 2008 (UTC)[reply]
As far as I know, blocking someone doesn't remove sysop rights. So this would hinder people with those actions that aren't the problem, but do nothing against those actions that are. - Andre Engels 08:56, 12 June 2008 (UTC)[reply]

Avruch, en.wp adopted the policy WP:GRU related to the global rights yesterday. And, of course, while stewards will not remove permissions if some action was done by mistake, if it was a clear abuse, even for rollback (but, en.wp allowed usage of rollback to global sysops) -- their rights will be removed. BTW, a lot of stewards are participating in building this policy. (As a proposer, I am a steward, too.) --Millosh 05:04, 12 June 2008 (UTC)[reply]

At en.wiki though, some "mistakes" result in emergency desysopping. For instance, an en.wiki admin who disclosed a deleted BLP would be emergency desysopped, even if it was a mistake. Also, afaik, if I block another en.wiki admin, they can unblock themselves and continue to block and unblock other users, while remaining blocked themselves. Would global sysops be unable to unblock themselves at the local level? MBisanz 07:46, 12 June 2008 (UTC)[reply]
OK, it is obvious that mistakes don't have the same weight. BTW, disclosing BLP may not be treated as a mistake. By mistake may be treated a technical actions which make one process (like, accidentally deleting some page). However, if someone made some systematic mistake (deleting a group of pages or so), even there wouldn't be any intention, permissions should be removed because of technical unreliability. --Millosh 08:18, 12 June 2008 (UTC)[reply]
Also, any global sysop will have to follow local rules at the projects which have explicitly defined policies. In the case of en.wp, global sysops will have to follow WP:GRU, as well as all policies which affect permissions which they may use (e.g. rollback policy and similar). --Millosh 08:18, 12 June 2008 (UTC)[reply]
And, at the end, as a steward, I really don't want to see any unreliable global sysop. While I am personally always for giving permissions to any constructive contributor, I am, also, very fast in removing permissions to unreliable permissions holders. --Millosh 08:18, 12 June 2008 (UTC)[reply]
I appreciate that, but I still see no reason not to allow a technical opt-out for opt-out minded wikis. I think if it were available, en.wp would go that route rather than adopt and enforce the global rights policy. Will the intention of stewards to immediately desysop in most instances of misuse of global rights be documented in the steward policy? Have you considered asking that this particular rights group be unable to take actions while blocked? Avruch 13:32, 12 June 2008 (UTC)[reply]
I see opting-out technically as somewhat socially problematic (are we the same community?). However, if there is a significant amount of large projects which are willing to opt-out, it should be implemented. So, the right place for the discussion about opting-out is your local project, firstly. Actually, I don't see any option how to stop, for example, en.wp or de.wp from opting-out technically if the software implementation exists. --Millosh 15:15, 12 June 2008 (UTC)[reply]
About the second issue we were discussed at the #wikimedia channel this morning, and I'll repeat what did I say here: --Millosh 15:15, 12 June 2008 (UTC)[reply]
  1. Removing permissions immediately if someone made a serious abuse of the rights is out of the question: It will be done at the time when the first steward is aware of that. --Millosh 15:15, 12 June 2008 (UTC)[reply]
  2. There is a possible set of "small abuses" or, better, the situation when a couple mistakes would be made by a global sysops. According to en.wp rules, the third minor mistake/abuse may lead (at least) to the rights removing. For example, if someone didn't use rollback button appropriately three times. In such cases, policy on en.wp states that such global sysop will be blocked until their permissions wouldn't be removed. So, we have a couple of possible options: --Millosh 15:15, 12 June 2008 (UTC)[reply]
    1. Stewards remove permissions to the user (steward did it or user asked it). --Millosh 15:15, 12 June 2008 (UTC)[reply]
    2. Stewards don't remove permissions (it is possible that some global sysop would be very useful in maintaining small wikis and that it is really stupid to remove his rights because of small abuse or a couple of minor mistakes): --Millosh 15:15, 12 June 2008 (UTC)[reply]
      1. Global sysop is indefinitely blocked from editing at en.wp (or any local project). --Millosh 15:15, 12 June 2008 (UTC)[reply]
        1. If a global sysop in such case remove their block, this is a clear example of heavy abuse of the permissions, so, the rule one will be used.) --Millosh 15:15, 12 June 2008 (UTC)[reply]
      2. Global sysop makes another account at en.wp and uses it as an ordinary user. --Millosh 15:15, 12 June 2008 (UTC)[reply]
In all of the cases, en.wp (or any other project) is protected very well from abusing the rights. And it is obvious that any kind of heavy permissions abusing is not a game with happy end for the global sysop who wants to do that. --Millosh 15:15, 12 June 2008 (UTC)[reply]

Final proofreading

So, it is the time to start with the final proofreading. We have two and half days before the voting. Please, read the article again and fix bad wordings, spelling errors etc. (I am not a native English speaker, so I can't be so useful for this task.) --Millosh 08:21, 12 June 2008 (UTC)[reply]

The shortcut is "AVF". Should we add another to represent the new name? A minor thing, but just asking :-). — Monobi (talk) 16:39, 12 June 2008 (UTC)[reply]
"You can edit". :-) Of course you can add a new shortcut, just use your best judgement. Cbrown1023 talk 17:30, 12 June 2008 (UTC)[reply]

Access to undelete

The answer given to "what about misuse by using the gloabl sysop access on a large wiki" has been "if people do, they will have it removed".

My sole concern given the safeguards in the proposal, is for the actions that can be undertaken for which no trace exists (eg, global sysops will be able to read all deleted revisions, on all wikis, large and small, and this is not usually logged).

Question - can the log be enhanced (I assume there'll be a log) so one can look up all global sysop actions with optional <action type>, <user>, <wikilanguage> and <wikitype>, thus:

All global sysop <actions of any kind> by <User:Example> on <en> <wiktionary>.
All global sysop <protection changes> by <any user> on <de> <wikipedia>.

I think that would be very reassuring. Without this, there's no way to know if any more "silent" options have been misused, or to check a users' entire usage or check for global sysop actions being done on a "large" project. The normal logs only show delete/undelete, protect/unprotect and block/unblock whereas global sysops will have ability to do more. FT2 (Talk | email) 04:56, 13 June 2008 (UTC)[reply]

I think that would be excessive. Not a lot of people will get Global sysops, so I doubt these issues will come up. Monobi (talk) 05:50, 13 June 2008 (UTC)[reply]
Why would global sysops need undelete in the first place? I thought they were vandal fighters, they should not be judging if something needs to be restored, merely if its blatant vandalism that needs deleting. MBisanz 08:03, 13 June 2008 (UTC)[reply]
For undeleting articles which they deleted (by mistake). I mentioned that a couple of times: When/if the permission "undelete if I deleted" would be made, it would replace undelete permission. (And it stays clearly in the policy.) --Millosh 08:35, 13 June 2008 (UTC)[reply]
About logs: Yes, logs should be seen at one place and developers should be asked for that. --Millosh 08:35, 13 June 2008 (UTC)[reply]
And, again, about "silent" options: English Wikipedia has more than 1500 admins. I would be happy to see 10% of that number as global sysops, but I don't think that we will have such number in the next year or so. This means that it is not a problem related [only] to global sysops, but to all other admins. --Millosh 08:35, 13 June 2008 (UTC)[reply]

propose slightly stricter criteria

I propose increasing the requirements from adminship, crat, or checkuser on one project to two. If we're going to be trusting these users to interact with other languages and other cultures, they should already be used to it. This will also help them be familiar with working with different policies and different attitudes across projects. And users with better credentials will better persuade those that don't like editors from other projects passing judgment where they see fit. This will only strengthen the case that these users are trustworthy enough to act on other projects. --DeadEyeArrow 05:51, 13 June 2008 (UTC)[reply]

This sounds like a very reasonable requirement especially with the large number of project that exist and which they will have access to with sysop rights. Dbiel 06:41, 13 June 2008 (UTC)[reply]
I agree, too. So, let's wait a day to hear what do others think. --Millosh 08:29, 13 June 2008 (UTC)[reply]
Personally I think this is a reasonable approach. I guess I might be fussy & say one of the project should be one of the "big" ones maybe. It would be possible to be a sysop on a couple of minor language non wp projects with half a dozen votes (I realise that someone like that probably would not be granted the rights but if you are going to "tighten" it?). --Herby talk thyme 08:38, 13 June 2008 (UTC)[reply]
I suppose that the preference for adminship would be one big project or Meta + another project. But, I think that it should stay as a preference, not as a requirement. "One big project" would mean that a lot of contributors are not in the same position. --Millosh 08:59, 13 June 2008 (UTC)[reply]
What's a big project? I don't know. Therefore, I agree with Millosh that it should stay a preference. Being an admin on two projects is a good idea though. I expect all global sysops will already be an admin on meta and thus also on another project, but there's no harm in actually making it a criterium. --Erwin(85) 10:12, 13 June 2008 (UTC)[reply]
No adminship on Meta is NOT a criteria and wil never be one, it will just bring a lot of unwanted people on Meta having multiple RfA and most of not all will fail the requirements since we may be forced to strengthen it, permanent multiple-project admin is a good requirement, and by permanent I mean not a temporary one granted by the stewards through Stewards request/permissions..--Cometstyles 12:31, 13 June 2008 (UTC)[reply]
I agree, there shouldn't be a requirement that one of the Wikis be a big or any specific one but I did mean for it to be a content wikis. If we narrow the criteria to a wiki or kind of wiki it could lead to problems. --DeadEyeArrow 06:27, 14 June 2008 (UTC)[reply]

To require adminship on one big wiki would be a bad idea. Example a user has adminship on 10 small wikis but no big wikis and is already doing GS type work on 10 more wikis; to require him to get adminship on a big wiki prior to becoming a GS is wrong. Granting him the GS tools would make his work much easier. Dbiel 14:23, 13 June 2008 (UTC)[reply]

Watching this I'm striking that aspect of my comment - I'm not bothered about it. --Herby talk thyme 14:36, 13 June 2008 (UTC)[reply]
I support two rather than one... I see the reasoning for wanting one of them to be "big" and would be OK with either one on that, but I'd hate to see requiring meta adminship. ++Lar: t/c 02:09, 15 June 2008 (UTC)[reply]

Candidates should be already administrator

Global sysop should be trusted in wikis. I propose this clause.

  • Global sysop candidate should be already administrator(or bureaucrat,CheckUser,Oversighter).

Thank you!--Kwj2772 13:33, 13 June 2008 (UTC)[reply]

ahh actually that clause is already there
Oops. I missed. sorry.--Kwj2772 14:44, 13 June 2008 (UTC)[reply]
Define "content project"... is Commons? Is Species? Meta? I'm thinking of the projects in the *.wikimedia.org, and if they are considered "content" projects. Monobi (talk) 16:12, 13 June 2008 (UTC)[reply]
Does mediawiki.org count as a content project? --MZMcBride 17:54, 13 June 2008 (UTC)[reply]
It should be. You may reword sentences to make it more clear. --Millosh 18:00, 13 June 2008 (UTC)[reply]
Okay, I guess if all these count as content projects, then what aren't content projects? Incubator? ArbCom wiki? Monobi (talk) 20:05, 13 June 2008 (UTC)[reply]
Incubator is definitely a content project, but ArbCom wiki, OTRS wiki, Internalwiki, etc. are not. Cbrown1023 talk 20:09, 13 June 2008 (UTC)[reply]
(un-denting) Okay, so basically a sysop on any public wiki? Monobi (talk) 20:22, 13 June 2008 (UTC)[reply]
I wouldn't say that, no. :-) For example, quality: and advisory are public but they are not "content projects". The wording now is probably the best we can get it. Cbrown1023 talk 21:03, 13 June 2008 (UTC)[reply]
Optionally, we may list "content wikis" at a separate page. --Millosh 21:07, 13 June 2008 (UTC)[reply]
Er, that would be a crazy idea. :-) It would be a lot easier to list non-content projects. Cbrown1023 talk 22:27, 13 June 2008 (UTC)[reply]
any wiki with a ".wikipedia.", ".wiktionary." ".wikibooks." ".wikinews." ".wikiquote." ".wikisource." ".wikiversity." in its domain can be classed as a content wiki + a few of those on the ".wikimedia." domain such as Incubator, wikispecies, and above all commons, I don't believe Mediawiki.org qualifies as a "content project" since it doesn't have a big community and most are developers..--Cometstyles 23:32, 13 June 2008 (UTC)[reply]
If we consider the wikis listed at Template:Sisterprojects (which are the most well known), then the wikis of the first, second and third row can be viewed as content wikis (of whatever language). But I'm not sure that the others qualify. Meta is a coordination wiki, not really a content wiki, but it may be an exception. Cenarium (talk) 01:29, 14 June 2008 (UTC)[reply]
We may rearrange the page Global sysops/Small and large wikis to include more details. There are already enough details there. --Millosh 09:59, 14 June 2008 (UTC)[reply]
Please remember that test.wikipedia does not count. Daniel (talk) 17:04, 14 June 2008 (UTC)[reply]
There he goes again, crushing my dreams ;-) . Monobi (talk) 22:04, 14 June 2008 (UTC)[reply]
I vaguely recall something about a +bureaucrat and Wikispecies... :) Daniel (talk) 01:35, 15 June 2008 (UTC)[reply]

Languages

This question may be out of place, but since this topic is directly related to wikis in all languages, is this discussion being held in the other languages of Meta? or just in English? I do not see many non-english signatures here. (note: I have a very limited knowledge on how Meta works) Dbiel 20:47, 14 June 2008 (UTC)[reply]

Losing privileges immediately?

It seems a bit harsh that they lose their privileges immediately if they use powers on a Wiki with admins. Anyone can make a mistake, especially on something trivial like rollback.--Cato 22:48, 14 June 2008 (UTC)[reply]

I tend to think that the reality of applying the policy of immediately removing privileges will be a bit slower than immediately; but the real point is simply the fact that the incorrect use of these powers will not be allowed and will be dealt with swiftly. Dbiel 00:58, 15 June 2008 (UTC)[reply]
Furthermore, local policy may mean that differing degrees of severity are used relating to blocking the accounts of users who use their tools, from accidental use to malicious use. Daniel (talk) 01:36, 15 June 2008 (UTC)[reply]

What happened to Global rollback?

Was Global Rollback absorbed with the global sysop? Personally I'd like to have it over the entire sysop bit... at the moment deleting, etc w/ global sysop is excessive and would go unused for most users, while being able to rollback instead of undo (which is epicly slow) would be great. Monobi (talk) 03:50, 15 June 2008 (UTC)[reply]