Perl 6 - the future is here, just unevenly distributed

IRC log for #moarvm, 2017-10-19

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

All times shown according to UTC.

Time Nick Message
01:56 ilbot3 joined #moarvm
01:56 Topic for #moarvm is now https://github.com/moarvm/moarvm | IRC logs at  http://irclog.perlgeek.de/moarvm/today
03:40 evalable6 joined #moarvm
03:40 nativecallable6 joined #moarvm
03:40 greppable6 joined #moarvm
03:40 quotable6 joined #moarvm
03:40 bloatable6 joined #moarvm
03:40 unicodable6 joined #moarvm
03:40 committable6 joined #moarvm
03:40 releasable6 joined #moarvm
03:40 benchable6 joined #moarvm
03:40 coverable6 joined #moarvm
03:40 bisectable6 joined #moarvm
03:40 squashable6 joined #moarvm
03:40 statisfiable6 joined #moarvm
05:23 patrickz joined #moarvm
06:02 domidumont joined #moarvm
06:03 ggoebel joined #moarvm
06:46 nativecallable6 joined #moarvm
06:46 bloatable6 joined #moarvm
06:46 quotable6 joined #moarvm
06:46 coverable6 joined #moarvm
06:46 greppable6 joined #moarvm
06:46 benchable6 joined #moarvm
06:46 committable6 joined #moarvm
06:46 bisectable6 joined #moarvm
06:46 unicodable6 joined #moarvm
06:46 evalable6 joined #moarvm
06:46 releasable6 joined #moarvm
06:46 squashable6 joined #moarvm
06:46 statisfiable6 joined #moarvm
07:05 samcv good * everybody
07:12 domidumont joined #moarvm
07:49 lizmat samcv o/
08:11 zakharyas joined #moarvm
08:15 samcv hi :)
08:33 moritz ho
08:41 robertle joined #moarvm
08:44 lizmat jnthn: trying to grok the performance of 1.contains versus "1".contains versus 1.Str.contains
08:45 lizmat what I see is that "add_to_cache" is being called for every call to Cool.contains
08:45 lizmat (line 2481 in BOOTSTRAP)
08:45 lizmat and that can only happen if @possibles == 1 && pure_type_result
08:46 lizmat so why does it think it can cache, but doesn't cache ?
08:50 lizmat afk for a few hours&
09:17 jnthn lizmat: Wasn't that the flattening args thing?
10:19 Zoffix Yes
10:19 Zoffix <jnthn> It'd be better named try_add_to_cache :) https://irclog.perlgeek.de/perl6-dev/2017-10-09#i_15279031
10:26 Zoffix The 1.contains goes through .contains: |c call which isn't getting cached
10:29 * ilmari keeps getting confused between samcv here and smcv on another channel
10:30 lizmat Zoffix: indeed
11:37 lizmat jnthn: is there a reason Channel uses classes / nqp::istype for CLOSE/FAIL, rather than using an Mu.new and nqp::eqaddr ?
11:43 lizmat actually, only for CLOSE
12:05 jnthn FAIL 'cus we have to transmit the exception
12:05 jnthn CLOSE for uniformity with FAIL :)
12:05 jnthn It could be a constant
12:06 jnthn Could eqaddr on the type object too :)
12:09 lizmat ah, yes
12:09 lizmat lemme try
12:11 dogbert2 jnthn: didn't you fix https://github.com/MoarVM/MoarVM/issues/721 and https://github.com/MoarVM/MoarVM/issues/724 a few days ago?
12:13 jnthn dogbert2: I believe so
12:25 zakharyas joined #moarvm
15:06 zakharyas joined #moarvm
15:08 timotimo jnthn: would merging spesh_faster_shutdown be fine for inclusion in our upcoming release?
15:09 zakharyas joined #moarvm
15:10 jnthn timotimo: Does it actually help?
15:11 jnthn For true global destruction we'd need to do something more sophisticated I think (not that --full-cleanup works in general anyway)
15:11 timotimo it does
15:11 jnthn And for normal exit...we don't tend to have any foreground threads, so the exit goes pretty fast anyway?
15:11 timotimo before i added that the time for perl6 -e '' was mostly 0.1, afterwards it was mostly 0.09
15:15 timotimo i'm not actually sure why it helps :)
15:15 timotimo can't really callgrind it
15:16 timotimo that's also why i asked for others to test it
16:47 domidumont joined #moarvm
16:51 lizmat joined #moarvm
16:52 timotimo with the lack of testing and measurement i'd be totally fine with not getting this in the release
17:14 robertle joined #moarvm
17:52 ilbot3 joined #moarvm
17:52 Topic for #moarvm is now https://github.com/moarvm/moarvm | IRC logs at  http://irclog.perlgeek.de/moarvm/today
17:58 timotimo jnthn: it'd be cool if we could get a Promise constructor for "already kept/broken" conditions into this release
17:59 timotimo oops, ww
18:04 coverable6 joined #moarvm
18:04 bloatable6 joined #moarvm
18:04 quotable6 joined #moarvm
18:04 evalable6 joined #moarvm
18:11 AlexDaniel` joined #moarvm
18:32 evalable6 joined #moarvm
20:14 brrt joined #moarvm
20:19 brrt good * #moarvm
20:19 brrt my double-load-copy-replacement code is broken somehow :-(
20:22 brrt luckily the optimizer isn't merged yet
20:37 * jnthn suggests merging after the weekend release ;)
20:43 AlexDaniel` by the way, who is doing the moarvm release this time?
20:44 AlexDaniel` jnthn again? I have a feeling that someone else was volunteering
20:44 AlexDaniel` but I don't really remember now
20:44 jnthn I think samcv++ volunteered
20:51 jnthn AlexDaniel`: Do we have any outstanding blockers you're aware of?
20:51 AlexDaniel` jnthn: no
20:51 jnthn Alright. I'm heading away for some vacation on Sunday.
20:51 AlexDaniel` jnthn: brrt++ fixed the JIT issue and that's all I know right now
20:51 jnthn Great. :) brrt++
20:52 jnthn I'll probably have a few hours tomorrow if there's anything. Don't plan to do anything drastic, will write some tests for the stuff I did today though
20:53 AlexDaniel` OK. jnthn++
20:53 AlexDaniel` .tell samcv https://irclog.perlgeek.de/moarvm/2017-10-19#i_15327049
20:53 yoleaux AlexDaniel`: I'll pass your message to samcv.
21:32 xi- joined #moarvm
21:37 MasterDuke joined #moarvm
21:47 xi- joined #moarvm
21:50 xi- joined #moarvm
21:56 lizmat joined #moarvm
22:01 ilmari[m] joined #moarvm
22:28 AlexDaniel joined #moarvm

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