Starting from scratch. Looking around the "tools for building new languages" ecosystem to see what peopel use. Encountered Racket. Noted Airflow for visualising workflows. Will work towards a sort of env where I can see in one shot (for any new lang) :
- { universe :
- [ runtime structure ] <> [ LL compilation targets ] <>
- [ "mylang" structure ] <> [ IR structure ] <>
- [ parsers + lexers ] <> [ syntax structure ]
- }... to form a map of economics/tradeoffs for the universe.
Then I can get some sense of how existing languages fit in, before I try to think of new language decisions. But if the map/tool is robust, it should form some sort of fabric that allows automated testing of hypotheses about ( consequences of syntax decisions, at runtime, at scale ).
No comments :
Post a Comment