October 26th, 2010

ILC 2010 part 2

Day two of the International Lisp Conference I was actually feeling pretty ill and jet-lagged, so I unfortunately missed Marc Feeley's talk about Gambit Scheme. By all accounts it was interesting.

After lunch Peter Seibel gave a talk ("Common Lisp Standardization: The Good, the Bad, and the Ugly") about the context and history of Common Lisp's emergence as a standard. He talked about the pressure on the American Lisp community to not only consolidate their dialects to keep DARPA money flowing, but to beat the European Lisp community in the standardization process, enlisting Dick Gabriel and (indirectly) John McCarthy to subtly stall the European progress on what would become ISLisp so ANSI Common Lisp could be finished earlier.

Peter explained that a lot of the issues people sometimes take as accidental or ignorant misfeatures of Common Lisp were actually hashed over in agonizing detail. Case (in)sensitivity was contentious. The email discussion over NIL's role as symbol, empty list, and false stretched into the hundreds of messages. People didn't make arbitrary, ignorant decisions about important issues — they talked things out and decided to pick a particular approach and standardize on it.

Peter had his work cut out for him, because some of the people involved in the standardization process were in his audience, and wanted to augment or dispute particular points Peter raised. At one point audience members were in danger of discussing history completely amongst themselves.

A number of people didn't get exactly what they wanted out of the standardization process. But Peter pointed out that many of the people involved with creating or improving Common Lisp implementations today (and there are still more than ten actively developed implementations) had no involvement with the standardization process at all; they just found something appealing about the language and decided to pitch in and work on it. There's been a jump out of the community of original standardizers into a new generation of Lispers.

Peter wrapped up with a kind plug for Quicklisp as a possible mechanism for future consolidation and standardization of CL libraries and extensions. I hope I can justify his optimism.

I'm only scratching the surface of the things he covered, and I hope I havent misrepresented anything too badly. Peter is working on getting audio of his talk online so you can hear it for yourself. Stay tuned.

For dinner another dozen of us wound up at the Blind Onion for some great pizza. We got to hear about Peter's experience writing Practical Common Lisp and Coders at Work, and about some highly trained chickens. A good time was had by all.

More stuff tomorrow...