9:01:55beachNo biggie. It is just very convenient for me. When someone posts a link, I go to my desktop (this is my laptop), whip up the logs and click on the link.
16:55:05beachFor the EDU calculation, I was considering using alists with lexical locations as keys and EDU as values, but that makes it somewhat expensive to do the <= comparison, depending on how many variables are live at any program point of course. Should I use hash tables instead? They probably take up a bit more room, but I can do one function at a time, so that should not be a problem.
16:55:42beachDo we have an idea of the (say average) number of live variables at any program point?
16:57:34beachI could sort the alists based on hash value of the lexical location I suppose. Or I can just do the quadratic thing.
17:03:39beachI guess I can use alists, and stick in a meter for the length.
17:04:02beachEither way, I should abstract out the representation.
17:09:51splittistAnd make sure you use one-letter vrbl names