June 7th 2004 #
random, June 7th 2004
ICFP last weekend. I took part teaming up with my brother: he wrote the simulator in Java and I made a symbolic assembler for the ant-language. I'm not too enamoured or impressed with our entry (the ant we ended up with is pretty basic), but it'll be interesting to see how it goes. Unfortunately we missed the lightning entry submission due to oversight: we had a few working atns by then, but just didn't get around to making the submission.
In other news, SBCL 0.8.11 has been tagged, which means the release is imminent, which in turn means it's time to for me to get ready to merge package locks to the main tree. Which in turn means that hopefully little active hacking on them remains, and I can start thinking about linkage table again.
On a tangent, I chanced to take a look at STEP in the CMU AI Repository a small while back, and to me it seems that the easiest way to give SBCL a nice STEP implementation starts by writing a real interpreter instead of the current "punt almost everything to compiler" approach — which as an upside could also give "dynamic" macro behaviour for interpreted code. Well, it's something like TODO item number 8739812, so there's a good chance that someone gets there before me. ;-)
Christopher Rhodes along with a few other smart people has begun work on a test suite for IEEE 754 conformance (that's the standard for floating point arithmetic) for Common Lisp. Cool stuff. I really should find time to read What every computer scientst should know about floating point arithmetic. Sigh — so much to read, so little time.
I got around to trying nxml-mode for Emacs (for editing this), and so far indenting seems much saner then what sgml-mode used to want. Still, I probably should generate the page and keep the entries elsewhere, but haven't bothered so far — and am not likely to do so too soon.
Daniel Barlow has stuff to say on habitable software, and I hope he continues to do so in a loud voice.