Jump to content

Talk:Stewards policy/2010

From Meta, a Wikimedia project coordination wiki
Latest comment: 13 years ago by Hercule in topic Traduction

Rules proposed during the voting stage

Anthere gave the following very reasonable suggestions in [1]:

I would like to suggest that at least two rules are necessary
rule 1 : no decision alone : depending on the urgency, Jimbo decision, arbitration committee decision, full vote (like for sysoping someone), poll (like urgent desysoping)
I think this rule is very important. No decision taken by an honorary developper alone
rule 2 : report mandatory. Depending on the action taken, on the mailing list, on the wikipedia sysop vote page, on the poll page
Breaking of any of these two rules (without a damn good reason) means suspension of "honorary developer" position.
No power should be entirely in a couple of people hands without rules. And the more power, the more the rules need to be enforced.
Last : when there are questions about the actions of an "honorary developer", the one questioning should be granted free speech and public place to talk.

What a steward can do

User:Tim Starling thinks that stewards should be able to configure the entire power structure of a wiki, on an individual basis.

—Eloquence thinks that stewards should essentially be bureaucrats with global access, that regular bureaucrats should get the additional ability of desysopping, and that the fundamental power structure should be the same for each wiki. See [2] for a detailed explanation.

Assigning rights to yourself

Please see Talk:Stewards#Assigning rights to yourself

Use bureaucrat abilities on your own wiki

I think people who are both bureaucrats and stewards should keep these two responsibilities separate. For example, if Anthere was a bureaucrat on fr:, she should make sysops on fr: using fr:special:makesysop, not test:special:makesysop. This has the advantage of logging the sysopping on fr:, where people will see it, rather than hidden away on test. Angela 22:28, 5 Apr 2004 (UTC)

That sounds like something I would agree with. Also, I would not want other stewards to do this on fr: either. We should try to let local sysops and bureaucrats handle things as much as possible. For Stewards remain then:
  • Actions that Bureaucrats cannot do
  • Actions on Wikipedias without Bureaucrats
  • Actions on Wikipedias where the Bureaucrats are inactive
  • Rare cases where a Bureaucrat refuses to follow up the will of a great majority of his/her Wikipedia
  • Cases like large-scale vandalism repair, where speed is more important than transparency and due process
For the record, I myself am bureaucrat on nl: and fy: and non-bureaucrat sysop on en:. Formally I also am sysop on meta: and a few smaller Wikipedias, but those are historical artifacts (meta because of the early attempt to have system-wide bureaucrats, the smaller wikis because of the March 10 spambot attack. - Andre Engels 10:09, 7 Apr 2004 (UTC)


I made a few changes to this page, adding "/she" to a place where the word "he" referred to "steward" (there are also female stewards). Under Stewards policy/2010#Checks, I changed Special:Listusers to Special:Listusers/sysop, which is a new page listing the sysops only. This change made the sentence "Stewards will mostly be dealing with the small Wikipedias, so this will not take long to check." unnecessary, so I removed it. The last change was changing "test" to "meta". That seems to have been an error, or is "test" an early name of Meta?

Anyways, I'm not sure if writing this is even needed (not sure of the policies about that here), I just felt it was better to be sure. Jon Harald Søby (talk, contrib) 18:32, 22 July 2005 (UTC)

Hi, thanks for the update. Test is not an earlier name of meta. Steward interface was initially only on test. It was moved to meta, as test log was frequently cleared and for transparency reasons. Anthere

Ah, then I see. =) Jon Harald Søby (talk, contrib) 09:49, 23 July 2005 (UTC)

Abuse of steward-rights

i propose to add the following to the steward policies:
Any abuse, or even the threatening of abuse, of the steward-rights, may lead to an immediate removal of these.

I think that's pretty obvious and doesn't need to be stated; the same could be said for any special access. —{admin} Pathoschild 06:47, 26 December 2006 (UTC)
Je pense qu'il faut aussi définir la notion d'abus et ne pas la confondre avec une erreur d'appréciation.--Bertrand GRONDIN – Talk 11:34, 26 December 2006 (UTC) Bertrand Grondin said: "I think that the notion of abuse needs to be defined, not to be confused with an appreciation error." Yann 12:00, 26 December 2006 (UTC)
Something that could go a long way in increasing transparency of Steward actions would be the inclusion of a "Reason" field in the Steward Makesysop tool. That would provide an explanation as to why a Steward adjusted their own account's rights on any project, or someone else's, and help people understand better what it is that the Steward might have been doing. Redux 12:26, 26 December 2006 (UTC)
I have no objections in principle. Adding a "Reason" summary field is a great idea. --Dbl2010 06:06, 27 December 2006 (UTC)
Yes, this seems to be a good idea. --.anaconda 01:55, 28 December 2006 (UTC)
This would seem to suffice as the entirety of the steward policy. The rest of the material should be sent to some sort of how-to page, with a list of steward duties and suggestions on how to perform them properly. — Dan | talk 06:38, 27 December 2006 (UTC)


I've sandboxed an overhauled version of the steward policies (diff). Aside from various grammatical and spelling changes, I've updated all sections to reflect current usage; most notably, I moved the focus away from the English Wikipedia and adminship, since there are now many other projects and user rights. I split suggestions and guidelines into their own section, to distinguish them from policy; for example, a discussion from 2004 about new users (which I summarized) and an explanation of how requests are made are moved out of the section which includes rules about transparency and the use of steward tools on a home project. I generalized phrases that were too specific, such as "[local] bureaucrat is responsible for creating new sysops" to "local bureaucrats are responsible for promoting users" (bureaucrats can also grant bureaucrat and bot status, not just administrator).

There were many other minor changes, but I think I summarised all the important changes above. You can see for yourself by looking at the differential link. What do you think? Is there anything else you think should be changed or corrected? —{admin} Pathoschild 09:17, 26 December 2006 (UTC)

  • Hi, I find this proposal quite good. Yann 12:12, 26 December 2006 (UTC)
  • I second that. BTW, what is "your own project"? For example, though I'm mostly active on ru:, where I'm a sysop, I also have over 2K edits on en:. So, can I desysop someone on en:? MaxSem 17:10, 26 December 2006 (UTC)
    Perhaps "a project you edit" would be better? —{admin} Pathoschild 00:45, 27 December 2006 (UTC)
    I believe we could stick with the expresion "home wiki/project", and if necessary, work on a more precise definition for it. As I understand it, a "home project" is where a user edits substantialy. Some users could be considered to have more than one "home project", some will have just one. And noticing that having an edit count that isn't negligible on a project is not necessarily sufficient to make that project a "home project". Redux 03:44, 27 December 2006 (UTC)
    Or "a project in which you judge yourself to have some vested interest". Still, this has always struck me as irrelevant; the decision to desysop a user will always be the user's own or that of a person or body representative of that community (e.g. an arbitration committee, Jimbo, etc.), and not a steward's. As long as stewards are executing and not deciding, they should be able to do so on their own wiki(s). — Dan | talk 06:17, 27 December 2006 (UTC)
    Not sure I agree with that... there is merit in avoiding the conflict of interest, or even the appearance of same, so I support "do not do things where you might give such appearance" as a guideline. I do however think that most of this boils down to "be reasonable", though... ++Lar: t/c 01:13, 28 December 2006 (UTC)
    Also, I don't think we need to "work on a more precise definition" for much of anything. This is a simple position, unburdened by rules and controversy; let it remain so. One likes to believe that the election process yields truly trustworthy users, and if in fact it does, we can assume that these users' judgement is sound, even when not guided by pedantically precise regulations. — Dan | talk 06:29, 27 December 2006 (UTC)
  • Looks good. --Dbl2010 05:59, 27 December 2006 (UTC)
  • Looks great, Pathsochild! Thunderhead 06:46, 27 December 2006 (UTC)

Implemented. I've merged in the new sections about elections and loss of steward access in a new 'Processes' header. —{admin} Pathoschild 00:12, 11 January 2007 (UTC)

Steward tools on home wikis

When there is some vandalism in es: that includes personal information, as there is no oversighters there and such requests aren't usually being made public, people would drop me a note. So am I not allowed to use commons sense and kill the revision after checking it even though hiding revisions specifically allows that case for oversight? I think RDsmith put it well, stewards are supposed to be trustworthy, sound and responsible users. I understand the appearance of conflict of interest for deadminning somebody or granting checkuser, and thus it's best not to act locally in such cases. But how about more janitorial stuff? drini 17:33, 10 January 2007 (UTC)

This seems eminently reasonable to me. While I am a strong supporter of avoiding even the appearance of impropriety, janitorial tasks like this don't seem an issue. ++Lar: t/c 16:36, 11 January 2007 (UTC)

Changing rights in own wiki?

There is a paragraph like this:

Stewards should not change rights on their own project with their steward tools, unless the subject requested it themselves. It is better to leave such cases to neutral stewards. This also applies to members of local arbitration committees.

How to define "own project"? I find this hard to define. Because I may register myself in all the wikis, but only active in a few of them, while only creating my own userpage in the others. What does it mean by "the subject requested it themselves"? Is it requesting to the particular steward? --Edmund the King of the Woods! 23:51, 29 November 2007 (UTC)

Your "own project" is defined as your home wiki, the project that you most actively edit on, it is self-defined by the steward. "the subject requested it themselves" means that you should only change a user's rights if that user has requested the change themselves (like a desysop request). (If the request pertains to your home wiki.) The reasoning behind this is so that the decision is left to more neutral stewards. Cbrown1023 talk 00:16, 30 November 2007 (UTC)
I've attempted to clarify this slightly and have reworded the section to reflect the fact that there may be more that one project where using steward rights might not be appropriate. Stewards are of course good standing members of the community who are trusted to use their own judgement to determine which projects exactly they feel this applies to. Adambro 00:32, 30 November 2007 (UTC)

But what if in my home wiki, when there was an election that have decided to grant someone a certain status, or someone have to be desysopped or whatever, may I perform the action and give the link to the relevant page? --King Edmund of the Woods 10:31, 2 December 2007 (UTC)

No, you shouldn't. A non-involved steward ought to do that, though you can help by translating the details for them. Angela 10:58, 2 December 2007 (UTC)

What if there is a private info need to be deleted by an oversight, and there is no oversight, may I turn myself into an oversight and perform? --King Edmund of the Woods 11:06, 2 December 2007 (UTC)

No, you should typically never use steward access on a home wiki. You might make an exception for self-requested removal of access, but even that might be better left to other stewards. —{admin} Pathoschild 18:26:24, 02 December 2007 (UTC)
Indeed - for the sake of the neutrality that Stewards need for their role to be fulfilled successfully, one really should not grant themselevs access at all on home wiki. May I ask this though - if one were to run an RFA on a wiki, and clearly passed (like, say, 100% support at end of 7 days): could one grant oneself adminship? --Anonymous DissidentTalk 18:52, 2 December 2007 (UTC)
Obviously, doing the above would be done at the risk of seeming self-indulgent ;) --Anonymous DissidentTalk 18:53, 2 December 2007 (UTC)
No, a Steward should never, ever grant themselves adminship after an RfA. -- 19:08, 2 December 2007 (UTC)
But granting checkuser or oversight access to oneself when there are no local checkusers and oversighters seems reasonable, since the steward would understand the language and the context of the needed actions. --filip 22:18, 15 December 2007 (UTC)

can stewards grant themselves status like adminship anywhere? The preceding unsigned comment was added by (talk • contribs) 22:12, 15 December 2007 (UTC)

The software gives them the power to do so, so they "can". However, policy is clear that there are certain wikis where a given steward should not... whatever wikis are their home wikis, plus whatever wikis have local processes for determining adminship, plus whatever wikis where even the appearance a conflict of interest might arise by doing so. Stewards can and do give themselves adminship on wikis all the time, but in the vast majority of cases it is for wikis where there are no local admins at all, or no active ones, and there is a clear and present pressing need, and as soon as that need is satisfied they turn their access off again. There are a lot of stewards, (although more are needed, why we're having an election now) so there should always be a steward that can act without conflict of interest on any given wiki. Hope that helps. ++Lar: t/c 04:47, 16 December 2007 (UTC)
As a Wikimedia Commons bureaucrat and a new steward, I may have the need to review the edit history of images transwikied from other Wiki sites. Sometimes the original image edit history has been deleted, but the Commons images have not fully credited the original sources or uploaders. Is it permitted for a Commons admin who is also a steward to grant himself or herself local temporary adminship to review the deleted original image edit history and then strip his or her local temporary adminship once finishing the review? For example:
I see a Commons image without fully crediting the original source or uploader. It is claimed to be copied from French Wikipedia where I am not an admin while I can speak French only in the basic level. I try to find it on French Wikipedia, but the original edit history has been deleted. If I cannot even grant myself temporary adminship for the sole purpose to review the deleted original image edit history on French Wikpedia for a few minutes, it can impede Commons administration as I think. This is why I have a question about the permissibility of temporary self-promotion to facilitate cross-Wiki administration. For my question, I would like to define being temporary as very few minutes.--Jusjih 02:29, 23 December 2007 (UTC) (Wikimedia Commons bureaucrat and a new steward)
Temporarily granting yourself adminship is, in my experience, generally okay. On an active wiki, you should probably only do it if you do not intend to perform any actions on that wiki (so in your example, I think it would be fine) -- except in emergencies, deletions and blockings and whatnot should be left to the local administrators, and you risk irritating them or "stepping on toes" by doing their work yourself. On a wiki without a substantial community, granting yourself temporary adminship is almost always okay. Also, if you forget to remove your adminship after you have finished your tasks, it is almost guaranteed that someone will become upset about it sooner or later. — Dan | talk 02:01, 25 December 2007 (UTC)
Thanks for your comment. Leaving no trace of privileged actions to local Wikis when temporarily promoting and then demoting oneself will be the best. Recently when I had my good reasons to temporarily promote and then demote myself, I explained them in the user rights. However, I have seen many stewards not leaving any words of reasons for temporary self-promotions. In case of lost Internet connection this can cause similar problems as you have said about forgetting to demote oneself. Someone, especially if unfamiliar with Meta, from that Wiki may then complain why any users became promoted there without local approval. Your scenario of "stepping on toes" include an argument at Talk:Stewards#Assigning_rights_to_yourself involving Angela and Daniel Mayer.--Jusjih 23:04, 25 December 2007 (UTC)

I can see why a steward might need to temporarily sysop themselves to check image licensing on a project without many local admins, but in other cases a local request will suffice. I agree with Majorly [3] and Angela [4] that Stewards should not need to sysop themselves on large projects like de.wiki or fr.wiki (or en.wiki for that matter). Not only are there plenty of admins to handle vandalism or, in this case, respond to queries about deleted images, there are a good number of users from that project that serve as Commons admins in any event. Much as I think Jusjih does great work on Commons and I can see how he could use steward tools to benefit this, I don't think present steward policy can justify so liberal an approach to stewards assigning themselves admin rights on large projects. WjBscribe 03:09, 9 February 2008 (UTC)

Please come to Requesting help from local Wiki sysops or bureaucrats with language barriers to discuss how to improve communication. Otherwise, language barriers are very frustrating.--Jusjih 02:02, 1 April 2008 (UTC)
Commons-l recently received a thought that Commons admins should be able to view deleted images Wikimedia wide. Supporting the great idea, I consider that a very limited privilege should be granted to Commons admins so they may view deleted images Wikimedia wide, without the privilege to undelete them without local adminship. This will not "step on the toes" of local admins, as viewing deleted images Wikimedia wide without undeleting will not leave local trace. No matter how active a local Wiki is, having to ask local admins to reply is unnecessary bureaucracy and delay, especially if language barriers are involved. After all, I am talking about efficiency without "stepping on the toes" of local admins, so please consider it carefully.--Jusjih 01:24, 4 June 2008 (UTC)

Stewards transferring admin access between 2 accounts

I want to clarify the approach that stewards should take where someone asks for their current admin rights on a project to be transfered to another account which they also control on the wiki in question. At the moment it seems that such requests are treated as uncontroversial and performed once it is confirmed the user making the request controls both accounts. It seems to me that the relevant part of the steward policy is:

Local bureaucrats are responsible for granting sysop, bureaucrat, and bot rights. Stewards should only grant these rights on a project if there is no active bureaucrat available on that project, or if the situation requires immediate action.

Where someone wishes their admin access transfered to a second account which they control, I think it should be correct practice to have a steward remove the access and to ask a local bureaucrat to give it to the other account. Although this may seem a little bureaucratic, this is not a situation that "requires immediate action" and there may be local considerations that could make the rights transfer controversial. For example the second account may have been undisclosed and have failed an RfA, or the community concerned may wish to require various conditions to ensure transparency for the rights transfer. It also ensures the maximum amount of information is available in the local logs. Thoughts? WjBscribe 01:42, 23 March 2008 (UTC)

I'm not a steward but I like this idea. John Reaves (talk) 03:42, 24 March 2008 (UTC)
It seems to me bureaucratic for a reason. It is imho a local community issue whether it is ok to transfer your "rights" to another account. It leaves of course major mazes for fraude etc, so I personally am very much against this practice, and would not consider transferring "rights". A user can make two seperate requests: removal of "rights" on RfP, and the setting of admin status with a local sysop, who is better able to judge whether it is allowed locally, and whether it is really the same person. Effeietsanders 11:39, 24 March 2008 (UTC)
(belated) tack on a "or if the local project has policies allowing for stewards to perform such transfers already" perhaps. Kylu 00:35, 10 November 2008 (UTC)


The sentence that currently reads

However, since stewardship is typically a position likely to get into trouble and since the steward group can easily control itself, the confirmation itself will be done by other stewards.

really ought to be reworded

However, since stewardship is typically a position likely perform controversial actions, and thus be viewed unfavorably by the subjects of those actions, and since the steward group can easily control itself, the confirmation itself will be done by other stewards.

Stewards don't cause trouble, they implement decisions that are controversial because of rights removal generally, so the people they de-userright are of course mad at them, this makes it a bit clearer. MBisanz talk 23:02, 30 August 2008 (UTC)

I wouldn't say they cause trouble or take controversial actions. (Don't they act based on consensus? I heard somewhere they're not even allowed to make decisions.) You would be more correct to say they may take unpopular actions.  — Mike.lifeguard | @en.wb 01:40, 10 November 2008 (UTC)
Perhaps emphasize the ... robotic ... nature of stewardship: While the actions themselves may well be unpopular, the stewards hold the position and perform these actions on the behalf of others. (I liken the idea to that of IP-Relay... an IP-Relay operator is required to merely translate text to speech, and are prohibited from exercising any decision-making at all during the call.) Kylu 02:27, 10 November 2008 (UTC)

Clarifying what is or isn't a Conflict of Interest

I try to use common sense about COIs but recently a situation arose where perhaps not everyone would have come to the same conclusion... so perhaps some clarification is in order. The scenario was that a duly constituted representative of en:wp ArbCom turned up on the Requests page asking that certain new arbitrators (already identified to the foundation and correctly listed on the Identification Noticeboard) be made CUs and/or Oversighters on en:wp (some were CUs already). A bit later I saw these, and since no other steward had yet acted, and since "common sense" told me that this was entirely noncontroversial, I went ahead and turned on the bits. A longtime and well respected contributor here suggested that perhaps I had a COI and should not have done it, since en:wp is a home wiki and since I voted in the elections. So I let the next one (which appeared a bit later) slide, but it was a good 16-18 hours before they got their bits turned on. An en:w pArbCom clerk later commented that they saw no issue with my doing the permissions changes, and that doing them right away was of such benefit that it outweighed any apparent COI issues. Comments? (I'm not trying to reopen any underlying controversy among the parties, which has been settled entirely amicably... just looking for comments on whether there is reason to tweak the wording here...) If we did tweak it, what should it say? ++Lar: t/c 14:24, 26 December 2008 (UTC)

IMHO in this special case as described I personally think it was even better that You did it because the request Steward_requests/Permissions/2008-12#New_ArbCom.40enwiki was so rare with information that I would have had to investigate where the voting was etc. and if the people are rightfully requesting this, but You already knew because You know who is clerk, where the voting was, etc.. Best regards, --birdy geimfyglið (:> )=| 15:32, 26 December 2008 (UTC)
I think conflicts are more likely when removing access, not granting it. I believe the rule is there to prevent a steward desysopping someone they're in an argument with, not to prevent all rights changes on your home wike. It's quite common for people to deal with these on their own wiki in non-controversial cases, like this one, and in situations like an admin requesting their own rights be removed. Angela 02:06, 30 December 2008 (UTC)
I'd like to suggest the policy be modified to state that Foundation-recognized ArbComs requests be considered non-controversial. The most conflict of interest I can see is if the steward refuses to perform the change... in which case it'd be filled by a different steward. Kylu 05:27, 4 February 2009 (UTC)

my question

(In my imagination) If the court of some country issued a warrant for investigation (maybe if mass personal informations such Resident registration numbers were released on wiki), Is steward allowed to perform checkuser or oversight action even in home wiki ignoring steward policy? (cf. in S Korea, when mass personal informations were released, millions of people raised legal suits and the responsible administrators were punished)--Kwj2772 07:23, 4 February 2009 (UTC)

No, public release of private information is a violation of the privacy policy. Court requests for information should go through the Wikimedia Foundation, not individual employees or volunteers. —Pathoschild 02:11:12, 05 February 2009 (UTC)

en-wiki update

En-wiki has updated its global rights policy with regard to when stewards may take local actions at [5], any feedback is welcome at w:WT:GRP. MBisanz talk 01:28, 10 March 2009 (UTC)

Thanks for the clarification. Kylu 02:14, 10 March 2009 (UTC)


In the last two years, activity level has been one of the most popular topics of discussion during steward elections and confirmations. The current activity policy was developed in 2009 to address this, by requiring a certain level of activity of all stewards for the first time. There have been suggestions that this bar is too low, or too difficult to assess over short periods of time, and gaps between expectations and policy lead to discomfort.

Proposed changes:

Current policy
Any steward inactive (as a steward) on meta for a full year will have their steward permissions removed. "Inactive" means no steward action in the past 6 months and fewer than 10 steward actions in the last year. They may re-apply through the regular process.
Proposed policy
Any steward inactive (as a steward) on meta for more than six months may have their steward permissions removed. Stewards who are on an extended wiki-break should say so publicly. To encourage the retention and development of current stewards, those found to be inactive may be invited to work on existing backlogs, and their activity revisited after three months. Stewards who lose their flags through inactivity may re-apply through the regular process.
"Inactive" means no steward actions in the past 3 months and fewer than 10 steward actions in the last six months.
A "steward action" can be a global rights change, use of steward global group permissions, resolution of some other steward request (even when this does not require the use of steward tools), or filing a steward-related bug report.

This would support assessments avery 3 months; a script could automatically notify people becoming inactive with talk-page messages and links to recent backlogs.

Open questions
  • Stewards could be removed under this policy for 6 months of no activity or fewer than 20 responses a year. Is this appropriate? Consider that many current steward tasks can now be handled by Global Sysops.
    • Are cross-wiki deletion requests still a steward task?
    • Are there other common steward requests (bot flags and bad-name hiding come to mind) that could be better split up or shared?
  • Can we script tracking of bug reports? What other tasks develop a backlog when stewards don't act on them?
  • Should there be accommodation for stewards on extended wiki-breaks?
  • Do we want to distinguish 'active' stewards with technical bits from others involved in discussions of cross-wiki policy and requests? (consider Anthere, still very interested in the latter, and birdy, still very active)

SJ+ help translate 05:29, 9 March 2010 (UTC)

<-= I'm glad sj started this I was coming to do so and should have a while ago. I die a little every time I see someone say someone must be confirmed because he hasn't proved mistrust and he has done 10 actions in the past year. I'm sorry, but fewer then 10 actions over a whole year (and to meet the current requirement you actually only have to have done one in the past 6 months) is laughably small. We say it isn't for life and then we have a policy which at least some people basically read to be "unless you've totally screwed up you're good to go" regardless of whether or not you are using the tools at all.

In my mind when you have access to such powerful tools as stewards do (and they have the most "powerful" ones available) you should be at least reasonably active with them in order to keep them and to be honest I can't think of any reasonable reason to consider someone who meets the current activity policies active or for that matter people who have 4-5x the activity of the current policy. That of course doesn't mean that they weren't good stewards or that they couldn't BE good stewards but I do think it shows that they no longer need the tools and it would probably be better if they served the community in other fashions. I may be alone in this but to be totally honest I don't think we need people hanging around with full access to the interface if they are not "actively" using it. Actively is of course somewhat up to interpretation but I don't think our current inactivity guidelines come even close. Which brings me to the current guidelines..... James (T C) 10:35, 9 March 2010 (UTC)

Personally, "steward activity" can now also mean the use of the "steward global group"-permissions, that aren't logged on meta (stewards hold all the permissions of sysop, 'crat (except userrights), import, and some of the oversight-rights globally), so if we should do something about the inactivity-rules of the policy, they should include the use of the global rights on all wikis. Also James: there is one group of users that has more access than stewards: sysadmins Laaknor 10:41, 9 March 2010 (UTC)
Language coverage has long been tied up with steward duties. We can separate them with some effort, by identifying other teams that could work closely with stewards, but for me language diversity and the ability to read and understand local policies remains as important as activity... moreso, because effective communication with local editors is necessary before they can understand how and when they can request global/steward help. SJ+ 11:04, 9 March 2010 (UTC)
Yes I realize that and I would say the same thing about them ( I wouldn't want someone laying around with root access not using it for a year either) but not my place ;) I can understand that sentiment and in many ways I have issues (see my ideas) with the "numbers" game at all but I would have to say that if someone is basically just a translator and negotiator with no use of tools then they don't need those tools. People like that are of immense value to the foundation and all of the projects (quite arguably much more value then normal stewards) but I do not see a reason to give out full interface rights for someone to do a job that is technically not even part of the Steward job title. Stewards are supposed to be implementors and in someways "xwiki protectors" , and thats basically it. I know Stewards have traditionally done much more (and I use them as such, and don't honestly have a problem with them being it I think it is inevitable) but I am not sure if this is really the right hat for people who do that instead of the more "technical" implementation jobs. James (T C) 11:08, 9 March 2010 (UTC)
Part of the conversation at the time the policy was being written was, that inactivity is something that can be measured automatically; and that removal from inactivity could be done regularly, without drama, and separately from confirmation discussions. I interpreted "no longer needed" to mean stewards who feel they are not going to be available in the future.
There is a tendency on many of our Projects to make capabilities or flags privileges which must be earned and maintained, or badges of honor to wear — rather than tools theoretically available to all to improve our collaborative work, whose use is limited only by trust and ability. Making capabilities competitive, transient privileges provides a short-term activity boost. But it also narrows the field of contributors who can participate, and supports a heirarchy and high barriers to entry. This is at odds with the traditional wiki philosophy, and I'm not sure that it builds a strong community over time. SJ+ 20:18, 9 March 2010 (UTC)
That's funny I seem to feel that tendency is to do the opposite (sysop privileges etc on en where people don't lose them unless they have major problems though sysop is very different then steward of couse) and the more powerful the tools the more I tend to disagree with that attitude. I don't think I feel that people need to be "horribly" active and in many ways would be fine including a lot of numbers that haven't been included in activity levels (edits on meta, even emails to stewards-l? not sure everything that could/should be included but it doesn't just need to be meta logs). I disagree however that people should just be able to hold these very powerful tools if they aren't using them at all and I would say the thing that goes against the wiki philosophy the most is the idea that you should ignore peoples inactivity concerns in a reconfirmation because they meet some random policy a small number of us hashed out on Meta. In my mind any steward who "votes" to reconfirm based on a policy like that when consensus exists to remove because the voters have another idea of what kind of activity is desired is ignoring their job as stewards. Obviously they can comment using their own ideas int he confirmation itself but once it closes their job is to decide consensus, not to decide inactivity criteria. The larger group discussing confirmation should trump the small group of us debating policy here on meta, hence why the inactivity policy should have nothing to do with the confirmations. James (T C) 22:23, 9 March 2010 (UTC)

To be totally honest when I first read the policy on steward removal (and multiple times since then) I didn't think it was "that" bad. I've always thought the inactivity threshold was shockingly low but I didn't think it mattered because of the 2nd part. For reference:

Loss of steward access

Any steward inactive (as a steward) on meta for a full year will have their steward permissions removed. "Inactive" means no steward action in the past 6 months and fewer than 10 steward actions in the last year. They may re-apply through the regular process.

Poll after a year

Stewardship is not a lifetime status. Users get it if they need it, keep it if people trust them, and lose it if they do not need it or are no longer trusted. Steward status is granted until the next yearly elections, where users will be invited to comment and in particular to ask for removal of status. Should the stewards determine that consensus exists for a steward's rights be removed, the steward will lose their status.

I've noticed that at least a couple people (and at least one steward that I know of and possibly a 2nd that I haven't talked to about it) seems to read this very differently then I did. Since day one, and every time I've looked back at it, I've seen this as two separate ways to lose stewardship. I saw the inactivity section as written to be the "automatic" removal part. This would be something that is done at any point in the year:

say a steward is elected in January, does "stuff" for a while and then starting in about September basically goes on wiki break and edits only infrequently (say does 8-9 changes on random days over a couple months). Perhaps he gets through the poll in January because he had said he was on break and had put up a statement and promoted a couple people the week or so before. He then basically disappears and come next September he hasn't done anything on meta since February and he is just removed per this policy.

This seemed to make sense to me, the numbers were very small but it seemed to make sense to have this written as a way to get rid of inactive stewards outside of the "normal" confirmations. The real confirmation was the "Poll after a year" section where Stewards would lose their tools if they do not need it or are no longer trusted with inactivity with the tools being almost the only reason I could think of for the use of the "no longer needed" phrase. The standing stewards would then decide if there was a consensus to remove based on need (inactivity) or trust. In my mind it strongly appears many stewards are saying (on the consensus discussion pages) that inactivity is basically an argument that has to be ignored and if thats the only argument they should be confirmed because "they are active per policy". In my mind that is an obvious misunderstanding of policy and to be honest an obvious misunderstanding of their job in deciding the confirmations since they should be determining if consensus exists saying they are too inactive, the inactivity policy is a different animal for a different time (NOT the yearly reviews). I may of course be wrong and in many ways I feel like a good portion of the removal policy is written well, but I may just be thinking that because I am interpreting it my way. Either way it is obvious that we need to at the very least qualify if not rework that policy. James (T C) 11:08, 9 March 2010 (UTC)

So in the end I'm somewhat conflicted. I honestly don't really like the "numbers" game. I agree that in many cases the idea of whether someone is active or not can be hard to determine by sheer numbers. That of course is two fold though, you should not be able to point to an activity policy and say "see!!! he has 20 edits and the policy says you can't have less then 20 so anyone who says he is inactive is just wrong move on to the next oppose reasons" just like you shouldn't be able to go and say "hey that guy's only done 40-50 edits and almost none recently so he should removed even though is an invaluable resource on the mailing list and is on IRC debating what to do for an oversight request with other Stewards all the time." That person who contributes his knowledge frequently and could well just not be doing the actual changes all the time because Pathoschild beats him too it with his bot or something.

In many ways I think the policy should be fairly flexible, a "moderately" small "auto-inactive" switch for people who are just not around and the confirmation process where people can air out all the reasons whether the steward should be kept or sent away with a nice chocolate package and a hearty thank you. I also think it should be quite clear that never the two shall meet, the inactivity "auto-switch" is in no way saying that someone can't make a legitimate and important argument for inactivity at the confirmation for someone who meets the switch (by almost any amount, to be honest someone who does a large amount of rights changes but never ever talks among his peers may be almost as bad to me, but thats a different argument). This of course is almost exactly what I feel the current policy says the only thing I really would like to change is to tweak the "auto-switch" and of course clarify were people obviously disagree with my interpretation ;).

While I obviously think 10 actions in a year (and 0 in the past 6 months) is crazy small I don't think the numbers need to be big. Part of me wants to go "huge" and say you need 50 or 60 over the year or something but I can be brought down :) I'm not totally sure whether we want to to have a smaller time segment to compare or use a whole year. If we were doing a year I would say we should do something like you need to do more then 30 actions over the past year with no more then half in the first 6 months of that year (if you've done 30 actions in the last 6 months then your golden. I may even be ok with 20/25 actions if it was set up in that fashion. If we were using something smaller I think we could break it up into 15 every 6 months (with at least half of that in the most recent 3 months? or something like that). I also think it's fine to put some kind of wikibreak feature in there but I don't honestly think it should be long. Maybe you can pause the "inactivity" timer if your on wiki break for up to 1-3 months? To be honest I think if your going to be gone for more then 6months to a year you should probably go through the election cycle again if for no other reason then you will have likely missed quite a bit of goings on. James (T C) 11:38, 9 March 2010 (UTC)

  • Question Being that much of what were formerly "steward-only" actions are now going to be handled by global sysops, that leaves a smaller pool of "actions" for stewards to do (CU's, OS's, permissions, and Global Block/Locks). How would that affect the future required activity levels being that the past is no longer fully credible? -- Avi 17:06, 9 March 2010 (UTC)
    A good point. Global blocks/locks are the fast, numerous part of that list (you can spend an hour deflecting a vandal and lock dozens of accounts). It's great that we now have global sysops... one problem with activity metrics is that we should be looking for ways to let non-stewards do the parts of our work that aren't sensitive, helping communities become self-sustaining, rather than racking up activity points. SJ+ 20:18, 9 March 2010 (UTC)
    Well nothing that has normally been counted in "steward actions" (logged at meta) are able to be done by global sysops since they can not do global blocks or locks. If we want to explicitly make stewards jobs less technical then perhaps we should separate out the jobs for different groups. I have no problems including things like xwiki activity (outside of stewards homewiki) as part of activity but again I would say in general that is the discussion for the confirmations. In many ways I wouldn't have a problem keeping the policy exactly as written I just feel the idea that idea that someone can only be removed for inactivity under that policy as ridiculous and that should be a perfectly acceptable reason for removal during confirmations. As the steward policy so nicely says stewardship is not for life if the community decides you no longer need the tools then you no longer need the tools and can better serve the community it other ways, you don't need the steward flag. James (T C) 22:33, 9 March 2010 (UTC)
    Yes, but define "need". I "need" forty million dollars :). Need, in my understandning, is demonstrated through volunteer work as a steward. This has a floor of 10 actions per annum with 1 in the last six months. Perhaps the floor needs to be changed. While I too like fluidity, there does need to be some concrete measure against which people can be judged fairly, IMO. -- Avi 22:53, 9 March 2010 (UTC)
    I would say the definition of "need" is up to the community. If the community decides you no longer "need" them that should be that regardless of whether they meet some floor or not. I have no problem with having a floor but I do have a problem with saying just because they meet that floor no one is allowed to say they are inactive and don't need the tools, that isn't for you to decide that is for the community to decide as a whole in my mind. If we want to just have concrete measures and other then that you are elected for life (except of course for major issues that arise) then I guess we can do that but I don't personally think we should. James (T C) 23:58, 9 March 2010 (UTC)
More questions - and a proposal: what exactly do we want this inactivity-policy to do, to regulate, to accomplish? set a validation number to opposes for reconfirmations, after which the word "inactive" will not be counted unless it corresponds to the number defined by this policy? should it determine when stewards should be forcibly removed, even in the middle of a year, or should it more like or also set a guideline for desired steward activity? in how much detail do we, in other words, want or need to measure, to count, administer even, all logged actions, also hidden logs, and perhaps, i wonder, steward-task related edits and reverts: should these be taken into account as well, and maybe even expert assistance on irc, activity on the mailing lists as such etc. exactly measuring stewards' presence and activity in a broader sense of duty might be a task we'd need a kind of pepys for? then again, do we perhaps simplify all this and just set an absolute minimum of logged actions only? or somewhere in between these extremes? do we just need this policy for reconfirmations of also for evaluation in general, should it be used cumulatively or per a certain date?
  • in view of this complexity, i would propose the following FORMAT:
each steward should be able to prove at least x different "cases" to have worked on (either by edits, logs, mails, or whatever), measured per (calender) year. activity is measured by "cases" not by actions.
for me x could be anywhere between 24 and 60, but once fixed, this is it, and can be submitted upon request to prove one's activity. it's another approach but imho this raises interesting new vistas and possibilities imho, and goes away from those edit- and log-counts which show and represent an all too one-sided view on the task. oscar 23:43, 9 March 2010 (UTC)
  • I actually really like the idea of cases for activity, it would be more work for the stewards I think in general though to be honest if they put the time to get those cases together every year it is highly unlikely I would personally say they were inactive enough to be removed. I do however still think that we shouldn't be limiting the confirmations. Regardless of whether they meet the inactivity guidelines I'd still argue that people should be able to use their own idea of activity in that arena, just like people can use their own idea of activity or qualification in the Steward elections (no one including myself would argue that the people who opposed me for steward because I was too green couldn't be counted because policy said that you were eligible to run if you had been around 3 months). James (T C) 00:03, 10 March 2010 (UTC)

Allow me to posit something, which is a danger with numerical requirements. In order for a steward to carry out a rights action, there has to be a need for such an action. Further, the action has to be not yet done at the time that the steward goes to carry it out. (that is, that no other steward already did it) With our current volume of work, at 10 actions each, there is no danger of there not being enough actions. But suppose for the sake of argument we raised the requirement to 1000 actions? It is just possible that there then would not be enough actions to go around. Even at 100 actions, it's possible that not all stewards would have enough actions available, especially if some stewards made it a habit of doing actions as fast as they could. This is a good problem to have, to be sure, and I am in no way advocating that we should not have minimum activity requirements! But it's something to keep in mind here... strict numerical limits can be taken too far. ++Lar: t/c 14:52, 10 March 2010 (UTC)

Personally I think all rights should be removed anywhere if they are not used. They are granted to benefit the community, if they ceased to be used they ceased to benefit the community. It would seem then there only purpose is to allow decoration of a trophy cabinet on a user page (& maybe the illusion of importance).
However - it is very difficult to assess "activity" in any objective way. I know - I've tried to deal with it in a number of places (including here I think) over time. The main reason I no longer have any of the tools I used to have - I've ceased to have a need for them. If I decide I have one I can always ask again. --Herby talk thyme 15:02, 10 March 2010 (UTC)
I don't know if anyone has pointed this out (I don't think they did), but I am surprised that Lar didn't mention it - as an Ombudsman, he would be prevented from Steward actions. A numerical standard would not take such things into consideration. Perhaps the change could be modified that the stewards who go inactive have the steward status removed but go through -confirmation- instead of -election-. The loss of the ability would not be through a breach of trust or abuse but in a sort of temporary leave (if it appears to be more than temporary, the stewards can determine that themselves). Would this be fair? Ottava Rima (talk) 15:32, 10 March 2010 (UTC)
As an ombudsman, the steward-on-leave has no access to the tools. They're removed from the Steward group during their term as ombudsman. It should probably be pointed out during confirmations when someone has been serving a term as ombudsman. Perhaps, as a bennie, they should be exempt from the activity requirements and instead simply state that they're willing/able to return to regular duty in lieu of this? Other objections such as lack of trust and civility would continue to apply. Sukida 20:00, 10 March 2010 (UTC)
So, with the above we have two different options that can compensate for the unique circumstances of Ombudsman. I think the future Ombudsman should add his own feelings on the matter too. :) Ottava Rima (talk) 21:43, 10 March 2010 (UTC)
I'm already an ombudsman, as of early February. Shihzao chose to stand for reconfirmation at the end of his term as an ombudsdman. I chose to stand at the beginning. I don't know if we want to set a policy about which is preferred if this comes up in future. Mind you, it is entirely possible that I will not be reconfirmed of course, and I certainly am taking the feedback I've gotten from the community very seriously. But until that happens, or until I resign, I am still a steward, on leave of absence from a "has the bit" perspective but still active on the mailing list. I would consider any steward exempt from the activity requirements that can only be met if one has the bit while they don't have the bit, of course. I don't know if that helps. Hope it does. ++Lar: t/c 03:41, 11 March 2010 (UTC)
I think it might be beneficial to tie the removal of rights to the confirmation: If someone stands at the end of their term, then go ahead and change steward rights out for ombudsman immediately. If someone stands before the ombudsman term, don't change the rights until after the confirmation. Right now, you're in a strange position of commenting as a steward (and being one 'by policy' until the confirmation is over and it's decided) but not actually able to act as one in any meaningful capability. Alternatively, perhaps the Foundation would consent to waiting until the confirmations were over before assigning the rights, in the future? Seems to be a bit of a mess this time around. Kylu 05:29, 11 March 2010 (UTC)

I might suggest the activity proposal be defined in ways that permit the advisory confirmations to empower the community. For example

  • Stewards must show regular activity in technical areas. (e.g. rights changes, global username changes, et cetera)
  • Stewards must show regular activity in policy discussion (e.g. Steward discussion index)

So the voting community would have the power to evaluate in their own subjective, how the stewards was performing to generalised guidance, as opposed to strict quantitative demands. Best, NonvocalScream 07:08, 11 March 2010 (UTC)

I know I for one would have no problem with this at all. I may have given a different opinion with all of my talk about the "number" being to low but to be totally honest my biggest problem is that I get frustrated when it seems like people are using the "number" (almost no matter what it is) to say that the community thoughts are invalid. In my opinion if the community decides that someone isn't active enough for the position then it should be removed basically no matter what "number" (if ANY) the community used to decide that. I would have absolutely no issue with totally getting rid of numbers and leaving it for the community to decide each election. Oh on a side not anyone notice that the "policy" has voter eligibility written down that we totally ignored this year...... o_O James (T C) 08:27, 11 March 2010 (UTC)
Inactivity is defined by policy. Yearly confirmations evaluate community trust. Having low levels of activity does not, by itself, compromise community trust, IMO. Stewards are not solely tasked with implementing community consensus; they MUST do so within the confines of policy. Yet, the community obviously has higher activity standards than the policy indicates. Thus the policy needs to be adjusted. I suggest we keep the current policy but add a "low activity" level above it (say twice as strict); Any steward may be removed during reconfirmation based solely on activity levels if they meet the low activity level and the community agrees that activity levels are too low. Stewards that fall in the inactive level will automatically not be confirmed. However, I think this should wait a while given that we now have global sysops, who will be performing many of the actions that only Stewards performed before. A definition of "Steward action" is also needed; I consider commenting on steward project pages, working on steward-related tools, and doing admin-actions on small wikis ALL Steward actions. Not to mention all the non-obvious actions (ML posts, IRC discussions/being on call, non-logged actions, bug reports...). --Daniel Mayer (mav) 14:52, 13 March 2010 (UTC)
I obviously know you feel that way since you were one of the major ones who commented like that but to be honest I must strongly object to that line of thinking. I don't even think that is what the policy actually says. The yearly confirmations are quite explicitly NOT just about trust they are about trust and whether or not the steward needs them anymore (please read the policy again, they are both explicitly mentioned). To say that the community is basically banned from using their own idea of inactivity to determine if they feel the steward needs the tools any longer is to spit in the community face and say fuck you. The activity policy as written isn't actually THAT bad, the activity policy as interpreted by about half of the stewards (and from what I can tell not the other half) is horrible and goes against the entire idea of consensus. If you want it to be that way then perhaps you can write one but the one we have no is 2 things: an automatic inactivity policy and a yearly confirmation determined on need and trust. I can not think of any possible way to legitimize saying that a community member can't use inactivity in their determination of need. In fact I would read need as almost SOLELY an activity related measure. So with all due respect I think you are amazingly off base here :/ James (T C) 15:16, 13 March 2010 (UTC)
If the steward performed (for example) 15 total actions in a given year and made at least one action each month, then obviously the steward needed the tools to perform those actions. Again, inactivity is defined by policy. Stewards must view all discussions in the light of established policy and practice when trying to determine consensus. If the policy is outdated or bad, then we should change it. Obviously we are interpreting this particular policy differently, so I'll leave things at that. --Daniel Mayer (mav) 15:37, 13 March 2010 (UTC)
The activity requirements in the policy pertain (only) to automatic removal. Falling below that line means removal without question. Above that line, the community needs to discuss it, which is what the confirmations are for. In short, the policy on inactivity is irrelevant for confirmations, since that is a community discussion and not automatic removal.
Since the community clearly has higher standards of activity than the policy outlines, we should adjust the policy to correspond more closely to the expectations we've seen in recent confirmations. As previously, there will be room for discussion about specific cases where inactivity is a concern, but the steward does fall above the line drawn by policy.  — Mike.lifeguard | @en.wb 20:10, 13 March 2010 (UTC)
I'd prefer to add a "low activity" layer on top of the current inactive definition; inactive = automatic removal but low activity would also require community consensus for removal. Making the low activity layer twice as strict as the current inactive definition seems reasonable to me. We also need to define what actions can be considered "steward actions." But I'm still a bit hesitant to enact this until we see how having global sysops affects things. --Daniel Mayer (mav) 21:05, 13 March 2010 (UTC)
I assume you mean a 2nd number? and in between number 1 and 2 you would have to have discussion and after number 2 the community can't use inactivity anymore? Why? I don't understand the insistence on removing the ability of the community to decide for themselves. It has the added benefit of not mattering what happens with global sysops, the community can discuss it on their own. In the end of the community decides that they no longer need the tools they no longer need them, PERIOD, it doesn't matter if they think they need them. I would prefer no numbers at all and let the community decide. The inactivity layer right now is ridiculously small I'd say that itself should be double (if we even have a number) but as mike said that is and should be totally separate from the confirmation. As the policy so aptly says Stewardship is not a lifetime appointment, if they have grown inactive enough that the community doesn't think they need the tools then it is time for them to move on or come back and run again when they are more active. James (T C) 22:31, 13 March 2010 (UTC)
If anything, a large part of the community wants a stricter policy on inactivity. That's why we are having a community discussion on what that new policy should be. --Daniel Mayer (mav) 23:41, 13 March 2010 (UTC)
Obviously if the community wants something else I'm fine with that I know my personal issue was two fold. I did have a problem with the ridiculously small "10 actions in the past year and nothing within the past 6 months" wording but really only because it was being used by some as why you couldn't say inactive if they had made anything in the past 6 months or more then 10 in the past year. My biggest problem was solely that people were trying to say that the community should be ignored because they met what I see as a totally separate policy. I didn't say to much DURING the confirmations because stewards, like anyone, are obviously welcome to use whatever litmus test they want when saying their bit but I got increasingly frustrated seeing what I saw as a disregard for the community in the steward discussion part. The confirm per policy comments on redux for example frustrated me to no end, most of the other "inactive per me" ones had enough people saying confirm that I was fine with them being confirmed but still feel it should have been "confirm per consensus" not policy. I'm going to post a new proposal (after thinking about it more) in a short bit below this so maybe we can have a bit more discussion of options. I also encourage others to create sections as well to discuss other options. James (T C) 00:15, 14 March 2010 (UTC)

Proposal from Jamesofur

After seeing the discussion here and elsewhere I thought I would write my own proposal to kick it off so we can start to discuss specific options. Obviously this one represents what I think is best and I welcome and encourage others to make their own proposals (or tweaks to mine). I'm not sure if we just want to have discussion at the moment or start to separate out into support/oppose etc. For now I'm leaving this without specific sections. James (T C) 04:59, 14 March 2010 (UTC)

Loss of steward access

Automatic Removal

Any steward inactive (as a steward) on meta for a full year will have their steward permissions removed. "Inactive" means fewer than 20 steward actions in last 12 months or less then 10 in the last 6 months. They may re-apply through the regular process or choose to appeal to the general community in the next confirmation period (see below). Nothing in this section shall prohibit users from using activity as a reason to remove during the confirmation period.


Stewardship is not a lifetime status. Users get it if they need it, and lose it if the community decides they are no longer trusted or active enough to need the tools. Steward status is granted until the next yearly elections, where users will be invited to comment and in particular to ask for removal of stewardship status. Should the stewards determine that community consensus exists for a steward's rights to be removed, the steward will lose their status and may re-apply if desired through the regular process.

James (T C) 04:59, 14 March 2010 (UTC)

  • 20 per year and 10 in the last six months sounds good to me so long as we define "steward action" liberally. But I don't like the "Nothing in this section shall prohibit users from using activity as a reason to remove during the confirmation period." part since that would allow any personal definition of "activity" to be used as a reason to remove any steward. This opens the door to recall of any but the most active stewards for any reason so long as it is labeled as "activity related." I would instead prefer for us to determine what the minimum activity levels we expect Stewards to have, set an absolute minimum for automatic removal, and set a level above that where the community can evaluate what was done to see if it was enough to keep the steward or not. Then make adjustments each year based on that. Below is my proposal along those lines. --Daniel Mayer (mav) 16:09, 14 March 2010 (UTC)

Proposal from mav

Here is my proposed policy. We can argue about what to enter for X and Y separately; but they should certainly be higher than the current policy. The intent of this proposal is to let stewards know what activity levels are expected while allowing changing community expectations of activity to be be used to adjust the policy each year. --Daniel Mayer (mav) 16:09, 14 March 2010 (UTC)

Loss of steward access
Low Activity

Any Steward with fewer than X steward actions* in last 12 months or less than Y in the last 6 months will automatically lose steward permissions. Any Steward with fewer than 2X actions in the last year or less than 2Y in the last six months may have steward permissions removed based solely on low activity during the next confirmation period (see below). They may re-apply through the regular process during the next Steward election. The numeric requirements of this policy will be adjusted after each confirmation to better reflect changing community expectations of steward activity.


Stewardship is not a lifetime status. Users get it if they need it, and lose it if the community decides they are no longer trusted to use steward permissions correctly or they are not active enough (see above) to need the tools. Steward status is granted until the next yearly confirmations, where users will be invited to comment and in particular to ask for removal of stewardship status. Should the stewards determine that community consensus exists for a steward's rights to be removed for reasons related to this policy, the steward will lose their status and may re-apply if desired through the regular process.

--Daniel Mayer (mav) 16:09, 14 March 2010 (UTC)
Comments on proposed 'Loss of access policy'

Another issue: what exactly is a "Steward action"? Below is my proposal. Please suggest additions/changes.

  • For the purpose of this policy, a "steward action" is defined as any rights change normally processed by Stewards, processing requests on subpages of Steward requests (including oversights), acting as an admin on small wikis the Steward does not have a local sysop flag for, participating in Steward-l or #wikimedia-steward IRC discussions, responding to emergency situations by using steward permissions, and working on Steward-related tools.

-- Daniel Mayer (mav) 16:09, 14 March 2010 (UTC)

Comments on proposed definition of 'steward action'
A "Steward action" is all "Crosswiki requests" on {{Requests}}? --Shizhao 13:34, 25 September 2010 (UTC)
I'm curious... there are a number of activities performable by stewards (acting as a local administrator on projects without one, for instance) which are difficult to enumerate for some of those who would otherwise be commenting on the confirmation: Frankly, I suspect that a great number of commentators simply plug the name of the steward into the log on Meta and base their activity on what they see there, when it's possible (though unlikely) that someone could be an effective steward without a substantial Meta log at all. Perhaps if we came up with a list of measurable steward activities and a tool to report on that activity, it would assist come confirmation time? Kylu 14:45, 25 September 2010 (UTC)



J'ai traduit cette page sur Stewards policy/fr. Je n'ai pas eu le temps de me relire, j'espère donc ne pas avoir écrit d'ânerie.

Si un steward francophone peut jeter un oeil pour vérifier que je n'ai pas fait de contresens, je lui en serais reconnaissant.


--Hercule 18:01, 9 December 2010 (UTC)