User:Skierpage

From MediaWiki.org
Jump to: navigation, search
Although I work for the Wikimedia Foundation, contributions under this account are exclusively in my individual, personal capacity.

Long-time but non-hardcore Wikipedia editor. In 2012 I started work for the Wikimedia Foundation. I have edited under this account for many years as a volunteer. To avoid confusion, I (try to) do all my work as a Foundation employee under the account User:S Page (WMF).

"skierpage" also has wiki homepages at w:User:Skierpage, at ontoworld.org, and elsewhere. (Where's w:Project Xanadu's transclusion when you need it? My blog post "it's one social network, not 1000 sites" is apropos.)

Semantic MediaWiki

I'm interested in the project since w:Cyc hasn't turned into either Wintermute or Neuromancer from w:William Gibson. I want to believe in the glorious dream of adding semantics to Wikipedia.

Back in 2007 I installed Cygwin, the XAMPPLite bundle, eclipse, PHP/eclipse, MediaWiki, and SeMediaWiki (about 3M lines of code for free!!) and started hacking on simple things like Special page improvements and additional datatypes.

I implemented additional datatypes in SMW 0.4 and 0.5, and custom unit handling thanks to Markus Krötzsch's excellent work and insights; contributed more type work to development work SMW 0.7; and documented SMW 1.0.

I put my personal properties (and a bunch of tests) on my user page on Ontoworld SMW server.

Related résumé

I've written technical documentation, I've created Software Development Kits, and have been a Web site engineer (UNIX, perl, ColdFusion, Apache, etc.) for 7 years.

I built a corporate search system using Verity that used HTML meta tags for better filtering and summary presentation, and watched it founder because the metadata grows stale and unusable (users can't even get the <title> tag right, let alone a KEYWORDS value!), so Google's brute force page ranking appears to do a better job. Nevertheless, call me an optimist because I never metadata I didn't like.

I've developed "knowledge bases" in Lotus Notes, Perl, ColdFusion, and Kanisa. But strip away the big-dollar words like "taxonomy" and these were all just structured documents with searchable organization. I've integrated Web site search using Verity, Atomz, and Google.

I have a few projects and forks on Github, mostly using node.js