Hello --
Given that the Renameuser extension now uses the job queue for renames involving over 10,000 edits (and thus no longer locking the site for large renames), is there any technical reason not to give administrators the renameuser right?
And, if there are no technical prohibitions, is this something that would require per-project voting or would switching it across all Wikimedia wikis be acceptable?
Thanks!
MZMcBride public@mzmcbride.com
MZMcBride wrote:
Given that the Renameuser extension now uses the job queue for renames involving over 10,000 edits (and thus no longer locking the site for large renames), is there any technical reason not to give administrators the renameuser right?
Renaming a user who doesn't expect it would be very disruptive. This isn't something we'd really want to give out to every sysop.
I'd much rather see a sensible system for self-renames (with decent throttling, checks, etc)
-- brion
Self-renaming would be great so long as appropriate logging was introduced alongside the functionality. I don't think the main rename log should be used, as at the end of the day these new usernames are not being endorsed to be appropriate by a 'crat.
But yeah, sounds good.
- Chris
On Fri, Apr 3, 2009 at 10:30 AM, Brion Vibber brion@wikimedia.org wrote:
MZMcBride wrote:
Given that the Renameuser extension now uses the job queue for renames involving over 10,000 edits (and thus no longer locking the site for
large
renames), is there any technical reason not to give administrators the renameuser right?
Renaming a user who doesn't expect it would be very disruptive. This isn't something we'd really want to give out to every sysop.
I'd much rather see a sensible system for self-renames (with decent throttling, checks, etc)
-- brion
Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l
Brion Vibber wrote:
MZMcBride wrote:
Given that the Renameuser extension now uses the job queue for renames involving over 10,000 edits (and thus no longer locking the site for large renames), is there any technical reason not to give administrators the renameuser right?
Renaming a user who doesn't expect it would be very disruptive. This isn't something we'd really want to give out to every sysop.
I'd much rather see a sensible system for self-renames (with decent throttling, checks, etc)
-- brion
And manual approval. 'niceuser' suddenly renaming itself to "Brion is a ***" is not good. And bureaucrats being throttled so it can't be unrenamed worse. (And worst if it's a legitimate account which has been compromised)
On Sat, Apr 4, 2009 at 1:14 AM, Platonides Platonides@gmail.com wrote:
Brion Vibber wrote:
I'd much rather see a sensible system for self-renames (with decent throttling, checks, etc)
-- brion
And manual approval. 'niceuser' suddenly renaming itself to "Brion is a ***" is not good. And bureaucrats being throttled so it can't be unrenamed worse. (And worst if it's a legitimate account which has been compromised)
Suggestions for throttling are of course only for self-renaming.
That rename is no more disruptive than some pagemoves with the same text, except, I suppose, that they would require a bureaucrat to reverse.
On Fri, Apr 3, 2009 at 8:14 AM, Platonides Platonides@gmail.com wrote:
And manual approval. 'niceuser' suddenly renaming itself to "Brion is a ***" is not good.
Well, neither is creating a brand-new account with that name and can't do much about that except block the account(s), which is what would happen in either case. Also in either case the account would be forcibly renamed to something completely different if the name is too offensive.
So what's the difference, the number of edits? Maybe you could let regular admins rename the really common cases where some new user misspelled their namesake pokémon, or forgot to capitalize their surname, or included a swearword, etc. but require that serious "bigrename" users (with, I don't know... 5,000–10,000 edits) have to talk to a crat.
Seems like a reasonable compromise to me.
—C.W.
On Sun, Apr 5, 2009 at 9:49 AM, Charlotte Webb charlottethewebb@gmail.com wrote:
So what's the difference, the number of edits? Maybe you could let regular admins rename the really common cases where some new user misspelled their namesake pokémon, or forgot to capitalize their surname, or included a swearword, etc. but require that serious "bigrename" users (with, I don't know... 5,000–10,000 edits) have to talk to a crat.
Seems like a reasonable compromise to me.
There's still the "Why can't I log in anymore?" problem. Sysops can currently block people, but at least they're told that they blocked, who blocked them, and why.
On Sun, Apr 5, 2009 at 8:35 AM, Aryeh Gregor Simetrical+wikilist@gmail.com wrote:
There's still the "Why can't I log in anymore?" problem. Sysops can currently block people, but at least they're told that they blocked, who blocked them, and why.
Yeah that would arise when somebody has been renamed without requesting it, but as far as I know all involuntarily renamed users except for Trollderella have been indef-blocked vandal accounts (who would have little to gain by still being able to log in, and whom we wouldn't be particularly interested in helping). The one exceptional case was part of the reason Uncle Ed was pressured to resign as bureaucrat.
My point is this isn't something that happens very often now and I doubt lowering the user-group requirement would make it happen much more often. Either way it would be easy enough to revert and probably still be considered grounds for summarily removing access.
In any case it might be helpful to provide a certain grace period during which the old name still works on the login screen as a valid alias, or possibly indefinitely unless there is a reason to for another user to claim the old name (in cases of deliberate impersonation or coincidental SUL conflicts, etc.)
—C.W.
On Fri, Apr 3, 2009 at 2:30 AM, Brion Vibber brion@wikimedia.org wrote:
MZMcBride wrote:
Given that the Renameuser extension now uses the job queue for renames involving over 10,000 edits (and thus no longer locking the site for large renames), is there any technical reason not to give administrators the renameuser right?
Renaming a user who doesn't expect it would be very disruptive. This isn't something we'd really want to give out to every sysop.
I'd much rather see a sensible system for self-renames (with decent throttling, checks, etc)
I believe the immediate motivation for MZMcBride's request is actually a discussion on enwiki that there weren't enough trusted individuals to forcibly rename accounts containing sensitive information in their username. If that is the goal, then a self-renaming function would not address it.
To be honest, I don't really understand why one would use renaming over revision delete to hide bad account names, but some ArbCom members and similarly highly placed persons have stated that there are circumstances where renaming is currently preferred over revision delete.
-Robert Rohde
On Fri, Apr 3, 2009 at 8:47 AM, Robert Rohde rarohde@gmail.com wrote:
On Fri, Apr 3, 2009 at 2:30 AM, Brion Vibber brion@wikimedia.org wrote:
MZMcBride wrote:
Given that the Renameuser extension now uses the job queue for renames involving over 10,000 edits (and thus no longer locking the site for large renames), is there any technical reason not to give administrators the renameuser right?
Renaming a user who doesn't expect it would be very disruptive. This isn't something we'd really want to give out to every sysop.
I'd much rather see a sensible system for self-renames (with decent throttling, checks, etc)
I believe the immediate motivation for MZMcBride's request is actually a discussion on enwiki that there weren't enough trusted individuals to forcibly rename accounts containing sensitive information in their username. If that is the goal, then a self-renaming function would not address it.
To be honest, I don't really understand why one would use renaming over revision delete to hide bad account names, but some ArbCom members and similarly highly placed persons have stated that there are circumstances where renaming is currently preferred over revision delete.
That said, I'm also not sure why we couldn't trust sysops with this. Assuming the limitation is not technical (i.e. the servers won't explode) and renames are undoable (they are right?), then given a well-formed policy on renaming I'm not sure why giving this to all sysops would be particularly dangerous.
I certainly understand the segregation of checkuser and oversight, but renaming has never struck me as belonging to the same category of rights that need to be highly restricted.
-Robert Rohde
On Fri, Apr 3, 2009 at 12:01 PM, Robert Rohde rarohde@gmail.com wrote:
That said, I'm also not sure why we couldn't trust sysops with this. Assuming the limitation is not technical (i.e. the servers won't explode)
I think Domas might have objections if the number of renames suddenly increased by a factor of 10, actually. He's complained about them before. They have to update an awful lot of rows.
and renames are undoable (they are right?)
In principle, yes. Of course they can get stuck half-done, and if you rename someone twice in quick succession I don't know what will happen with the job queue.
Renaming is not currently a simple operation, because of our schema. We encode the username all over the place. If it were just in the user table, renaming would be less of a big deal, since we'd change one row plus maybe some stuff in RC. But we have to change a ton of rows in a ton of tables, almost all of which (for long-established users) will be cold and so will be dug out from disk. So we have to defer things, so they can be in a half-finished state, so it's slow and messy.
Also keep in mind that if someone is renamed without their knowledge, they'll be unable to log into their account. The error message will only say that the user doesn't exist, or -- if a user is created with the same name -- it will just tell them their password is wrong! Forcible renames are scary for this reason as well.
Hello,
where can I ask for Commit access to Wikimedia SVN?
Viele Grüße Jan Luca
On Fri, Apr 3, 2009 at 1:18 PM, Jan Luca jan@jans-seite.de wrote:
where can I ask for Commit access to Wikimedia SVN?
This is an okay place. You'll want to provide your public key, desired commit name, and what you want to use the commit access for.
Hello,
Key: http://toolserver.org/~jan/files/SSH2-RSA.pub
Commit name: jan Alternative: jan_luca
I will use my commit access for commit extensions. At the moment I create a extensions for my poll-script on Wikimedia Toolserver.
Viele Grüße Jan
-----Ursprüngliche Nachricht----- Von: wikitech-l-bounces@lists.wikimedia.org [mailto:wikitech-l-bounces@lists.wikimedia.org] Im Auftrag von Aryeh Gregor Gesendet: Freitag, 3. April 2009 19:22 An: Wikimedia developers Betreff: Re: [Wikitech-l] Commit access
On Fri, Apr 3, 2009 at 1:18 PM, Jan Luca jan@jans-seite.de wrote:
where can I ask for Commit access to Wikimedia SVN?
This is an okay place. You'll want to provide your public key, desired commit name, and what you want to use the commit access for.
_______________________________________________ Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l
Hello,
when will there a answer?
Gruß Jan Luca
-----Ursprüngliche Nachricht----- Von: wikitech-l-bounces@lists.wikimedia.org [mailto:wikitech-l-bounces@lists.wikimedia.org] Im Auftrag von Jan Luca Gesendet: Freitag, 3. April 2009 19:58 An: 'Wikimedia developers' Betreff: Re: [Wikitech-l] Commit access
Hello,
Key: http://toolserver.org/~jan/files/SSH2-RSA.pub
Commit name: jan Alternative: jan_luca
I will use my commit access for commit extensions. At the moment I create a extensions for my poll-script on Wikimedia Toolserver.
Viele Grüße Jan
-----Ursprüngliche Nachricht----- Von: wikitech-l-bounces@lists.wikimedia.org [mailto:wikitech-l-bounces@lists.wikimedia.org] Im Auftrag von Aryeh Gregor Gesendet: Freitag, 3. April 2009 19:22 An: Wikimedia developers Betreff: Re: [Wikitech-l] Commit access
On Fri, Apr 3, 2009 at 1:18 PM, Jan Luca jan@jans-seite.de wrote:
where can I ask for Commit access to Wikimedia SVN?
This is an okay place. You'll want to provide your public key, desired commit name, and what you want to use the commit access for.
_______________________________________________ Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l
_______________________________________________ Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l
On Wed, Apr 8, 2009 at 2:50 AM, Jan Luca jan@jans-seite.de wrote:
Hello,
when will there a answer?
I don't know. It's not up to me.
Aryeh Gregor wrote:
Also keep in mind that if someone is renamed without their knowledge, they'll be unable to log into their account. The error message will only say that the user doesn't exist, or -- if a user is created with the same name -- it will just tell them their password is wrong! Forcible renames are scary for this reason as well.
Any reason for not displaying an excerpt from the rename log for that user on login failure, much like on editing deleted pages?
On Fri, Apr 3, 2009 at 7:29 PM, Platonides Platonides@gmail.com wrote:
Any reason for not displaying an excerpt from the rename log for that user on login failure, much like on editing deleted pages?
That would be a sensible idea. In the overwhelming majority of cases, either there'd be no rename entry or the login would succeed, so it wouldn't be obtrusive.
On Fri, Apr 3, 2009 at 9:51 AM, Aryeh Gregor <Simetrical+wikilist@gmail.comSimetrical%2Bwikilist@gmail.com
wrote:
On Fri, Apr 3, 2009 at 12:01 PM, Robert Rohde rarohde@gmail.com wrote:
That said, I'm also not sure why we couldn't trust sysops with this. Assuming the limitation is not technical (i.e. the servers won't explode)
I think Domas might have objections if the number of renames suddenly increased by a factor of 10, actually. He's complained about them before. They have to update an awful lot of rows.
and renames are undoable (they are right?)
In principle, yes. Of course they can get stuck half-done, and if you rename someone twice in quick succession I don't know what will happen with the job queue.
Renaming is not currently a simple operation, because of our schema. We encode the username all over the place. If it were just in the user table, renaming would be less of a big deal, since we'd change one row plus maybe some stuff in RC. But we have to change a ton of rows in a ton of tables, almost all of which (for long-established users) will be cold and so will be dug out from disk. So we have to defer things, so they can be in a half-finished state, so it's slow and messy.
Also keep in mind that if someone is renamed without their knowledge, they'll be unable to log into their account. The error message will only say that the user doesn't exist, or -- if a user is created with the same name -- it will just tell them their password is wrong! Forcible renames are scary for this reason as well.
I agree that admins might not be necessarily trusted with permissions which would screw up site operations. We do no administrator ID verification, and the length and breadth of high load website operations/architecture experience is not a selection criteria. Limiting it to people who understand what effects it could have, and who are better known to the community/foundation, seems sensible.
Any reason not to set this up as a new userright we can have Arbcom hand to admins on request/review/background verification?
We already have them as a mechanism to extend admin rights to CU. This is less sensitive than that - but we don't have any mechanism in place to handle anything between "admins can add themselves to this" and CU right now.
OTRS might sorta count there, but it's sideways of the on-wiki administrative structure.
On Fri, Apr 3, 2009 at 6:06 PM, George Herbert george.herbert@gmail.com wrote:
I agree that admins might not be necessarily trusted with permissions which would screw up site operations. We do no administrator ID verification, and the length and breadth of high load website operations/architecture experience is not a selection criteria. Limiting it to people who understand what effects it could have, and who are better known to the community/foundation, seems sensible.
Any reason not to set this up as a new userright we can have Arbcom hand to admins on request/review/background verification?
Uh, what's you're point?
Bureaucrats are not required to identify themselves and about half of them do not, cf.: http://en.wikipedia.org/wiki/Special:Listusers/bureaucrat http://meta.wikimedia.org/wiki/Identification_noticeboard
I understand a couple of them are believed not to be legal adults either.
—C.W. —C.W.
wikitech-l@lists.wikimedia.org