00:31 vendethiel joined 01:57 vendethiel joined 02:18 jimmy_ joined 02:21 vendethiel joined 02:53 vendethiel joined 03:32 vendethiel joined 04:47 vendethiel joined 05:56 vendethiel joined 06:19 vendethiel joined 07:16 vendethiel joined 07:44 vendethiel joined 07:59 FROGGS joined 08:10 zakharyas joined 09:45 vendethiel joined
nwc10 mmm moar::cinclude is not a very useful config variable 09:45
09:45 FROGGS joined
nwc10 moar::cincludes= -I3rdparty/libuv/include -I3rdparty/libuv/src -I3rdparty/libatomic_ops/src -I3rdparty/libtommath -I3rdparty/dynasm -I3rdparty/dyncall/dynload -I3rdparty/dyncall/dyncall -I3rdparty/dyncall/dyncallback -I3rdparty/linenoise 09:45
JimmyZ_ wonders what nwc10++ is hacking
nwc10 because that's only useful if you're building MoarVM
10:52 danaj_ joined
jnthn nwc10: I don't think we're very selective about what makes it in there and what doesn't... 11:00
11:13 vendethiel joined 12:00 vendethiel joined 12:23 vendethiel joined 12:51 brrt joined
brrt \o 12:58
anybody break the JIT these days
jnthn Doubt my patches coulda :) 13:05
Not to mention I'm working in a branch :)
moritz brrt: would you please write a blog post for the perl 6 advent calendar? 13:09
brrt but but but
what would it be about
i know nothing
tadzik hmmmmmmmmm
moritz brrt: about your gsoc project 13:10
brrt yeah.. i could do that, but that wouldn't be much of a surprise right now
moritz brrt: sergot is also writing about his
brrt perl6 has a jit? old hat ;-)
moritz brrt: it's not meant to surprise anybody
tadzik but it will surprise some :) 13:11
brrt very well
moritz brrt: it's meant for people who usually don't read much about p6
brrt very well :-)
moritz we've had about 1k views per day on the first two days
brrt oh? not bad
moritz which is much more than the usual p6 blog post gets
brrt who is writing today
moritz github.com/perl6/mu/blob/master/mi...4/schedule says ab5tract 13:12
brrt cjfields was from bioperl, right 13:14
jnthn Yeah, advent calendar isn't about enws to us or clsoe followers, it's for the wider world :) 13:17
brrt ok :-) i can't do it before the 17th because of exams 13:19
jnthn Thankfully, advent lasts until 24th :)
brrt but in the week after that i should be able to manage something
jnthn OK. I'll do a more general MoarVM one before then 13:20
Will likely mention spesh, but not JIT if I know you're going to post on that :)
brrt what kind of detail would be suitable? 13:21
yeah, i probably will :-)
i'll probably not emphasize it's toyish nature too much 13:22
13:22 vendethiel joined
jnthn Right 13:23
And it's not exactly toyish.
When taken in totality with spesh.
I mean, it's doing speculative type-based specializaing (implying deoptimization), inlining, on-stack replacement, etc. :) 13:24
brrt in totality, no
that is true
jnthn Perl 5's VM does none of that. :)
brrt fair enough :-)
jnthn Put another way, just because we can see how much mroe is possible doesn't mean we should forget how much we've already achieved. :) 13:25
brrt that is true :-) 13:26
ok, i'll write a blogpost
(and hopefully one about the static-perl6-thingy, not that it's very useful but it is fun)
14:06 vendethiel joined
timotimo +1 14:21
14:31 vendethiel joined
lizmat hmmm... a MoarVM build from scratch now fails for me on OS X 14:43
14:44 FROGGS joined
lizmat In file included from 3rdparty/libuv/include/uv-unix.h:55: 14:44
3rdparty/libuv/include/uv-darwin.h:26:11: fatal error: 'mach/mach.h' file not
I guess I shouldn't have updated Xcode :-(
brrt fuuu 15:03
apple why u ...
15:15 vendethiel joined
lizmat seems to get fixed by adding a symlink to the "mach" directory of your OS X version to the libuv/include directory 15:22
both in nqp/moarvm as well as in install (for rakudo, that is)
brrt hmmmm 15:27
no, we can do better than that 15:28
15:34 dalek joined
brrt can't confirm it, by the way, but i'm not sure if my build tried to make reaclean 15:37
lord clang is slow
(realclean-ing now)
15:40 vendethiel joined 16:10 vendethiel joined
timotimo lord clang 16:26
timotimo bows
brrt lol 16:28
16:28 brrt left
TimToady actually, P5 does a small amount of inlining and constant folding 16:45
but it's not pervasive as it is in P6
17:02 vendethiel joined 17:31 vendethiel joined 17:53 FROGGS_ joined 17:57 vendethiel joined 18:11 vendethiel joined 18:34 woolfy joined 21:34 colomon joined
timotimo so, brrt's task was shanghai'd by nwc10? 21:53
22:58 colomon joined 23:26 vendethiel joined