sergot hi o/ 08:42
nwc10 \o
diakopter interesting; instrumented/debug build of moar.dll,moar.exe,perl6_ops_moar.dll run the parse stage 4% faster than non-instrumented/debug 12:23
brrt o.O 12:25
consistently? 12:26
although 4% is far below the usual variation in stage-parse-time on my machines
lizmat does it spectest ok? 12:27
diakopter oh
dalek arVM: a924435 | jonathan++ | src/6model/serialization.c:
Don't try to serialize file handles.
15:21
arVM: 16bd593 | jonathan++ | src/ (3 files):
Provide mechanism to mark allocating extops.
arVM: b37f93e | jonathan++ | src/profiler/instrument.c:
Instrument allocating extops.
zakharyas Rohit-- in general 15:30
whoops 15:31
jnthn (wrong window)-- in general :P 15:32
hoelzro so I found a really "fun" precomp bug over the weekend: github.com/MoarVM/MoarVM/issues/142 15:41
I'm guessing that what's happening is that the Example namespace is loaded in the first module, so the lazy loader think it doesn't have any more work to do 15:42
brrt \o 19:07
timotimo o/ 20:09
brrt was your shirt too large or too small :-) 20:10
timotimo let's investigate the evidence 20:15
the shirt was produced in america, but i forgot to correct the size i ordered for that fact 20:16
brrt americans are not very large? (dutch perspective) 20:26
but they are ehm.... heavy, on averge
average
i think it's probably too large 20:27
lizmat brrt: don't worry, once you get older you will fill it and blame it on having it washed too hot :-) 20:29
brrt has had a very stable weight for the last 10 years though :-) 20:30
also, mine fits
(for now)
lizmat :-) be happy! 20:35
brrt i am :-) 20:36
timotimo brrt: have ideas for stuff i could try to work on for moarvm in the near future? :S 21:28
brrt no.. not now 22:09
sorry
i'm going to sleep now
timotimo gnite!
lizmat brrt is always away too fast to say good night :-)