02:05 TEttinger joined 06:34 ShimmerFairy joined 06:52 TimToady joined 07:24 zakharyas joined
jnthn hoelzro: Indeed :) 07:46
07:54 rurban_ joined 08:51 brrt joined
brrt good morning :-) 08:55
jnthn o/ brrt 08:57
JimmyZ \o 09:00
brrt long work on GLR today? 09:01
i should really blog today 09:02
but i don't know what to write about
jnthn brrt: No, I got some $dayjob tasks and my YAPC::Asia talk to work at today 09:03
brrt you're going to YAPC::Asia?
jnthn Aye
brrt wow
your cumulative jetlag over the year must be enormous
jnthn One of the invited speakers, even. :)
brrt oh, very nice :-)
jnthn Yeah, I try to keep the longhaul down... 09:04
Yeah, I'd not be going if it wasn't for the invite, tbh.
jnthn really hopes his sleep continues to improve before he has to do the long flight... 09:05
brrt wow, the english of that 09:06
jnthn I...did something hard to parse? :) 09:07
brrt no, the yapc asia site 09:12
but, cool, good luck there 09:13
JimmyZ japanese? 09:29
brrt i think so 09:34
jnthn Aye 10:07
10:15 TimToady joined 10:17 ely-se joined
ely-se Does MoarVM have optimisations for immutable data structures? 10:17
jnthn ely-se: Anything in particular you're thinking of? I can't think of a great deal such things, although the GC is optimized for the use case where most objects quickly die. 10:20
Which is a good fit for regularly throwing away stuff 'cus you copied rather than mutated 10:21
ely-se Nice. 11:07
How about TCO? I can't find anything about it.
jnthn No, we don't do that one. 11:10
ely-se is there functionality to do it yourself? 11:11
like jumps to other functions
like perl5 goto
jnthn Not really. The best you'd manage at the moment is using the VM extension op interface to hack something up 11:13
It'd not be hard to all a tailcall op to the VM
*add
ely-se nice 11:14
I need TCO since my language guarantees it, so I guess I'll go that route 11:15
12:05 brrt joined
brrt hmm.. i have a feeling we should maybe want to clean up branches on MoarVM 12:14
wait, something horribly bad happened to the tiler table generator 13:13
jnthn Oh? 13:14
brrt yes, and i don't yet know what :-( 13:17
it doesn't generate all required rulesets anymore 13:18
not even for the small example grammar 13:20
hmmm 13:21
maybe it's because i uppercased the terminal names 13:22
probably because i uppercased the terminal names
brrt, you dumb-ass.... 13:23
arnsholt It's because you uppercased the terminal names? =) 13:24
jnthn OOPS 13:26
brrt yeah, pretty much
hmm
although
removing that doesn't quite fix the problem yet
jnthn Maybe there's some titlecasing too? <ducks> 13:30
brrt it's probably some naming issue, yes 13:31
hah 13:39
i removed a lc somewhere while debugging the earlier issue 13:40
dalek arVM/even-moar-jit: ebe7d4d | brrt++ | tools/tiler-table-generator.pl:
Don't uppercase terminal names so early

That interacts rather badly with table generation
13:42
brrt hmmm 13:44
i'm wondering how to deal with copy
hmm, bug found 13:57
14:23 brrt joined
brrt hmm 14:30
brrt thinking
bbl
14:44 ely-se joined 15:24 brrt joined 15:50 rurban_ joined
hoelzro article.gmane.org/gmane.comp.lang.l...eral/75426 18:06
more interesting articles!
19:07 colomon joined 19:49 raiph joined
raiph hoelzro++ # mike pall (LuaJIT) wow which led to www.reddit.com/r/programming/comme...64/c0ls6zr 20:15
21:16 colomon joined 21:24 TEttinger joined 22:08 oetiker joined 22:27 colomon joined