Perl 6 - the future is here, just unevenly distributed

IRC log for #moarvm, 2014-09-30

| Channels | #moarvm index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
07:02 njmurphy joined #moarvm
07:04 FROGGS joined #moarvm
07:06 zakharyas joined #moarvm
08:06 kjs_ joined #moarvm
12:52 brrt joined #moarvm
13:31 * jnthn heads home tomorrow morning :)
13:31 jnthn So the tuit supply should improve soon :)
13:32 FROGGS \o/
13:58 [Coke] jnthn: ++(welcome home)
14:12 jnthn Yes, it'll be nice ot see it again :)
14:14 Juerd To whoever upgraded bash on feather: thanks :)
14:15 Juerd Only today I realised I was forgetting these.
14:16 jnthn .oO( at least it was easy for pretty much anyone to get in and do it... ;) )
14:17 Juerd Nah, I don't think anything with root was susceptible for that.
14:18 Juerd Oh, or perhaps something was. bash is still the system shell on feather
14:18 Juerd Let's fix that. rm /bin/sh; ln -s dash /bin/sh
14:19 Juerd feather's going away soon, by the way
14:19 Juerd What's the best mailinglist to announce this?
14:20 jnthn Juerd: #perl6 is the best IRC channel...many more folks there than here...
14:20 FROGGS and #parrot?
14:20 jnthn Ah, true
14:21 jnthn Of mailing lists, I guess perl6-compiler and perl6-language can't hurt.
14:21 Juerd Okay
14:22 Juerd My current plan is to switch it off in January and provide a new VPS specifically for those who need it for IRC.
14:24 Juerd -rw-r--r-- 1 root root 8 May 25  2005 /etc/hostname
14:24 Juerd Or maybe I shoud keep it running until May 25 2015 so that we can say it ran for 10 years.
14:24 Juerd :P
14:43 JimmyZ joined #moarvm
14:52 FROGGS[mobile] joined #moarvm
14:52 * JimmyZ needs it for irc ...
15:09 moritz Juerd: fwiw I think I updated bash on all feathers, but I have lost overview :-)
15:10 jnthn Time for some rest...long day tomorrow... &
15:12 Juerd moritz: Thank you :)
15:13 JimmyZ joined #moarvm
15:15 JimmyZ_ joined #moarvm
15:41 FROGGS joined #moarvm
18:25 kjs_ joined #moarvm
18:49 Ven_ joined #moarvm
19:25 kjs_ joined #moarvm
20:21 kjs_ joined #moarvm
20:21 brrt joined #moarvm
20:21 brrt \o
20:22 brrt i've calculated the average (geometric mean) speedup caused by moar-jit in it's first release was 30%
20:23 FROGGS nice!
20:23 FROGGS I had guessed that :o)
20:27 odc` joined #moarvm
20:30 brrt well, i had not. i had imagined there were too many benchmarks on which there would not be a lot of benefit
21:03 itz_ joined #moarvm
21:07 Ven joined #moarvm
21:08 itz joined #moarvm
21:08 ggoebel111118 joined #moarvm
21:12 timotimo so ...
21:12 timotimo we'll want to make a thread not only block but ideally do work for another thread when it encounters a promise that another thread has created, right?
21:12 timotimo but what would that look like?
21:12 timotimo we can't really longjmp into the other's frame or something
21:13 timotimo and we also can't just call into the other piece of work from await or something similar
21:14 brrt no... that's basically green-threads-over-os-threads
21:14 brrt let your kernel deal with that
21:14 brrt make more threads instead
21:14 timotimo mhm
21:14 brrt we live in 64 bits these days
21:15 timotimo so make 2**64 threads
21:15 brrt all sorts of hacks aren't really all that necessary anymore with sufficient address space :-)
21:21 brrt that said i don't know how / when threads are created in moarvm
21:29 timotimo i suppose the scheduler code in perl6 land has some nqp op that spawns a thread and gives it a code object to run
21:30 timotimo nqp::newthread
21:34 brrt i see
22:04 brrt left #moarvm
23:59 daxim joined #moarvm

| Channels | #moarvm index | Today | | Search | Google Search | Plain-Text | summary