3 ms·
SSA and the Lambda Tribe
- chalst 15y agoThe slides to Zadeck's keynote at the Static Single-Assignment Form Seminar, 2009, are worth reading for the history: http://www.cdl.uni-saarland.de/ssasem/ http://www.cdl.uni-saarland.de/ssasem/
- wingo 15y agoIndeed! I liked Wimmer's notes on SSA in HotSpot too, though for other reasons. Too bad there were no slides for the GCC experience report.
- contextfree 15y agoTo someone like me who's interested in compilers but doesn't have deep understanding of their workings, this is a really fascinating article! I appreciate the history, maybe should dig in to some of the linked papers ...
- bitdiddle 15y agogreat article! It reminds me of Curien's work on categorical combinators, where the machine definition is given by the equational axioms of the model, in this case the cartesian closed categories. I don't think McCarthy was informed by the lambda calculus either. I've heard his background was in functional analysis and he was motivated by symbolic differentiation when inventing lisp. I'd be interested to know this history better.
- pavpanchekha 15y agoYeah, his first papers and presentations on lisp already delved into basic pattern-based symbolic differentiation. What a brilliant guy...
- kenjackson 15y agoI skimmed this article, but it really deserves a nice bowl of cereal and a comfy chair to go with it. This looks like serious fun. More like this please!
- kenjackson 15y agowingo, is this line a typo? let t1 = f t1
- wingo 15y agoYep. Fixed, thanks.