01:17
MasterDuke left
02:45
tbrowder left
02:46
tbrowder joined
08:07
sena_kun joined
08:22
sena_kun left
|
|||||||||||||||||||||||||||||||||||||||
[Tux] | Sorry for that noise. I just realized that I posted that in the wrong channel | 08:50 | |||||||||||||||||||||||||||||||||||||
|
09:17 | ||||||||||||||||||||||||||||||||||||||
tux.nl/Talks/CSV6/speed4-20.html / tux.nl/Talks/CSV6/speed4.html tux.nl/Talks/CSV6/speed.log | |||||||||||||||||||||||||||||||||||||||
nine | Are grammars thread safe in general? | 12:57 | |||||||||||||||||||||||||||||||||||||
[Coke] | no.... Failed to write 21402732 bytes to filehandle: No space left on device | 13:09 | |||||||||||||||||||||||||||||||||||||
jdv | good job | 13:10 | |||||||||||||||||||||||||||||||||||||
[Coke] | after 62 hours. :P | 13:11 | |||||||||||||||||||||||||||||||||||||
jdv | after it went over half a day i think it was a lost cause anyway | 13:12 | |||||||||||||||||||||||||||||||||||||
lizmat | nine: grammars rely on $/ which has its thread safety issues? | 13:13 | |||||||||||||||||||||||||||||||||||||
[Coke] | jdv - it's 1/4 as big as the machine you run on | ||||||||||||||||||||||||||||||||||||||
lizmat | otoh... that would only apply if you start code inside an action method ? | 13:14 | |||||||||||||||||||||||||||||||||||||
jdv | oof | ||||||||||||||||||||||||||||||||||||||
[Coke] | also, this is at least something I can easily throw disk at. | ||||||||||||||||||||||||||||||||||||||
got this much: gist.github.com/coke/d0423c8c6dd2d...65ba5a14fd | |||||||||||||||||||||||||||||||||||||||
Do you have as many as 931 "AlwaysFail" ? | 13:15 | ||||||||||||||||||||||||||||||||||||||
jdv | looks like most of it | ||||||||||||||||||||||||||||||||||||||
its very roughly 50/50 iirc, so yeah | 13:16 | ||||||||||||||||||||||||||||||||||||||
github.com/rakudo/rakudo/issues/5584 | 13:17 | ||||||||||||||||||||||||||||||||||||||
[Coke] | Thanks | 13:21 | |||||||||||||||||||||||||||||||||||||
jdv | a lot of abandoned dists? idk. someone could look into it;) | 13:22 | |||||||||||||||||||||||||||||||||||||
Geth | Blin: 8e900ea2c2 | (Will Coleda)++ | 3 files Convert PERL6LIB to RAKULIB As Blin is used to test recent Rakudos, we want to avoid triggering the deprecation warning Part of #34 |
||||||||||||||||||||||||||||||||||||||
Blin: 82bec79885 | (Will Coleda)++ | lib/Blin/Processing.rakumod Remove unused code Closes #34 |
|||||||||||||||||||||||||||||||||||||||
[Coke] | doubling VM sizes again... | 13:27 | |||||||||||||||||||||||||||||||||||||
Also what was the actual commit you ran Blin on last time? I can use that as the start point. | |||||||||||||||||||||||||||||||||||||||
(the issues was for a few releases ago) | 13:28 | ||||||||||||||||||||||||||||||||||||||
(that issue above, I mean) | |||||||||||||||||||||||||||||||||||||||
(ah, would we not have saved it if there were no blin issues?) | 13:33 | ||||||||||||||||||||||||||||||||||||||
jdv | uh, i threw it away:) oh wait. it was on the bump i did for blin. see the last bump fore the release. | 13:41 | |||||||||||||||||||||||||||||||||||||
[Coke] | (trying to resize disk at the moment.) | 13:42 | |||||||||||||||||||||||||||||||||||||
nine | I have definite thread safety issues in what should be a relatively straight forward compiler that's offered as a service throu Cro. Can't find any globals or otherwise shared data in use but the problems do go away when I wrap the compile call in a global lock. | 13:55 | |||||||||||||||||||||||||||||||||||||
lizmat | meh | 13:56 | |||||||||||||||||||||||||||||||||||||
feels like a visit to NQP land is warranted then | 13:57 | ||||||||||||||||||||||||||||||||||||||
nine: could you provide a golf ? | |||||||||||||||||||||||||||||||||||||||
nine | I was gonna write that if I could golf, I could problably fix it myself. But that's actually worth a try. E.g. replacing Cro with a loop and start | 13:59 | |||||||||||||||||||||||||||||||||||||
timo | with some wrapping of subs and methods, or replacing of $/ variables with proxies, or something like that, surely we can get some interesting information out of such an experiment | 16:37 | |||||||||||||||||||||||||||||||||||||
the patch that makes Match.Bool no longer travel through an assignment to $/ and back will have been very important, surely also for grammars? | 16:50 | ||||||||||||||||||||||||||||||||||||||
or do grammars never call Match.Bool ... would probably be better from a performance standpoint honestly | |||||||||||||||||||||||||||||||||||||||
jdv | 7 | 16:51 | |||||||||||||||||||||||||||||||||||||
[Coke] | moved some things around on the blin box, now getting a zef error (moved blin install from ~/Blin to /data). /home/azureuser/raku/bin/perl6 -I data/zef data/zef/bin/zef --config-path=data/zef-config.json update | 16:53 | |||||||||||||||||||||||||||||||||||||
^^ that now updates the mirrors, but dies with "Must specify a non-empty string as a path" | |||||||||||||||||||||||||||||||||||||||
guessing this is a result of running out of disk on the previous run. | 17:04 | ||||||||||||||||||||||||||||||||||||||
yup, clearing out the old blin/data folder got there. | 17:07 | ||||||||||||||||||||||||||||||||||||||
OK this run looks like it's on track to take less time, we'll see if it starts slowing down. :| | 17:51 | ||||||||||||||||||||||||||||||||||||||
23.5% after 1h16m - that's much better. | 18:18 | ||||||||||||||||||||||||||||||||||||||
19:00
finanalyst joined
|
|||||||||||||||||||||||||||||||||||||||
finanalyst | lizmat: I just noticed that in your Highlight module you have 'use v6.e.PREVIEW' and 'use experimental :rakuast'. What is the purpose of the former? | 20:12 | |||||||||||||||||||||||||||||||||||||
[Coke] | are you asking in general? (It pulls in languages features from the next release of the language, 6.e, that aren't available in the default version of the compiler, 6.d) or which of those features specifically she needs? | 20:27 | |||||||||||||||||||||||||||||||||||||
jdv | [Coke]: good luck! | 20:28 | |||||||||||||||||||||||||||||||||||||
20:37
sena_kun joined
21:22
sena_kun left
23:55
finanalyst left
|