00:13 colomon joined
timotimo wow 00:19
i got some passing tests
this is incredible
and such a simple mistake, too
ok 1 - localref of type int with value 23 assigned to it 00:20
ok 2 - localref of type num with value 99e2 assigned to it
ok 3 - localref of type str with a value assigned to it
now i'll just have to write more tests to also cover lexicalrefs and make sure they work, too 00:22
at least in that direction
i probably also have to test the other direction 00:23
not 100% sure what the tests'll look like ... first i'd have a local set to some value, grab a localref from the local and access it as if it were a regular local again?
anyway. sleep first.
i pushed it to a branch "mast_localref_2" which is rebased on recent work, but dalek didn't report anything 00:25
[Coke] [6~[6~[6~/away 01:09
nwc10 jnthn: I think that I've found it. deserialize_method_cache_lazy() is cruel and unforgiving 06:46
07:00 lizmat joined 07:11 FROGGS joined 08:19 sylvarant joined 09:28 TimToady joined 10:34 FROGGS_ joined
timotimo found what exactly? 10:38
nwc10 why changing the size of stored strings made a lot more CPU work on startup 10:39
timotimo oh? 10:40
"the size of stored strings"?
11:00 sylvarant joined 11:27 colomon joined 11:56 colomon joined 12:39 colomon joined 13:51 zakharyas joined 15:00 colomon joined 15:38 colomon joined 16:28 colomon joined 16:35 Ven joined 16:44 zakharyas joined 16:59 colomon joined 17:08 tadzik joined
dalek arVM: a3c1c21 | usev6++ | src/io/syncpipe.c:
Loop over waitpid to get correct exit code for closed pipe
18:51
arVM: 8731d86 | FROGGS++ | src/io/syncpipe.c:
Merge pull request #209 from usev6/freebsd_pipe

Loop over waitpid to get correct exit code for closed pipe
19:15 zakharyas joined
[Coke] t/spec/S17-procasync/kill.t is passing, but slow. Any reason not to add it to spectest.data ? 21:36
wrong window. 21:39
22:41 colomon joined 22:59 lizmat joined