libera/#commonlisp - IRC Chatlog
Search
1:18:33
nij_
Oh.. I should rephrase "I wonder how *usual* (i.e. non-CL) types are implemented then.. can they be implemented by the users?"
1:18:34
pjb
Note also that in lattices, there are elements that are not comparables (sets that are not subset one of the other).
1:33:46
jcowan
More broadly considered, in statically typed languages, types are the types of expressions, not only variables.
1:34:35
Bike
lisp types are attached to values rather than expressions or variables, which is part of why they're different
1:43:35
pjb
Because with change-class, a given object may have several different direct classes, at different times.
1:48:30
aeth
But if you do things that are so dynamic that they might break things, that's on you... if someone CHECK-TYPEs and you get around that, well, you're probably now breaking the assumptions of the program.
4:28:56
beach
nij-: It is types in other languages that are strange and complicated, yet restricted. They must be so that types can be determined statically. And they don't even succeed very well. Determining the exact type of an expression is undecidable in general, so those languages do what they can to make it decidable.
13:13:18
pve
Hey! Does "hook" refer to the variable to which functions (or symbols) are added or does it refer to the *functions* that are added to the list variable? I think I've seen it used both ways.