Search
20:28:16
kpoeck
Thats why I believe that the loop macro generates wrong code (and clasp complains correctly)
20:28:51
Bike
(loop-check-data-type data-type 'real), that's probably worng
20:29:43
Bike
yeah, if i change that to number it stops complaining
20:31:15
Bike
hm... maybe it ought to check for downto/upto/below/above though
20:31:35
Bike
oh, and by says it has to be a positive number, which implies being real
20:32:13
Bike
though i don't see why "by" needs that restriction... whatever
21:00:58
kpoeck
sbcl loop generates (DECLARE (TYPE NUMBER C)) instead of (DECLARE (TYPE REAL C))
21:01:43
kpoeck
thesame thing you proposed I assume
21:02:09
Bike
yeah but some more thinking might be required for upto and downto and stuff.
21:04:05
Bike
does clear up those tests, though
2:30:05
Bike
so yeah i did the function names thing and also increased the source information coverage in instructions, though it needs a sicl update
2:31:11
drmeister
Or is it still in the works?
2:32:22
Bike
i pushed some changes to clasp and some changes to Cleavir, but I haven't updated clasp's Cleavir commit since that's a bit of a pain
2:32:25
Bike
but i can do that tomorrow
2:57:43
beach
Good morning everyone!