00:57 Summertime left, sivoais left, camelia left
Geth rakudo: vrurg++ created pull request #4130:
Make enums return Failure on bad conversion
01:16
01:17 Summertime joined, sivoais joined, camelia joined 02:07 lucasb left 03:21 cog__ joined 03:24 cog left 04:21 leont left 05:21 frost-lab joined 07:04 Altai-man joined 08:25 domidumont joined
AlexDaniel` ah crap, just a few days short of 2021 08:52
so, I've seen in some science-denying circles that when people are told the numbers about total covid cases they don't believe them because, well, they don't have anybody close to them who actually had it 08:53
so, yeah, got a positive result today :S AMA
goddamn 2020 08:54
nine AlexDaniel`: oh, sorry to hear that :( 08:57
AlexDaniel` I'm feeling OK-ish today, no cough at all for now 09:00
btw not suggesting that anyone is denying anything here. Just in case. 09:05
09:09 sena_kun joined 09:11 Altai-man left
Geth rakudo: bef6b4f69e | (Stefan Seifert)++ | src/Perl6/Actions.nqp
Fix use of undeclared variable in &Perl6::Actions::lower_signature

Contrary to most methods in Actions, lower_signature actually doesn't have a $/ argument. Compilation of Perl6::Actions doesn't fail however as the lookup of this lexical is deferred to runtime and the runtime has an odd exception for object lexicals and just returns VMNull instead of raising an exception.
10:36
dogbert17 AlexDaniel`: get well soon 10:40
nine: do you have an early review of your new system? Good, bad, noisy ... 10:41
nine Cut down stage parse from ~52s to 30s. Spec test is done in just 77s. Even feels a bit snappier in general desktop use. So far system is running flawlessly. Still have to adjust the fan curves a bit, so they stay silent for short term load. When idle it's already totally quiet 10:45
sena_kun nine, what's the cooling, I assume NH-D15? 10:47
nine correct 10:48
sena_kun nine, yay, this one is just awesome. Glad for you, have fun. :) 10:49
dogbert17 what have you set TEST_JOBS to, 40?
nine 35 seems to be the sweet spot according to my first test runs 10:50
sena_kun Isn't it good to just set number of threads available, i.e. 32 or 24? 10:51
I am running with 24, should I try higher someday?
nine In general that's a good starting point, yes. Using a little more than that can still help because some threads may be blocked on IO
sena_kun Interesting point, thank you. 10:52
nine But then, it always is subject to experimentation. The above advice comes from a time when that meant that you'd run make -j3 and IO was done on spinning disks 10:53
dogbert17 I guess we can assume you're using NVME-drives now :-) 10:54
nine Then ther's other factors as well. Building MoarVM is faster with make -j90 because of the (comparitively) very long time files like interp.c take. So you benefit from starting them as soon as possible. 10:55
Geth rakudo: f0d64ea5e7 | (Moritz Lenz)++ | docs/ChangeLog
Changelog: fix typo. bbkr++
11:07
sena_kun Ugh. 11:16
moritz++
11:39 Kaiepi left 11:40 Kaiepi joined
lizmat bisectable6: dd <a b c d e>[0,(1,2,3),4]:k 12:02
bisectable6 lizmat, Will bisect the whole range automagically because no endpoints were provided, hang tight
lizmat, ¦6c (50 commits): «(0, 4)␤» 12:03
lizmat, Nothing to bisect!
lizmat so, incorrect from the very beginning and always 12:04
Geth rakudo/faster-slice-access: eaf54efbe5 | (Elizabeth Mattijsen)++ | 4 files
Re-imagine @a[Iterable]:adverbs

This commit makes array slice accesses about 3x as fast *and* fixes several issues found in the previous implementation. It also creates a framework for handling adverbs that are allowed with postcircumfix adverbs (in a separate source-file Rakudo/Internals/PostcircumfixAdverbs).
... (14 more lines)
12:06
rakudo: lizmat++ created pull request #4131:
Re-imagine @a[Iterable]:adverbs
timotimo whoa 12:12
12:22 MasterDuke joined 12:26 frost-lab left
MasterDuke AlexDaniel`: hope it's an uneventful recovery 12:28
12:45 Kaiepi left 12:46 Kaiepi joined
timotimo oh shit. hope you'll do fine AlexDaniel`; i hear there's a genetic component to whether you'll have a harmless experience, i hope you're lucky in that way or another 13:03
lizmat also: it has been proven that if you're Vitamin-D deficiient (which most IT people are vulnerable to due to not getting out in the sun much), you have a greater risk of complications after infection 13:07
so please take your vitamin-D3 pills !
13:08 Altai-man joined 13:11 sena_kun left 13:31 Kaiepi left, Kaiepi joined 13:55 leont joined
vrurg AlexDaniel`: wish you get over it without consequences. Get well soon! 14:14
jdv79 AlexDaniel`: Hope its not too bad for you. I had it back in april; it wasn't fun. 14:23
lizmat notable6: weekly 15:18
notable6 lizmat, 4 notes: gist.github.com/d9ac7349591022b164...bf8de70237
lizmat notable6: weekly reset 15:19
notable6 lizmat, Moved existing notes to “weekly_2020-12-21T15:19:21Z”
16:55 Xliff joined
lizmat and yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2020/12/21/2020-...releasmas/ 17:04
Xliff lizmat++ 17:07
Is there a way I can force hyper to use all cores? 17:08
I;ve been attempting to write code all weekend that uses a heavy computation and rakudo only uses at most, 2-3 cores. I have 20.
lizmat degree => cores
Xliff Yep. That's what I suspect.
lizmat .hyper(degree => 20). 17:09
Xliff However even when degree is used, it still uses only 2 cores.
Doesn't hyper default to $*KERNEL.cpu-cores, anyway?
17:09 sena_kun joined
lizmat then maybe your batch size needs to be adapted: if the batch size (default 64) is too small, there will never be a reason to start up additional threads 17:09
batch => 1024 maybe ?
Xliff That's what I thought, too. I can bump the batch size to 256k, and there still isn't enough load to bring up more than an additional core. 17:10
17:10 lucasb joined
Xliff Also... memory leaks. 17:10
Well...apparent ones. Which is why I don't use a higher batch. 17:11
lizmat Xliff: weird
Xliff Indeed.
17:11 Altai-man left
Xliff So I will have to run things through valgrind after work, and hope I can decipher the output to make a report. 17:11
lizmat is going to be afk for the rest of the day, so won't be able to help you with that immediately
++Xliff 17:12
Xliff OK. Was hoping I could get a quick answer.
So I guess my computation wasn't as heavy as I thought??
lizmat could be?
Xliff Fair enough. Thanks, lizmat! 17:13
Geth ¦ rakudo: lizmat self-assigned my %hash; my %n = |%hash<>:p.map({ .key => .value }); cause moar panic github.com/rakudo/rakudo/issues/4132
AlexDaniel` yeah, I've been taking vitamin D regularly for a long time
still no cough at all, I probably have the mildest symptoms of the 3 people in our “circle” or whatever you call that 17:15
I think I feel more of an immune response than anything caused by the virus. Joint pain is pretty annoying, but, if I'm understanding correctly, that's not caused by the virus but by the immune system itself. 17:18
I'm not a doctor so don't quote me on that though :)
I guess I'm more sad than sick, was hoping to get at least through 2020 without getting it 17:21
Thank you for your kind words, everyone!
17:41 domidumont left 17:42 Xliff left 17:44 b2gills left 17:45 b2gills joined 17:47 domidumont joined 17:50 domidumont left 19:04 Kaiepi left, Kaiepi joined 19:59 MasterDuke left 20:05 MasterDuke joined 20:25 Xliff joined 21:04 vrurg left 21:08 Altai-man joined 21:12 sena_kun left 21:28 Xliff left 21:46 vrurg joined 22:01 Altai-man left 22:55 leont left 23:03 leont joined 23:22 SmokeMachine left