libera/#sicl - IRC Chatlog
Search
12:44:50
Bike
beach: trucler returns nil on no info. which is why i don't know why trucler even has condition types
14:15:54
Bike
i think it's better for trucler not to signal an error. the consequence of no information is ultimately up to the application, and many applications wouldn't want an error (like macro-function). and even if they do want an error, they probably want their own condition type, not trucler's
15:24:59
Bike
i mentioned a while back i was doing trucler for ecl - i'm trying it with the tests now. and, question: if the implementation actually does not track something, is it ok for the tests to pass? ecl seems to ignore dynamic-extent, so trucler can't accurately report that information
16:39:57
Bike
beach: is there more up to date documentation on clostrum than the paper? are there any systems using clostrum right now?
17:56:25
Bike
i am thinking about using clostrum for the bytecode/cross-compilation/sandbox stuff, in which case i can document as i go. i'm just trying to get a handle on how to use it