02:48 ilbot3 joined 04:37 samcv joined 05:37 TimToady joined 07:20 brrt joined 07:27 domidumont joined 07:33 domidumont joined 08:22 zakharyas joined 08:53 domidumont joined 11:49 domidumont joined 11:56 domidumont joined 12:52 brrt` joined 13:15 ggoebel joined 13:21 domidumont joined
brrt to add to japhb and timotimo's answers: we don't use luajit per se, we use a piece of code that was developed together with luajit but has no 'system' relation to luajit 14:02
we use one of the tools that luajit also uses
and if you went the way of the LLVM, you'd first want to make sure that LLVM can actually generate good code, which means you have to (re)structure your VM/interpreter to simplify structure accesses 14:03
which is the *actual* difficult bit of having a high-performance VM
especially when you'd wish to combine that with 'portable' 14:04
14:04 domidumont joined
brrt and once you are there, then in all probability, LLVM or whatever backend will provide benefit in the order of percentages compared to what you could hack together yourself 14:05
so the interesting metric is, I think, 'system-complexity versus performance-gain-on-naive'
case in point: CPython want to minimize the system complexity part, and PyPy guys want to maximize the performance part 14:06
I guess we'd be in the middle of that 14:17
neither afraid of complexity nor racing headlong into it 14:20
15:10 domidumont joined
nebuchadnezzar hello 15:36
yoleaux2 31 Oct 2016 10:23Z <FROGGS> nebuchadnezzar: Is there anything else I can do for you? otherwise I'd leave it in this state: gist.github.com/FROGGS/9c06ac23b49...aa1687a43c
6 Nov 2016 20:40Z <FROGGS> nebuchadnezzar: can we somehow put /usr/share/perl6/site/bin in PATH?
6 Nov 2016 21:45Z <FROGGS> nebuchadnezzar: forget what I said :o)
15:39 FROGGS joined
FROGGS o/ 15:39
japhb o/ 15:40
[Coke] o/~ 15:42
japhb [Coke]: Waving a pennant? Rythm gymnastics? ASCII 8th note? 15:44
FROGGS mst / nine: thanks again for the help yesterday evening 15:45
15:45 hoelzro joined
mst FROGGS: did you get a resolution? 15:45
FROGGS mst: I was able to prove that that bug does not exist anymore
mst awesome
FROGGS (on a debian testing system)
[Coke] japhb: why not... armpit hair? 15:49
FROGGS I was thinking you were fighting a "laser" beam with your lightsaber 15:50
15:51 zostay joined
japhb [Coke]: Armpit hair almost as long as your arm? That's ... impressive? 15:53
15:54 zakharyas joined 15:55 geekosaur joined 15:57 domidumont joined 16:43 brrt joined
brrt damnit kids.... I had a whole bit about killing tile->template and I find I haven't even shaved that yak yet 16:43
japhb brrt: Please shave all the yaks. kthxbye. 16:46
;-)
brrt hehehe 16:47
so much to do, so little time...
at least I've implemented the heapifying logic
japhb .tell brrt No fair asking me questions via PM and then going offline. Ah well, ping me when you're back online again. 17:19
yoleaux2 japhb: I'll pass your message to brrt.
FROGGS nebuchadnezzar: hi, is there a reason that some bugs in moarvm/nqp/rakudo do not get closed, even though they are resolved? 19:16
nebuchadnezzar FROGGS: they are close only when a new package enter unstable with a changelog containing Closes #<bugID> 20:24
FROGGS ahh, so one cant close them just now? 20:25
sad
nebuchadnezzar FROGGS: they can be manully set to pending if the fixes are in repository waiting for new packge 20:28
FROGGS I see 20:29
well, then we have to wait for a new package to contain said Changelog lines
22:46 flaviusb joined 23:47 geekosaur joined