freenode/#clasp - IRC Chatlog
Search
15:27:33
Bike
in the compiled dtree, getting the stamp and branching based on the where tag are all in the same instruction, since you never need to do them separately anyway
15:31:19
Bike
by the way, we might be able to differentiate classes on wrapped/derivedness pretty easily? it looks like wrapped object classes are of class core:clbind-cxx-class, and derived object classes have a core:derivable-cxx-class in their class precedence list
17:24:35
drmeister
Ok - how does that help? Would it be better to confine the wrapped stamps and derivable stamps to separate ranges?
17:25:06
drmeister
Let me put it this way. If there would be an advantage to confine those stamps then I could arrange the runtime to do that. It would take some work though.
17:30:24
drmeister
Great - that saves me a bunch of work. I also don't like the idea of putting limits on stamp ranges for certain kinds of objects.
17:31:10
Bike
the point is having some way to tell the _classes_ apart. at compile time where we can be a little slow, too
21:25:40
Bike
i'm getting what is i think a recursive C++ exception from a bad format string - i think we had that in here earlier