A few months ago, one of the instructors I serve while wearing my webmaster hat asked me to attach a wiki to his class’ website. I’d been meaning to start testing various wiki systems for a while, and this was the perfect opportunity to dig in.
The only wiki I had previously installed anywhere was Wakka. The experience had gone very well - easy to set up, did most of the things I needed it to do, was fairly easy to customize. But when I went looking for it again, was disappointed to find it had become open-source roadkill - an abandoned project without so much as a homepage remaining to memorialize it.
Being partial to PHP/MySQL solutions, I headed for phpWiki. At first blush, phpWiki seemed like a great replacement with a solid development community behind it. A few installation glitches, and some confusion over whether to use a provided wizard for installation configuration or to edit the configuration text file, but once up and running, I was fairly satisfied, even though it didn’t support linking of Chinese characters to full Chinese URLs (yes, this is possible! However, it is not possible to paste such URLs into the O’Reilly weblog back-end without them breaking, so you’ll have to click a link from within the Chinese wikipidia once there to see what I mean).
Months passed before the class was ready to rock with their wiki. Meanwhile, I worked on other projects, including an upgrade to PHP5 (required by a CMS I’m testing for another project). Finally, the time came to prepare for launch. Checked back in with the wiki, only to find it completely broken - turns out the current version of phpWiki is incompatible with PHP5. Worse, there were rumblings in the phpWiki user forums that the project was starting to lose momentum and direction. No new downloads had appeared on the SourceForge site since last March. Another great open source project left to rust by lack of direction/enthusiasm.
That left me scrambling, with a week to find a replacement. Checked out MoinMoin (obscure installation procedure left a bad taste in my mouth) and TikiWiki (everything-including-the-kitchen-sink portal software that just happens to include a wiki module — I want a streamlined, dedicated app). And today TikiWiki’s homepage is inaccessible.
Finally, the light went on. Since we were using WikiPedia as an inspiration for the class project, why not use what they were using? Duh. MediaWiki is the platform on which the incredible and vast WikiPedia is built. Found most of the links to the download broken at the SourceForge mirrors, but eventually found one that worked. Installation, unlike many of the wikis I sampled, was a breeze.
MediaWiki is a great platform, and does support full Chinese URLs, but its documentation is scattered and obtuse. Even simple things like figuring out where the administrative access page for the system is, or which files need to be edited to customize the system (the “templates” directory contains only one file, which does not appear to be template related) were stupidly difficult. The problem wasn’t confined to administrative documentation - the system didn’t come with any user-level documentation either — I had to copy/paste help pages out of WikiPedia’s wiki into our own just to provide assistance for our users. Ridiculous.
Yes, I know - if I don’t like the state of an open source project, I can help fix it. And I’d love to. But I just. Don’t. Have. Time. I don’t mean to look a gift horse in the mouth here, I’m just saying that the public distribution didn’t feel quite ready for prime time.
Overall, I was left with the overwhelming impression that the wiki world desperately needs its own Movable Type - a system that’s cleanly designed, clearly and logically documented, easy to customize, has momentum and traction, an expansive and supportive user community… I’m predisposed to open source solutions, but would certainly pay a small/reasonable sum for a commercially supported, high-quality wiki (our .edu budget doesn’t leave us piles of money to throw around - I’m talking $100 here, not thousands).
On a side note, I’m rapidly coming to the conclusion that wikis aren’t necessarily the best place to provide documentation — yes, they’re easy to contribute to, but they tend towards meandering, non-hierarchical layouts and can be rough for end-users to follow.
Yes, there are wiki services out there we could take advantage of, such as SocialText and SeedWiki, but I’m committed to keeping all school content hosted on school servers - I’ve been burned going down that road before when services went belly up.
It seems that the wiki universe is being built by extreme geeks. Which is great - that’s where all the good stuff comes from. But at a certain point, a good software project needs the kind of vision that will make the product more accessible, more polished, less prone to abandonment or neglect. The blogging world got there a while ago. It’s time for the wiki world to catch up.
What great wiki opportunities am I missing here?