Magnus brings up an important point, which has to do with how this will interact with namespaces, and particularly the stable namespace. I think we need to have some manual control over what goes into the stable namespace, so that we can be sure that the people approving articles for the stable namespace actually know what they are talking about. It may be that people with enough KP should be trusted to only approve articles on subjects where they have actual experience, but we may still want some kind of manual check in place for that function.
Another good point Magnus brings up is that blocked IP's will only work against people with a static IP address, or against one log in session. Additionally blocking an IP could stop someone else from being able to use the wikipedia, if they are later assigned that IP. Tis' true, but I still think people with enough KP should be able to block an IP but only temporarily, and by this I mean a really short time, like an hour or two. This will be enough to make your average vandal get board and leave, and will be highly unlikely to effect someone else who might log on with that IP wanting to edit a page. (Obviously blocking an IP should only disallow edit access and not mere read access). Actually we should probably have an automatic IP blocker, which shuts down an IP address for an hour or two if someone makes more than 10 edits in 30 seconds, to make programmatic vandalism harder.
Yours Mark
-----Original Message----- From: Magnus Manske [mailto:Magnus.Manske@epost.de] Sent: Thursday, October 18, 2001 3:41 PM To: wikipedia-l@nupedia.com Subject: RE: [Wikipedia-l] A proposal for the new software
As the (main) author of the new software, I'd like to contribute some things to this debate:
- Watching the actions of a signed-up user will be very simple, even if he/she logs in from different machines. - Counting edits/new articles will be as simple. - After each "karma point" addition, the status could be checked and basic rights could be given. - All pages can be locked to give write access only to people with the necessary user rights.
So, no technical problem with that. But, think about what I originally had in mind (I mentioned that somewhere already) :
- Have about a dozen "sysops"/administrators. Larry, Jimbo, a few others (and currently myself, for maintnance;) - Sysops can do everything: edit other user's rights, delete pages (and I mean delete, not just remove the contents), mess directly with the database etc. - Sysops can create "editors", which have less rights, but of whom there are many. - *Everybody* can edit pages in the normal wikipedia namespace - Good articles can be advanced into an "approved" namespace (by everybody, or by a special "reviewer" class) - Editors can advance articles from the "approved" namespace to the "stable" namespace, or remove it from "approved" - The "stable" namespace can only be edited by sysops
"Reviewers" and maybe "editors" could also be generated by karma points, or by LSD ;)
Additionally, central pages could still be protected, and my new variables will change the date and the number of articles on the HomePage automatically.
A word to "blocked IPs": Almost everyone who goes online via an ISP gets a random IP from the ISP every time he/she dials in. Blocking such an IP would not stop trolls, but it would stop other harmless people who come in through the same ISP at a later time. We don't want "wikipedia colateral damage", now do we?
Magnus
[Wikipedia-l] To manage your subscription to this list, please go here: http://www.nupedia.com/mailman/listinfo/wikipedia-l
On Thu, 18 Oct 2001, Mark Christensen wrote:
Magnus brings up an important point, which has to do with how this will interact with namespaces, and particularly the stable namespace. I think we need to have some manual control over what goes into the stable namespace, so that we can be sure that the people approving articles for the stable namespace actually know what they are talking about. It may be that people with enough KP should be trusted to only approve articles on subjects where they have actual experience, but we may still want some kind of manual check in place for that function.
Golly, I hope we don't have to resolve that issue in order to resolve Jimbo's issue. I have more than enough on my plate as it is! :-) If you want to hold forth on the issue, though, please do on
http://www.wikipedia.com/wiki/Wikipedia_approval_mechanism
Many people have already made several good proposals and some surprisingly :-) polite debate has been ongoing for some time now.
Larry
Larry Sanger wrote:
On Thu, 18 Oct 2001, Mark Christensen wrote:
Magnus brings up an important point, which has to do with how this will interact with namespaces, and particularly the stable namespace. I think we need to have some manual control over what goes into the stable namespace, so that we can be sure that the people approving [...]
http://www.wikipedia.com/wiki/Wikipedia_approval_mechanism
Many people have already made several good proposals and some surprisingly :-) polite debate has been ongoing for some time now.
I was thinking here. What if you could name and address every version of an article, e.g. http://www.wikipedia.com/wiki/Radio would link to the "current" or "latest" version of the article on Radio, but http://www.wikipedia.com/wiki/Radio/12 would link to version 12 of that article. Then somebody could "approve" version 12, or maybe have copyright to version 23, and this would not stop anybody from continuing to modify the article into version 34. You could create "Nupedia 2.0" as a web of pointers to approved versions of Wikipedia articles. The articles (and their versions) would be part of Wikipedia, but the approval mechanisms would be part of Nupedia. Any modification to a Wikipedia article would create a new version, and not change the approved version.
By separating article storage from approval, you could also have multiple approval webs. You could have Nupedia with high academic standards, and PC-pedia with high demands on political correctness, only linking to politically correct articles. Or a leftist-pedia web that only links to article versions with non-capitalist views. Or a parental-guidance-pedia that only links to articles with non-explicit contents.
I think this sort of separation would do the least harm to wikipedia, because people could continue to write wikipedia articles at the speed of light without having to think too much about getting approval. The approval could be a slower process without hampering the creativeness.
--- Lars Aronsson lars@aronsson.se wrote:
I was thinking here. What if you could name and address every version of an article, e.g. ... http://www.wikipedia.com/wiki/Radio/12
That already works. The format is: http://www.wikipedia.com/wiki.cgi?action=browse&id=&revision=
For instance, revision 770 of HomePage is http://www.wikipedia.com/wiki.cgi?action=browse&id=HomePage&revision...
It would be trivial to configure wikipedia.com to translate a flat URL like Lars suggests so it returned the requested revision. The only problem today is that old revisions are purged. (HomePage only goes back to #709 on October 5, for example.)
How 'bout it, O beneficent and philanthropic owner of the database (Jimbo)? Would you be willing to disable the purge of old revisions? Seems to me mostly to be a matter of disk space.
<>< [[Tbc]]
__________________________________________________ Do You Yahoo!? Make a great connection at Yahoo! Personals. http://personals.yahoo.com
Tim Chambers wrote:
How 'bout it, O beneficent and philanthropic owner of the database (Jimbo)? Would you be willing to disable the purge of old revisions? Seems to me mostly to be a matter of disk space.
Yes, and disk space is increasingly very very cheap. However, the machine this is on at the moment isn't exactly overflowing with extra space.
For now, I've upped the number of days to keep the revision history from 14 to 28. That will have the effect of nothing expiring for the next 14 days. If disk space isn't an issue at that time, I'll bump it up further. If disk space *is* an issue at that time, I'll set the wheels in motion to move wikipedia to a machine with essentially infinite disk space. :-)
Either way, I'm not opposed in principle to keeping _all_ of revision history, but I'm also not promising beyond what I can really deliver.
wikipedia-l@lists.wikimedia.org