The site nominally exists to facilitate discussion of research issues. But it mainly gives me an excuse to keep up on various issues related to web design and computer programming more generally.
In designing the basic layout, my goal was to have high density of content and low degree of flash (literally and figuratively). I also aspired to have a nice looking site. I took the basic three column design from Matthew Levine's article on A List Apart.
The background image [unobscured] on most pages considers the fundamental question in economics. The image was produced in TeX using PSTricks; the TeX was generated with a little script that drew Greek letters uniformly—producing some runs of upper case Xi.
There is also a story behind the logo image in the upper left of the pages. This image arose as part of my research trying to understand the statistical bootstrap in the simplest cases. Pretty much reveals why the bootstrap is such a good idea!?
As an economist, I have been really amazed watching the evolution of web standards and the adherence to thereto. When I first started doing this stuff, browswers were a mess and the range of CSS hacks used to make sites work in various browsers was too wonderful for words. I used to be an obfuscated C fan, and it was fun seeing production code use those skills.
Now all browser developers seem to be stumbling over each other to do even a tiny bit better on litmus tests like the acid tests. When I say “all”, I obviously mean all but Microsoft.
As usual, there are some shenanigans required. I'm using the FontSpring approach to bulletproof specification of @font-face css and also a useful tip about how to specify a single font-family with multiple styles.
The current color palette is motivated (as is so much in my life) by Sponge Bob. The palette on that show (particularly for Squidward) is spectacular. Two images I like for their palettes are Squidward by Megan Beavers and [scroll down to squidward]. How I got from those images to the site palette, well that's hard to say.
As noted above, my real interest in doing the site is to think about issues of language design. The site is maintained in a language I wrote called GJ. Of course, this required study of intelligent language design, and given my Berkeley roots, I was naturally led to certain Davidsonian prior issues on language. This all led to GJ. It is currently undocumented and, hence, remains a private language. Coming to terms with the notion of a private language has further delayed the documentation process.