From: Akash Mehta draicone@gmail.com wrote:
Do we have a way where we can scan talk pages for such warnings and come up with a record of those with 20+? I know AmiDaniel's vandalproof can look for warnings, and I'm sure there could be a similar function implemented.
I'm pretty sure it would mostly find IPs, which would be meaningless for most pre-planned blocking, as we would be unable to assess which were static and which were dynamic in the same way we generally are now.
I personally like the idea of blocking the lot of them and separating the sheep from the goats at our leisure, but this is a *wiki* and we don't and can't have this luxury. The benefits of having it outweigh the problems we would have, except in one vital way: we don't have enough excellent editors. If we prevent en masse the the chaff, we will always catch some wheat. And I speak as someone who has never, not even once, not even by accident, edited without logging in.
There may be a couple, or a dozen, or even a thousand account-holding editors who have had a set of repeated warnings for various disruptions and could therefore be streamed out by a bot. But then any sensible blocking admin would check the talk page before blocking anyway (and shame on both those that don't, and those editors who complain if they do: which someone did to me recently!) and see them and act accordingly.
If the talk page has been blanked, a human editor would check the history (one hopes). A bot could not be expected to do so without making mistakes.
Certainly, a bot can't spot idiotic warnings from trolls, which a qualified eye can, and generally does - again I speak as someone who has suffered from mad complaints from trolling users. And a bot can't spot an idiot user who suddenly has touched the clue stick and wised up, something a human /can/ do.
No. Bots have their place on Wikipedia, and I wouldn't want to be without them. But no bot can substitute for human intuition and judgment. Where such talents are required, humans must be the ones to act.
-> REDVERS
___________________________________________________________ All new Yahoo! Mail "The new Interface is stunning in its simplicity and ease of use." - PC Magazine http://uk.docs.yahoo.com/nowyoucan.html
I'm not saying we have blocks with sysop privs going around blocking users, just compiling a list of the non-IP editors with these blocks. I'm sure we could come up with a whitelist to reduce the load, and they could work from a database dump so that the server isnt affected. However, if a bot came up with a decent list, a respected user(s) could go through the list and point out problematic users to sysops for attention.
-- Akash/Draicone
On 9/13/06, Redvers @ the Wikipedia wikiredvers@yahoo.ie wrote:
From: Akash Mehta draicone@gmail.com wrote:
Do we have a way where we can scan talk pages for such warnings and come up with a record of those with 20+? I know AmiDaniel's vandalproof can look for warnings, and I'm sure there could be a similar function implemented.
I'm pretty sure it would mostly find IPs, which would be meaningless for most pre-planned blocking, as we would be unable to assess which were static and which were dynamic in the same way we generally are now.
I personally like the idea of blocking the lot of them and separating the sheep from the goats at our leisure, but this is a *wiki* and we don't and can't have this luxury. The benefits of having it outweigh the problems we would have, except in one vital way: we don't have enough excellent editors. If we prevent en masse the the chaff, we will always catch some wheat. And I speak as someone who has never, not even once, not even by accident, edited without logging in.
There may be a couple, or a dozen, or even a thousand account-holding editors who have had a set of repeated warnings for various disruptions and could therefore be streamed out by a bot. But then any sensible blocking admin would check the talk page before blocking anyway (and shame on both those that don't, and those editors who complain if they do: which someone did to me recently!) and see them and act accordingly.
If the talk page has been blanked, a human editor would check the history (one hopes). A bot could not be expected to do so without making mistakes.
Certainly, a bot can't spot idiotic warnings from trolls, which a qualified eye can, and generally does
- again I speak as someone who has suffered from mad
complaints from trolling users. And a bot can't spot an idiot user who suddenly has touched the clue stick and wised up, something a human /can/ do.
No. Bots have their place on Wikipedia, and I wouldn't want to be without them. But no bot can substitute for human intuition and judgment. Where such talents are required, humans must be the ones to act.
-> REDVERS
All new Yahoo! Mail "The new Interface is stunning in its simplicity and ease of use." - PC Magazine http://uk.docs.yahoo.com/nowyoucan.html _______________________________________________ WikiEN-l mailing list WikiEN-l@Wikipedia.org To unsubscribe from this mailing list, visit: http://mail.wikipedia.org/mailman/listinfo/wikien-l
Redvers @ the Wikipedia wrote:
I'm pretty sure it would mostly find IPs, which would be meaningless for most pre-planned blocking, as we would be unable to assess which were static and which were dynamic in the same way we generally are now.
Thankfully "IPs" lack the technical ability to upload images, regardless of copyright status. Permitting them to do so would be sheer madness.
On a similar note I've heard talk about requiring e-mail address confirmation prior to image uploading. This might be a safeguard for those casual editors who check their e-mail inbox more frequently than their talk page, and might help prevent a legitimate image from being deleted due to failure to properly tag it (fringe cases). I do not know whether this has been implimented or shot down for being too "unwiki-like", but I think the "you have Wikipedia e-mail from OrphanBot" idea is a good one. It would give the uploaders one less excuse, anyway.
However the point is probably moot now that image uploads are now stored for fucking ever and can now be undeleted at will.
--f.o.n.