Search
11:49:22
stassats
now i need to write read-evaluated-form-of-type
11:50:23
scymtym
what will CONTINUE do, use NIL?
11:56:09
stassats
well, it's for non-existent packages, so, *package*
12:41:53
stassats
Enter a form to be evaluated (of type LIST):
12:42:09
stassats
will people confuse it and think that the form should be a LIST, not the result?
12:42:22
stassats
Enter a form to be evaluated (returning a value of type LIST):
12:42:58
stassats
Enter a form evaluating to a value of type LIST:
14:02:59
stassats
do we really need a restart for (intern "x" :a)? as it will imperceptibly slow down INTERN
14:16:37
stassats
huh, restart-case conses
14:31:11
stassats
but restarts should be even cheaper
14:31:47
stassats
i'm descending into yak shaving
14:33:01
stassats
there can be a prefabricated block of restarts, now how to achieve that
14:39:31
stassats
i think it can be done relatively painlessly
15:11:17
pfdietz_
It's not clear to me why adding restarts should have any cost for the non-error case.
15:11:33
stassats
well, how else would they work?
15:12:13
stassats
ok, that's only when you wrap the error directly
15:12:18
pfdietz_
You only create the restart(s) when you are going to signal the error?
15:12:42
stassats
but if you wrap around code that might produce an error, it's not zero cost
15:12:56
pfdietz_
Right. So don't do that.
15:26:00
pfdietz_
What's the overhead for handler-case?
17:49:01
karlosz
so i think im running into the cheneygc + 64bit thing again with the code component not being recognized by ICF
17:49:22
karlosz
is the code folding supposed to work on all platforms?
18:13:56
karlosz
is doing that reprecussionless? then trying to save lisp and die gives me *thread-initial-bindings* is unbound
18:14:02
karlosz
im not building with threads, so im stumped
18:14:42
karlosz
specifically when trying to coalesce-ctypes
18:14:48
karlosz
okay, thats good to know
18:22:19
stassats
what says *thread-initial-bindings* is unbound?
18:23:15
karlosz
https://paste.gnome.org/pbvenojb8
18:23:22
karlosz
i get this backtrace when save-lisp-and-die is run
18:36:56
stassats
just remove coalesce-ctypes from deinit
18:40:49
stassats
or don't and fix fast-symbol-global-value
18:43:25
stassats
try (define-vop (fast-symbol-global-value fast-symbol-value) (:translate sym-global-val))