It seems like we should have almost everyone at the same level, and have that level be available for the asking, and have that level be mostly nondestructive, i.e. every action is reversible, although some might be a pain in the neck to reverse. So it might look like this:
1. Newcomers -- can do everything except a small handful of actions, actions which are "pretty serious" like temporarily blocking an ip or deleting a page.
2. "sysop" or better terminology might be "community member" -- can do some extra things like nondestructive delete, temporary ip blocking (which should affect only newcomers, not other "community members"). The important _wiki_ ideal here is that community members shouldn't be able to do anything extra "in a content fight" if you see what I mean. We'll all the same, newcomers and community members, when it comes to the content -- one "check" here is that anyone can become a community member just by asking... it shouldn't be all that special, and certainly not an exclusive "club".
3. "developers" -- the main extra thing that developers should have access to would be "raw" stuff that's pretty technologically "dangerous" if you don't know what you're doing. I.E., ability to enter arbitrary SQL select statements, even ones that might be really slow. This extra power should only be used for development purposes, and not "in a content fight".
4. "sysop" -- of which there might only be 1, me, or a very small number. The sysop can do things that might be legally necessary, like immediately and totally deleting copyright violations, or things that might be necessary to fight a serious troll attack, like semi-permanently banning whole ranges of ip numbers. For most things of this nature, there's probably no need to worry about it in advance. Anyhow, "sysop" powers must never be used "in a content fight". This would be like Superman cheating at poker by looking through the cards. :-)
--Jimbo
From: "Jimmy Wales" jwales@bomis.com
- "developers" -- the main extra thing that developers should have
access to would be "raw" stuff that's pretty technologically "dangerous" if you don't know what you're doing. I.E., ability to enter arbitrary SQL select statements, even ones that might be really slow. This extra power should only be used for development purposes, and not "in a content fight".
I think I'm metadata-aware enough to be trusted with read-only SQL access to wikipedia, but I've been wrong in the past and can't think of anything immediately useful I'd do if I did have select access.
Since it appears to be a fairly strongly copyleft project, I assume you have published its SQL data model?
Lorraine Lee wrote:
From: "Jimmy Wales" jwales@bomis.com
- "developers" -- the main extra thing that developers should have
access to would be "raw" stuff that's pretty technologically "dangerous" if you don't know what you're doing. I.E., ability to enter arbitrary SQL select statements, even ones that might be really slow. This extra power should only be used for development purposes, and not "in a content fight".
I think I'm metadata-aware enough to be trusted with read-only SQL access to wikipedia, but I've been wrong in the past and can't think of anything immediately useful I'd do if I did have select access.
I'm in the same boat. And I think read-only access should be fine for just about anyone. I mean, unless it proved to be a problem for some reason.
I personally don't want to be able to delete things directly in the database via the web. I mean, if I REALLY have to delete something from the database, then for now, I'm better off going directly into mysql on the command line and in a very paranoid and cautious fashion typing it in. :-) I've never had to do that, but...
Since it appears to be a fairly strongly copyleft project, I assume you have published its SQL data model? [Wikipedia-l] To manage your subscription to this list, please go here: http://www.nupedia.com/mailman/listinfo/wikipedia-l
On dim, 2002-03-31 at 14:22, Lorraine Lee wrote:
I think I'm metadata-aware enough to be trusted with read-only SQL access to wikipedia, but I've been wrong in the past and can't think of anything immediately useful I'd do if I did have select access.
Since it appears to be a fairly strongly copyleft project, I assume you have published its SQL data model?
The table setup statements are in "wikipedia.sql" in the source tree; nothing's really documented as such, but it's mostly pretty self-explanatory*. You can browse the files in the CVS repository: http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/wikipedia/phpwiki/fpw/
* The cur_id, old_id, cur_old_version, and old_old_version fields are used to put together linked lists of version histories.
A dump of the database (80 megs compressed) is available at: http://www.wikipedia.com/tarballs/wikipedia_dump_20020321.sql.gz
-- brion vibber (brion @ pobox.com)
wikipedia-l@lists.wikimedia.org