00:18 jnap joined 00:59 jnap joined, colomon joined 01:11 btyler joined
dalek arVM: e995699 | Carlin++ | src/io/ (2 files):
make default mode 0666 per pmichaud++
01:16
arVM: d2e0732 | jimmy++ | src/io/ (2 files):
Merge pull request #106 from carbin/master

make default mode 0666 per pmichaud++
01:49 woolfy joined 02:41 cognominal joined 02:55 raiph joined 05:08 raiph joined 06:23 woolfy joined 06:34 lizmat joined 06:58 FROGGS joined 07:04 zakharyas joined 07:08 lizmat joined 07:12 zakharyas1 joined 07:18 lizmat_ joined 07:21 lizmat joined, woolfy joined 07:41 lizmat_ joined 07:58 odc joined 09:00 lizmat joined 09:12 donaldh joined 09:25 brrt joined
brrt jnthn here? :-) 09:25
.tell jnthn that i hadn't considered that, that it is a good point (i.e. the inliner needing the effective bytecode, which is impossible if i've replaced it with magic jit bytecode) 09:26
and... that he's right and that i'll fix it, or at least that i think that what i think he meant is right 09:28
nwc10 everything apart from t/spec/S17-lowlevel/lock.rakudo.moar happy at master/master/nom 09:59
10:11 tokuhirom joined 12:03 brrt joined 13:10 FROGGS joined 13:23 btyler joined 13:26 jnap joined 13:45 btyler_ joined 15:25 raiph joined 17:18 woolfy1 joined 20:18 cognominal joined
dalek arVM/moar-jit: 51f3ea5 | (Bart Wiegmans)++ | / (8 files):
Added simple return value handling.

This won't do everything that the moarvm interpreter can do, but it doesn't have to. Ops that are really too big too JIT - or perhaps too likely to change in implementation - should be single function calls.
21:33
22:16 btyler joined