I've been tooling around on the php wikipedia, and it is very nice! Congratulations.
Great job with the statistics page, it'll be very useful.
The parser works well with the couple pages I imported, generally I'm very impressed.
Are you planning to post the code for this on one of the php.wikipedia.com pages as you mentioned in a previous e-mail. I'd be very interested in seeing it.
The one problem I see thus far is that UseModWiki allows subpage links on a subpage to other subpages of the main page. I know that's not the best description, but in the original wikipedia software a [[/subtalk]] link on [[MainPage/talk]] would lead to [[MainPage/subtalk]]. We may not want to emulate this, as it is certainly not intuitive, but there are a lot of pages, like poker, with subpages that link to other subpages -- either our parser needs to automatically translate these links or the PHP wikipedia should deal with them in the same way as UseModWiki.
Yours Mark Christensen
-----Original Message----- From: Magnus Manske [mailto:Magnus.Manske@epost.de] Sent: Saturday, August 25, 2001 3:30 AM To: Wikipedia-L@Nupedia. Com Subject: [Wikipedia-l] PHP Wikipedia, Part 2
Jason is activating php.wikipedia.com for the script to test, which should be working later today. So, soon you can flood me with bug reports ;)
Some points that were mentioned on the list while I was asleep: - Larry, I don't oppose CVS as such, I just thought why bother... So, I wouldn't mind a CVS at all.
- Edit locks : I thought they'd protect a page that is edited for a certain time, e.g., 5 minutes, so there won't be two edits of the same text at the same time. Now that I know it's only for writing, I am glad to not have wasted time in implementing such a thing in my script ;) The MySQL server will take care of the write-at-the-same-time problem, for sure.
- /Talk pages : Changing the standard text for new documents so they'd have a /Talk page should do it, right? I could also have the parser look for "/Talk" and append it if necessary in a "top-level" article.
- Conversion to SQL format : The easiest way I can think of is a script that goes through all articles in the current wikipedia and generates a complete article text in chronoligical order (oldest first). After each "version" is generated, a variant of my script can store it in the DB. That would ensure identical data. Anyone to write a "generation" script?
- Lame names : How about "Aide-Pikiw" (wikipedia spelled backwards)? That must be the lamest, for sure? ;)
Magnus
[Wikipedia-l] To manage your subscription to this list, please go here: http://www.nupedia.com/mailman/listinfo/wikipedia-l