Nobody is currently in the process of standing for adminship on en.wikipedia ... the last three ran their time and were closed earlier today.
I don't recall seeing the page empty the entire time I've been here...
On 8/23/07, George Herbert george.herbert@gmail.com wrote:
Nobody is currently in the process of standing for adminship on en.wikipedia ... the last three ran their time and were closed earlier today.
I don't recall seeing the page empty the entire time I've been here...
* '''Support'''. We don't need new admins. ~~~~ :p
Maybe now somebody will realise that something is fundamentally wrong with our adminship process. It does not scale; it worked when we were smaller, but now it seems to me that the old indicators of trustworthiness for adminship are unreliable; in the first place, the RfA process seems to diminish the importance of the trust factor, emphasising more the ILIKEIT and IDONTLIKEIT factors.
Johnleemk
Do you have any evidence to support this claim of a reduction in the trust factor? I only recently (last month or two) began contributing there, and I see people arguing for or against RFAs specifically with the word trust in ever RFA.
On 8/22/07, John Lee johnleemk@gmail.com wrote:
On 8/23/07, George Herbert george.herbert@gmail.com wrote:
Nobody is currently in the process of standing for adminship on en.wikipedia ... the last three ran their time and were closed earlier today.
I don't recall seeing the page empty the entire time I've been here...
- '''Support'''. We don't need new admins. ~~~~ :p
Maybe now somebody will realise that something is fundamentally wrong with our adminship process. It does not scale; it worked when we were smaller, but now it seems to me that the old indicators of trustworthiness for adminship are unreliable; in the first place, the RfA process seems to diminish the importance of the trust factor, emphasising more the ILIKEIT and IDONTLIKEIT factors.
Johnleemk _______________________________________________ WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: http://lists.wikimedia.org/mailman/listinfo/wikien-l
On 8/22/07, Steven Walling steven.walling@gmail.com wrote:
Do you have any evidence to support this claim of a reduction in the trust factor? I only recently (last month or two) began contributing there, and I see people arguing for or against RFAs specifically with the word trust in ever RFA.
On 8/22/07, John Lee johnleemk@gmail.com wrote:
On 8/23/07, George Herbert george.herbert@gmail.com wrote:
Nobody is currently in the process of standing for adminship on en.wikipedia ... the last three ran their time and were closed earlier today.
I don't recall seeing the page empty the entire time I've been here...
- '''Support'''. We don't need new admins. ~~~~ :p
Maybe now somebody will realise that something is fundamentally wrong
with
our adminship process. It does not scale; it worked when we were
smaller,
but now it seems to me that the old indicators of trustworthiness for adminship are unreliable; in the first place, the RfA process seems to diminish the importance of the trust factor, emphasising more the
ILIKEIT
and IDONTLIKEIT factors.
Johnleemk _______________________________________________ WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: http://lists.wikimedia.org/mailman/listinfo/wikien-l
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: http://lists.wikimedia.org/mailman/listinfo/wikien-l
What people say in their vote and why they actually voted are 2 different things. If you're new to the process you wouldnt notice it.
I can call my golden retriever a tabby cat all I want, wont make her a feline.
That said, RfA itself isnt broken, but its implementation doesnt scale. Take the same concept behind adminship, work out a better system for granting bits from that concept..
Maybe now somebody will realise that something is fundamentally wrong with our adminship process. It does not scale; it worked when we were smaller, but now it seems to me that the old indicators of trustworthiness for adminship are unreliable; in the first place, the RfA process seems to diminish the importance of the trust factor, emphasising more the ILIKEIT and IDONTLIKEIT factors.
What happens on RfA doesn't tell us anything about how well our adminship process is working. You need to look at what actual admins do, not what is said about admin candidates. No-one standing for adminship doesn't mean we don't have enough admins, you need to look at the size of admin backlogs to determine that. People giving reasons you don't like on RfAs doesn't means admins are becoming less trustworthy, you need to look at what admins do wrong to determine that.
RfA may be the cause of these problems, but you need to look at the symptoms to determine if there actually is a problem.
On 23/08/07, John Lee johnleemk@gmail.com wrote:
- '''Support'''. We don't need new admins. ~~~~ :p
Maybe now somebody will realise that something is fundamentally wrong with our adminship process. It does not scale; it worked when we were smaller, but now it seems to me that the old indicators of trustworthiness for adminship are unreliable; in the first place, the RfA process seems to diminish the importance of the trust factor, emphasising more the ILIKEIT and IDONTLIKEIT factors.
Johnleemk
Actually, the problem is Wikipedia's decision-making apparatus in general. It's allegedly "by consensus" - but if it really were, there would be paralysis (as most of the time, there are dissenters who don't back down - or even quite simply a clear lack of consensus - contributors to the particular decision-making discussion split pretty evenly two or more ways). It's allegedly not by voting (although that begs the question, why there are frequent votes), by majority or supermajority (although realistically, often decisions are taken on the basis of "most contributors to the discussion support it"). Despite idealists chosing to believe it doesn't happen, frequently the status quo is just determined by a small but persistent group, or even an individual. Considering people on the project are volunteers, no-one should have their impact on decisions decided solely by how much time they can afford, and so frequently contributors choose not to participate in decision-making (plenty of people like myself have stepped back from regular contribution due to the hassle it entails).
Zoney
Actually, the problem is Wikipedia's decision-making apparatus in general. It's allegedly "by consensus" - but if it really were, there would be paralysis (as most of the time, there are dissenters who don't back down - or even quite simply a clear lack of consensus - contributors to the particular decision-making discussion split pretty evenly two or more ways). It's allegedly not by voting (although that begs the question, why there are frequent votes), by majority or supermajority (although realistically, often decisions are taken on the basis of "most contributors to the discussion support it"). Despite idealists chosing to believe it doesn't happen, frequently the status quo is just determined by a small but persistent group, or even an individual. Considering people on the project are volunteers, no-one should have their impact on decisions decided solely by how much time they can afford, and so frequently contributors choose not to participate in decision-making (plenty of people like myself have stepped back from regular contribution due to the hassle it entails).
Yep, that's the problem. We all know that. Do you have a solution?
Admin powers should be granted automatically to users who have been around a while, basically.
On 8/23/07, Thomas Dalton thomas.dalton@gmail.com wrote:
Actually, the problem is Wikipedia's decision-making apparatus in general. It's allegedly "by consensus" - but if it really were, there would be paralysis (as most of the time, there are dissenters who don't back down - or even quite simply a clear lack of consensus - contributors to the particular decision-making discussion split pretty evenly two or more ways). It's allegedly not by voting (although that begs the question, why there are frequent votes), by majority or supermajority (although realistically, often decisions are taken on the basis of "most contributors to the discussion support it"). Despite idealists chosing to believe it doesn't happen, frequently the status quo is just determined by a small but persistent group, or even an individual. Considering people on the project are volunteers, no-one should have their impact on decisions decided solely by how much time they can afford, and so frequently contributors choose not to participate in decision-making (plenty of people like myself have stepped back from regular contribution due to the hassle it entails).
Yep, that's the problem. We all know that. Do you have a solution?
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: http://lists.wikimedia.org/mailman/listinfo/wikien-l
On 8/23/07, The Cunctator cunctator@gmail.com wrote:
Admin powers should be granted automatically to users who have been around a while, basically.
Sockpuppetry is the biggest problem with that idea.
Any reasonably designed mechanism would properly ensure that the number of good actors vastly outnumbers the number of bad actors.
On 8/23/07, Anthony wikimail@inbox.org wrote:
On 8/23/07, The Cunctator cunctator@gmail.com wrote:
Admin powers should be granted automatically to users who have been around a while, basically.
Sockpuppetry is the biggest problem with that idea.
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: http://lists.wikimedia.org/mailman/listinfo/wikien-l
Admin powers should be granted automatically to users who have been around a while, basically.
Not everyone that's been around a while is suitable to become an admin. We generally try and avoid banning people if it can be helped, if not banning someone would result in them becoming an admin, we would end up with no choice but to ban them.
Also, it's necessary for admins not only to be trustworthy, but to be trusted. We have enough problems with that as it is, making everyone that's been around for 6 months (or whatever) an admin would reduce that trust even more.
On 8/23/07, Thomas Dalton thomas.dalton@gmail.com wrote:
Admin powers should be granted automatically to users who have been around a while, basically.
Not everyone that's been around a while is suitable to become an admin. We generally try and avoid banning people if it can be helped, if not banning someone would result in them becoming an admin, we would end up with no choice but to ban them.
No, you just remove the admin powers.
Also, it's necessary for admins not only to be trustworthy, but to be trusted. We have enough problems with that as it is, making everyone that's been around for 6 months (or whatever) an admin would reduce that trust even more.
I really don't know what you mean by that.
Rich Holton proposed this back in February:
Select at random 100 editors who meet some minimal criteria* and make them admins. Make it clear to them that they may turn down adminship without prejudice.
Then, we watch these 100 "probationary" admins for 3 months. If they abuse their admin powers in that time, their admin status is removed. Otherwise, we treat them as regular admins. The only difference with a "probationary" admin is the level of scrutiny they receive.
If this works, then after 3 months we do it again. And again every three months. Soon, adminship loses almost all of its "status" appeal. It's just something you'll get if you hang around and keep your nose clean.
Of course, you can still apply through RfA. But I predict that RfA will quickly become much less political and controversial.
*My suggestion for "minimal criteria": At least 50 edits to at least 10 different non-own-user pages for each of the past three months, and No blocks in the past three months
Essentially, just enough to give a good indication that the user is involved and isn't a trouble-maker. Nothing more.
No, you just remove the admin powers.
So you are proposing introducing a quick and easy way to desysop people? Perhaps allow crats to desysop (requires a little developer work, but not too much), and leave it pretty much to their discretion, similar to how blocks are pretty much at admins discretion now.
Also, it's necessary for admins not only to be trustworthy, but to be trusted. We have enough problems with that as it is, making everyone that's been around for 6 months (or whatever) an admin would reduce that trust even more.
I really don't know what you mean by that.
It's all about perception. An admin may be trustworthy, but if the community don't trust them, then it can cause numerous problems, lots of time wasted dealing with complaints, etc.
Rich Holton proposed this back in February:
Select at random 100 editors who meet some minimal criteria* and make them admins.
Yes, and very few people liked the idea then, too.
On 23/08/07, Thomas Dalton thomas.dalton@gmail.com wrote:
No, you just remove the admin powers.
So you are proposing introducing a quick and easy way to desysop people?
When on the three months' probation.
Perhaps allow crats to desysop (requires a little developer work, but not too much), and leave it pretty much to their discretion, similar to how blocks are pretty much at admins discretion now.
Though unblocking is at other admins' discretion, and there's lots of back-and-forth on WP:ANI if an admin doesn't seem to have a feel for the spirit of the blocking policy. (Not to mention here.)
- d.
Perhaps allow crats to desysop (requires a little developer work, but not too much), and leave it pretty much to their discretion, similar to how blocks are pretty much at admins discretion now.
Though unblocking is at other admins' discretion, and there's lots of back-and-forth on WP:ANI if an admin doesn't seem to have a feel for the spirit of the blocking policy. (Not to mention here.)
Yes, and it works quite well. Wheel warring is sufficiently frowned upon that admins don't go around undoing every block they disagree with, but when an admin makes a serious error in judgement, someone will undo it. I think the same principle for desysopings would work equally well (although may require a few more crats than we have at the moment).
On 23/08/07, Thomas Dalton thomas.dalton@gmail.com wrote:
Yes, and it works quite well. Wheel warring is sufficiently frowned upon that admins don't go around undoing every block they disagree with, but when an admin makes a serious error in judgement, someone will undo it. I think the same principle for desysopings would work equally well (although may require a few more crats than we have at the moment).
I'm not sure the 'crats want the headache ;-)
- d.
On 8/23/07, The Cunctator cunctator@gmail.com wrote:
No, you just remove the admin powers.
I support this strongly.
If our goal is actually to have good admins then we must be willing to remove the powers from people who misuse them. If we are already doing that, then we have no reason to worry about giving adminship to someone who will misuse it, since we could just reverse their actions and remove the adminship.
Rich Holton proposed this back in February:
Select at random 100 editors who meet some minimal criteria* and make them admins. Make it clear to them that they may turn down adminship without prejudice.
So long as we make the minimal criteria high enough to require a substantial investment of productive time.
We don't want people minting tons of admin accounts for the purpose of defacing the main page. We can avoid this if we make the time investment long enough that the 20 second long main page defacement won't justify the many hours of work required to get adminship.
[snip]
*My suggestion for "minimal criteria": At least 50 edits to at least 10 different non-own-user pages for each of the past three months, and No blocks in the past three months
Too easy to just mint accounts. If we are going to take a fixed number of probationary users, we should use criteria like:
*Account is not an admin, and has no arbcom ruling preventing them from being an admin. *Account is at least 6 months old. *Account has had no blocks in the last three months *Account has at least 50 edits to non-userspace in each of the last three months. *Account has edits to at least 10 distinct non-userspace pages in each of the last three months.
Then we make a list of accounts meeting that criteria, order them by number of non-userspace edits, and pick 100 people at random from the top 50%.
This final step means that the criteria to get accepted depends on the activity level of all the other non-admins.
I'd gladly produce a list of users if we have a 'crat that wants to try this.
On 23/08/07, Gregory Maxwell gmaxwell@gmail.com wrote:
If our goal is actually to have good admins then we must be willing to remove the powers from people who misuse them. If we are already doing that, then we have no reason to worry about giving adminship to someone who will misuse it, since we could just reverse their actions and remove the adminship.
At the moment the arbcom can remove admin powers, though it takes blowing it quite severely for them to act fast.
Most of the whining that "it's impossible to de-admin someone" means the complainant is upset that someone can't be deadminned because the complainant is upset, and there's all this *tedious procedure* to go through. In practice, it's easier now than it's ever been.
- d.
On 8/23/07, David Gerard dgerard@gmail.com wrote:
At the moment the arbcom can remove admin powers, though it takes blowing it quite severely for them to act fast.
Most of the whining that "it's impossible to de-admin someone" means the complainant is upset that someone can't be deadminned because the complainant is upset, and there's all this *tedious procedure* to go through. In practice, it's easier now than it's ever been.
... They should also recognize that the best way to make deadminship less of a big deal is to allow it to become necessary a little more often.
On 8/23/07, Gregory Maxwell gmaxwell@gmail.com wrote:
On 8/23/07, David Gerard dgerard@gmail.com wrote:
At the moment the arbcom can remove admin powers, though it takes blowing it quite severely for them to act fast.
Most of the whining that "it's impossible to de-admin someone" means the complainant is upset that someone can't be deadminned because the complainant is upset, and there's all this *tedious procedure* to go through. In practice, it's easier now than it's ever been.
... They should also recognize that the best way to make deadminship less of a big deal is to allow it to become necessary a little more often.
Well, not "necessary". The people who argue that deadminship should be easier most likely want it to be done at times when it isn't "necessary".
What about "semi-adminship"? The power to block, delete, and view deleted articles, but not to unblock, undelete, edit while blocked, or edit while protected?
Maybe the block power could even be time limited.
Where are the biggest backlogs that make more admins necessary in the first place?
Anthony schreef:
What about "semi-adminship"? The power to block, delete, and view deleted articles, but not to unblock, undelete, edit while blocked, or edit while protected?
That would give the "destructive" admin powers to semi-admins, while withholding the "constructive" ones.
It would make more sense to me to give deletion-related powers to admins first, because those can do less permanent damage. (Deleted articles can always be undeleted; but we may never be able to get contributors back who have been blocked.)
Eugene
On 8/23/07, Eugene van der Pijll eugene@vanderpijll.nl wrote:
Anthony schreef:
What about "semi-adminship"? The power to block, delete, and view deleted articles, but not to unblock, undelete, edit while blocked, or edit while protected?
That would give the "destructive" admin powers to semi-admins, while withholding the "constructive" ones.
I strongly disagree with your characterization of those powers as "destructive" and "constructive". All admin powers are capable of being both "destructive" and "constructive".
It would make more sense to me to give deletion-related powers to admins first, because those can do less permanent damage. (Deleted articles can always be undeleted; but we may never be able to get contributors back who have been blocked.)
I don't know, I think it really comes down to a question of where the backlog is.
The backlog is probably in the area of making sound judgements (especially in deciding whether or not to delete). This can be done without adminship at all, though.
The backlog is probably in the area of making sound judgements (especially in deciding whether or not to delete). This can be done without adminship at all, though.
That's an interesting point. People seem to miss that there is more to adminship than the technical admin tools. Closing AfDs is an admin power, being automatically whitelisted on most vandal-fighting tools is an admin power, being able to reject unblock requests is an admin power, etc., etc. Just because they are explicitly included in the $wgPermissionsGroup['sysop'] part of LocalSettings.php doesn't mean they aren't admin powers.
On 23/08/07, Thomas Dalton thomas.dalton@gmail.com wrote:
Just because they are explicitly included
That should be "aren't", of course.
On 8/23/07, Thomas Dalton thomas.dalton@gmail.com wrote:
The backlog is probably in the area of making sound judgements (especially in deciding whether or not to delete). This can be done without adminship at all, though.
That's an interesting point. People seem to miss that there is more to adminship than the technical admin tools. Closing AfDs is an admin power,
Last time I checked policy stated (*) that only closing highly disputed AfDs was "an admin power", though those are the only ones that take a significant amount of time anyway. This was a step up in power from the rules during the early days, when the rules specifically said that non-admins were allowed to close VfDs.
(*) I can't find the policy at the moment, though.
On 8/23/07, Anthony wikimail@inbox.org wrote:
On 8/23/07, Thomas Dalton thomas.dalton@gmail.com wrote:
The backlog is probably in the area of making sound judgements (especially in deciding whether or not to delete). This can be done without adminship at all, though.
That's an interesting point. People seem to miss that there is more to adminship than the technical admin tools. Closing AfDs is an admin power,
Last time I checked policy stated (*) that only closing highly disputed AfDs was "an admin power", though those are the only ones that take a significant amount of time anyway. This was a step up in power from the rules during the early days, when the rules specifically said that non-admins were allowed to close VfDs.
(*) I can't find the policy at the moment, though.
And here it is: http://en.wikipedia.org/wiki/Wikipedia:Deletion_process
"Non-administrators closing discussions"
According to this, non-admins can close keep results, but shouldn't close delete results. They can even close high disputed keep results, but they are discouraged from doing so if they "are not familiar with deletion policy or the workings of deletion discussions".
The rationale for saying that non-admins shouldn't close delete results is that "they lack the technical ability to delete pages". But that ignores the fact that the vast majority of the work of deletion is deciding what to do, not in hitting the delete button.
And here it is: http://en.wikipedia.org/wiki/Wikipedia:Deletion_process
"Non-administrators closing discussions"
According to this, non-admins can close keep results, but shouldn't close delete results. They can even close high disputed keep results, but they are discouraged from doing so if they "are not familiar with deletion policy or the workings of deletion discussions".
The rationale for saying that non-admins shouldn't close delete results is that "they lack the technical ability to delete pages". But that ignores the fact that the vast majority of the work of deletion is deciding what to do, not in hitting the delete button.
I'm not sure that page accurately reflects how things actually are. Policy is determined more by what actually happens than by what is written down.
On 8/23/07, Thomas Dalton thomas.dalton@gmail.com wrote:
And here it is: http://en.wikipedia.org/wiki/Wikipedia:Deletion_process
"Non-administrators closing discussions"
According to this, non-admins can close keep results, but shouldn't close delete results. They can even close high disputed keep results, but they are discouraged from doing so if they "are not familiar with deletion policy or the workings of deletion discussions".
The rationale for saying that non-admins shouldn't close delete results is that "they lack the technical ability to delete pages". But that ignores the fact that the vast majority of the work of deletion is deciding what to do, not in hitting the delete button.
I'm not sure that page accurately reflects how things actually are. Policy is determined more by what actually happens than by what is written down.
Perhaps, but policy which contradicts things that are written down is much easier to change :)
Perhaps, but policy which contradicts things that are written down is much easier to change :)
Changing what's written down involves clicking the edit button (and hoping nobody notices, which they often don't until it's used against them). Changing what people actually do involves changing mindsets - much harder.
Last time I checked policy stated (*) that only closing highly disputed AfDs was "an admin power", though those are the only ones that take a significant amount of time anyway. This was a step up in power from the rules during the early days, when the rules specifically said that non-admins were allowed to close VfDs.
(*) I can't find the policy at the moment, though.
Well yes, uncontroversial keeps can be closed by anyone. My point still stands, though.
On 23/08/07, Eugene van der Pijll eugene@vanderpijll.nl wrote:
Anthony schreef:
What about "semi-adminship"? The power to block, delete, and view deleted articles, but not to unblock, undelete, edit while blocked, or edit while protected?
That would give the "destructive" admin powers to semi-admins, while withholding the "constructive" ones.
I wouldn't classify them as "destructive" and "constructive". I'd classify them as actions "against" (can't think of a better word) regular users, and actions "against" admins. It makes sense for semi-admins to have "power over" (again, can't think of a better phrase) regular users, but not (other) admins.
Thomas Dalton schreef:
On 23/08/07, Eugene van der Pijll eugene@vanderpijll.nl wrote:
Anthony schreef:
What about "semi-adminship"? The power to block, delete, and view deleted articles, but not to unblock, undelete, edit while blocked, or edit while protected?
That would give the "destructive" admin powers to semi-admins, while withholding the "constructive" ones.
I wouldn't classify them as "destructive" and "constructive". I'd classify them as actions "against" (can't think of a better word) regular users, and actions "against" admins. It makes sense for semi-admins to have "power over" (again, can't think of a better phrase) regular users, but not (other) admins.
No, it makes more sense for semi-admins to be able to act against admins. Admins are better equiped to handle "hostile" acts, and should have enough experience to handle semi-admins who do something wrong.[0] We cannot expect that from new users, and it is they who must be protected against less well thought out actions of inexperienced admins.
Eugene
[0] In theory.
No, it makes more sense for semi-admins to be able to act against admins. Admins are better equiped to handle "hostile" acts, and should have enough experience to handle semi-admins who do something wrong.[0] We cannot expect that from new users, and it is they who must be protected against less well thought out actions of inexperienced admins.
But admins shouldn't need actions taken against them except it extreme circumstances. In extreme circumstances we would want full admins handling it, because they are more experienced. It is the routine stuff (and actions against regular users are generally routine) that can be done by semi-admins.
On 23/08/07, Anthony wikimail@inbox.org wrote:
What about "semi-adminship"? The power to block, delete, and view deleted articles, but not to unblock, undelete, edit while blocked, or edit while protected?
Maybe the block power could even be time limited.
Suggested hundreds of times in hundreds of different forms, and rejected every time. Also, admins can't edit while blocked, they can, however, unblock themselves, so it's generally a moot point. While there are certainly reasons for introducing such a system, and I may even support if done right, I don't think we would get a consensus to do so.
Where are the biggest backlogs that make more admins necessary in the first place?
To be honest, I don't think there are any excessive backlogs. I think we have enough admins as it is. I think admin numbers are increasing roughly linearly, while everything else is increasing roughly exponentially, so we may run into problems eventually, but at the moment, things are working. That's the major point people complaining about RfA seem to miss (and a point I made earlier in this thread).
On 23/08/07, Thomas Dalton thomas.dalton@gmail.com wrote:
To be honest, I don't think there are any excessive backlogs. I think we have enough admins as it is. I think admin numbers are increasing roughly linearly, while everything else is increasing roughly exponentially, so we may run into problems eventually, but at the moment, things are working. That's the major point people complaining about RfA seem to miss (and a point I made earlier in this thread).
I'm not convinced - we have too many burnout cases, and although good admins are getting through RFA pretty reliably, they too frequently strike me as future burnout cases.
- d.
On 8/23/07, Thomas Dalton thomas.dalton@gmail.com wrote:
To be honest, I don't think there are any excessive backlogs. I think we have enough admins as it is. I think admin numbers are increasing roughly linearly, while everything else is increasing roughly exponentially, so we may run into problems eventually, but at the moment, things are working.
Actually things are no longer increasing exponentially. At least, number of articles isn't, and hasn't been since about September 2006. New articles per day has been holding steady in the 1750 range for months now (down from a peak of around 2400/day in August 2006). We'll probably end up with fewer new articles in 2007 than there were in 2006.
Anecdotal evidence suggests that a similar leveling off has happened in terms active users; the active user pool may (or may not) still be growing, but I don't think it's growing exponentially anymore, either.
(I don't know if we keep good data on edit rates, but I think it's the same story there as well.)
-Sage (User:Ragesoss)
On 8/24/07, Sage Ross ragesoss+wikipedia@gmail.com wrote:
On 8/23/07, Thomas Dalton thomas.dalton@gmail.com wrote:
To be honest, I don't think there are any excessive backlogs. I think we have enough admins as it is. I think admin numbers are increasing roughly linearly, while everything else is increasing roughly exponentially, so we may run into problems eventually, but at the moment, things are working.
Actually things are no longer increasing exponentially. At least, number of articles isn't, and hasn't been since about September 2006.
This was discussed here: http://www.andrewlih.com/blog/2007/07/10/unwanted-new-articles-in-wikipedia/
Complete with the obligatory reference to 1984. :)
On 8/23/07, Anthony wikimail@inbox.org wrote:
What about "semi-adminship"? The power to block, delete, and view deleted articles, but not to unblock, undelete, edit while blocked, or edit while protected?
Maybe the block power could even be time limited.
Where are the biggest backlogs that make more admins necessary in the first place?
None. Current requirement for admin actions could be trivialy met by existing admins. In practice it is met by a group of less than 100.
11K actions would get you into the top 50 most active admins of all time.
http://en.wikipedia.org/wiki/User:Misza13/admin_actions
Snapshots over a shorter time period show an even more extream patturn of a small number of admins carrying out most of the admin actions.
We don't /need/ the majority of existing admins as admins.
On 8/23/07, geni geniice@gmail.com wrote:
On 8/23/07, Anthony wikimail@inbox.org wrote:
What about "semi-adminship"? The power to block, delete, and view deleted articles, but not to unblock, undelete, edit while blocked, or edit while protected?
Maybe the block power could even be time limited.
Where are the biggest backlogs that make more admins necessary in the first place?
None. Current requirement for admin actions could be trivialy met by existing admins. In practice it is met by a group of less than 100.
11K actions would get you into the top 50 most active admins of all time.
http://en.wikipedia.org/wiki/User:Misza13/admin_actions
Snapshots over a shorter time period show an even more extream patturn of a small number of admins carrying out most of the admin actions.
We don't /need/ the majority of existing admins as admins.
It looks like most of what the most active admins do is predominantly deletions.
There's a much wierder distribution for blocks - 2 accounts with about 25k blocks each (Curps, Can't sleep...), 3 with more than 5k each (Naconkantari, Luna Santin, Freakofnurture), 4 with 4-5k each, 9 with 3-4k each, 17 with 2-3k each, about 50 more with 1-2k, and then it drops off from there (I have 274 blocks, and am about 200 down the # of blocks list from the top).
On 8/23/07, geni geniice@gmail.com wrote:
We don't /need/ the majority of existing admins as admins.
The vast majority of the admin actions of which you speak are anti-vandalism ones. We need the majority of existing admins as admins not because we need that many to do the load, but because better decisions get made the more admins there are and the more time an admin has to fix a problem.
Looking purely at numbers ignores the fact that some admin actions are made with little thought, while some require complex decision-making.
IMO, a good number of of the '100 most active admins' are those most likely to burn out.
In fact, in some respects, the concentration of admin actions is a PROBLEM; IMO, it shows that too many admin actions are being done too quickly.
-Matt
On 8/23/07, Matthew Brown morven@gmail.com wrote:
On 8/23/07, geni geniice@gmail.com wrote:
We don't /need/ the majority of existing admins as admins.
The vast majority of the admin actions of which you speak are anti-vandalism ones.
Nope mostly image copyright issues.
We need the majority of existing admins as admins not because we need that many to do the load, but because better decisions get made the more admins there are and the more time an admin has to fix a problem.
Which would be great if the majority of existing admins were spreading the load.
Looking purely at numbers ignores the fact that some admin actions are made with little thought, while some require complex decision-making.
Includeing that gets you exactly the same result. Small number of admins perform most admin tasks while the rest are fairly inactive as admins.
IMO, a good number of of the '100 most active admins' are those most likely to burn out.
Only because if you ah burn out from minimal activity it is somewhat hard to tell.
In fact, in some respects, the concentration of admin actions is a PROBLEM; IMO, it shows that too many admin actions are being done too quickly.
So figure out how to get the less active admins more active.
On 8/23/07, geni geniice@gmail.com wrote:
None. Current requirement for admin actions could be trivialy met by existing admins. In practice it is met by a group of less than 100.
[snip]
The actions which get done are the mindless ones.. We don't need more admins doing those. (we need some admined bots, but thats another matter).
Actions which are more complicated and require more thought are the ones don't do quite so well. It's hard to tell what the real need is since our need seems to scale to match our resources.
Most of the whining that "it's impossible to de-admin someone" means the complainant is upset that someone can't be deadminned because the complainant is upset, and there's all this *tedious procedure* to go through. In practice, it's easier now than it's ever been.
Yes, but the reason we make it quite difficult to desysop someone is because it's rarely necessary, since we are so careful not to promote anyone questionable. If we relax promotions, we would need to relax demotions too. And if we relax demotions, we would be able to relax promotions. It's rather circular, but the end result is that we have to do both at the same time for it to work.
Yes, but the reason we make it quite difficult to desysop someone is because it's rarely necessary, since we are so careful not to promote anyone questionable.
Sorry, I should clarify that. It's rarely necessary, so most requests should be ignored, so we make it hard to weed out those ignorable requests.
On 23/08/07, Thomas Dalton thomas.dalton@gmail.com wrote:
Most of the whining that "it's impossible to de-admin someone" means the complainant is upset that someone can't be deadminned because the complainant is upset, and there's all this *tedious procedure* to go through. In practice, it's easier now than it's ever been.
Yes, but the reason we make it quite difficult to desysop someone is because it's rarely necessary, since we are so careful not to promote anyone questionable. If we relax promotions, we would need to relax demotions too. And if we relax demotions, we would be able to relax promotions. It's rather circular, but the end result is that we have to do both at the same time for it to work.
What criteria for deadminning are you thinking of? "Unpopular for deleting 1000 fair-abuse images" or something that'll be usable for that shouldn't be amongst them, for instance.
- d.
What criteria for deadminning are you thinking of? "Unpopular for deleting 1000 fair-abuse images" or something that'll be usable for that shouldn't be amongst them, for instance.
The same at the criteria for a block. Violating policy. When such a violation requires action and what action is required should be at the discretion of the crat, just as it's at the discretion of the admin for blocks.
On Aug 23, 2007, at 12:46 PM, Thomas Dalton wrote:
If we relax promotions, we would need to relax demotions too. And if we relax demotions, we would be able to relax promotions.
I'm not sure what is needed is a relaxation of promotion either - at least, not relaxation in the sense of "easier." In the sense of "thorazine," perhaps. That is to say, I think what is needed is not lower standards, but a refusal to allow stupid standards to hold sway.
-Phil
If our goal is actually to have good admins then we must be willing to remove the powers from people who misuse them. If we are already doing that, then we have no reason to worry about giving adminship to someone who will misuse it, since we could just reverse their actions and remove the adminship.
I agree, the main thing stopping relaxing adminship requirements is the lack of an easy way to desysop people when we make mistakes. However, even with an easy way to desysop, I don't think there would be net gain from relaxing requirements as much as you are suggesting. We could go with something inbetween. How about:
1) User requests adminship 2) Other users have a week to give reasons why they shouldn't be promoted, and respond to other people's reasons. No '''Support''' votes, support is the default. 3) After the week, a crat comes along, reads the reasons, and if, based on their personal judgement, there is not a good reason to withhold the mop, they promote the user.
The main thing is getting rid of support votes. You do away with them, and you remove any semblance of a vote. I've been trying to come up with a slightly stricter suggestion along these lines which involved coming to a consensus (or possibly just voting) on each reason not to promote, but I couldn't think of any way that would actually work. Just leaving it up to the crat makes it much easier (informal guidelines would probably arise naturally, and be ignored where appropriate). It requires trusting crats a little more than we do now, but I think RfB works reasonably well (and shouldn't change much, if at all), so that shouldn't be a problem.
On 23/08/07, Thomas Dalton thomas.dalton@gmail.com wrote:
- Other users have a week to give reasons why they shouldn't be
promoted, and respond to other people's reasons. No '''Support''' votes, support is the default.
Ooh, I like that. Of course, support should be the default.
I wonder if there's any room to say "this objection is specious."
- d.
On 23/08/07, David Gerard dgerard@gmail.com wrote:
On 23/08/07, Thomas Dalton thomas.dalton@gmail.com wrote:
- Other users have a week to give reasons why they shouldn't be
promoted, and respond to other people's reasons. No '''Support''' votes, support is the default.
Ooh, I like that. Of course, support should be the default.
I wonder if there's any room to say "this objection is specious."
That's what I meant by "respond to other people's reasons". Just saying an objection is specious would be rather unhelpful, of course, reasons should always be given. "You are talking complete nonsense" may be an acceptable reason in some cases, of course.
On 8/24/07, The Cunctator cunctator@gmail.com wrote:
Admin powers should be granted automatically to users who have been around a while, basically.
Yikes, I can think of any number of long term editors who shouldn't be admins.
Hang on. Most of them are...
This has been addressed before;
The standard perhaps should be relaxed, a little.
Navou
-----Original Message----- From: wikien-l-bounces@lists.wikimedia.org [mailto:wikien-l-bounces@lists.wikimedia.org] On Behalf Of Skyring Sent: Sunday, August 26, 2007 8:15 PM To: English Wikipedia Subject: Re: [WikiEN-l] No RFAs in progress...
On 8/24/07, The Cunctator cunctator@gmail.com wrote:
Admin powers should be granted automatically to users who have been around a while, basically.
Yikes, I can think of any number of long term editors who shouldn't be admins.
Hang on. Most of them are...
On 8/23/07, Thomas Dalton thomas.dalton@gmail.com wrote:
Actually, the problem is Wikipedia's decision-making apparatus in general. It's allegedly "by consensus" - but if it really were, there would be paralysis (as most of the time, there are dissenters who don't back down - or even quite simply a clear lack of consensus - contributors to the particular decision-making discussion split pretty evenly two or more ways). It's allegedly not by voting (although that begs the question, why there are frequent votes), by majority or supermajority (although realistically, often decisions are taken on the basis of "most contributors to the discussion support it"). Despite idealists chosing to believe it doesn't happen, frequently the status quo is just determined by a small but persistent group, or even an individual. Considering people on the project are volunteers, no-one should have their impact on decisions decided solely by how much time they can afford, and so frequently contributors choose not to participate in decision-making (plenty of people like myself have stepped back from regular contribution due to the hassle it entails).
Yep, that's the problem. We all know that. Do you have a solution?
The first step is to stop lying about it: stop calling it consensus, stop saying it isn't a vote. Describe what's really happening. Admit there's a problem. Stop drinking the cool-aid.