Last modified: 2011-08-02 01:57:20 UTC
The Wikimedia cluster is facing a lot of issues mostly related to the crapy MediaWiki 1.17 software. Although it came with lot of AMAZING features in the last few years, it comes with way too much bugs forcing to us to do smashbugging party in Berlin every year. I hereby request to downgrade to the good old 1.3 version which proven its success since the WikiMedia websites are now a major actor on the internet. Implementing this backport will solve most of the bug in bugzilla.
+1
adding URL to the archive + ccing our bugmeister since this bug has a HUGE impact.
How right you are! It was manufactured by some great committers such as vibber that no longer contribute. We need it back!
And who should re-open all these bugs which were solved during the years?
+1 It was made back when MediaWiki did not have such creepy commiters as I.
Revert the bugzilla database to its previous state!
Besides, it makes MediaWiki compatible with PHP 4.
(In reply to comment #4) > And who should re-open all these bugs which were solved during the years? Only if the report was about 1.3 or before.
This is probably the best way to handle the parser rewrite. No templates makes things so much more easy.
We do not need to revert the bugzilla database. We could just ask Brion to reopen the sourceforge bug tracking database. Easier to handle.
Going back to the cvs version would also make accessing that code simpler.
But is it compatible to HipHop?
Rather than bothering to keep maintaining wiki software ourselves, why not just use SharePoint? I'm sure we'll see a lower TCO as opposed to pouring time and money into developing a wiki engine using PHP.
Hey, why waste time and money on SharePoint at all? Just make some .doc files accessible via a network share.
(In reply to comment #14) > Hey, why waste time and money on SharePoint at all? Just make some .doc files > accessible via a network share. Plain text files.... Full cross comparability for the most part
Don't forget SneakerNet support for offline readers...
No templates... this is the obvious solution to the rendering speed issue! Why didn't we think of it before?? No new development needed, could be done in a day at most, in comparison to other alternatives which would take months or years.
Outdated.
That's the point, we need to stop using such new software.