As for why we switched to MindTouch Deki, we did so primarily because MediaWiki is a hassle for us to maintain, and we think that over time, the WYSIWYG editor offered by Deki is likely to increase contributions to our codebase.

I think over the next few months as we expand what we’re doing with Deki’s capabilities, more people will see the advantages and understand better why we chose to make the switch.

Obviously I’m biased. I’m from MindTouch. But I’ll try to be as objective as I can. I can not agree with your assertion that MediaWiki syntax won. I do not believe any wikitext is a winner. XHTML/CSS wins. MindTouch Deki stores and operates on XHTML. This is the only approach that makes sense. Also, if you work with MediaWiki’s codebase you’ll quickly realize it’s a snarl of spaghetti. Extending that codebase is painful. Simply put, it’s brittle. MediaWiki has difficulty innovating because any minor change has sweeping ramifications across many millions of pages of content and they’re storing in plaintext (wikitext) and have evolved from a old codebase that was not implemented as an extensible platform.

I suspect some may find these statements an affront to the wonderful people who maintain MediaWiki. It absolutely is not. They’re wonderful smart people who are doing great work. Keeping Wikipedia running is a herculean endeavor. However, MindTouch Deki is an obvious choice for folks to upgrade MediaWiki to because: http://mindtouch.com/Technology

Here is a conversation at Dr. Dobbs with Shaver about the selection of MindTouch Deki: http://www.ddj.com/web-development/207600577

I just found support.mozilla.com is using yet another wiki, Tikiwiki. Individually these choices may make sense, but cumulatively this is ridiculous! MediaWiki syntax won, MediaWiki has the awesome extensions, MediaWiki is friction-free.

Also, can the Mozilla sites implement OpenID and/or single sign-on? I’ve got half a dozen logins I have to manage across *.mozilla.org and *.mozilla.com.