Geth roast: de0ec138d7 | (Vadim Belman)++ | integration/advent2010-day03.t
Make sure that test current directory is roast root

Because it is testing against some roast-specific files.
00:03
rakudo: vrurg++ created pull request #3696:
Update INSTALL.txt for the latest changes
00:28
01:04 AshlyeGraam joined 01:06 AshlyeGraam left
[Coke] vrurg++ - keep install.txt simple 01:13
[Tux] MasterDuke, not really (for smokes): 06:46
Rakudo version 2020.05.1-121-gadfa192bc - MoarVM version 2020.05-12-g3605efdbf
csv-ip5xs0.837 - 0.863
csv-ip5xs-208.600 - 8.833
csv-parser26.171 - 28.634
csv-test-xs-200.417 - 0.430
test7.749 - 8.045
test-t1.956 - 2.000
test-t --race0.867 - 0.944
test-t-2032.372 - 33.595
test-t-20 --race9.379 - 9.988
booting, absolutely; starting desktop, sure! 06:47
maybe this is good: it shows my tests are pure computing speeds independant of the disk speed 07:19
08:25 sena_kun joined
tyil weekly: best-friends.chat/@nvsr/104177209614718661 08:47
notable6 tyil, Noted! (weekly)
08:55 domidumont joined
Geth ¦ problem-solving: lizmat assigned to jnthn Issue Gist type objects as `(Foo:U)` rather than `(Foo)` github.com/Raku/problem-solving/issues/194 09:57
10:03 domidumont left 10:25 Altai-man_ joined 10:28 sena_kun left 11:26 jjmerelo joined 11:42 hankache joined
hankache Hello * 11:42
jjmerelo Hey! 11:44
hankache The RC for Rakudo Star 2020.05.1 on Windows is available at hankache.com/files would you be kind enough to test it and report possible issues? I was having trouble setting permissions and one specific file "repo.lock" kept me from installing modules. I figure out a way to adjust permission for said file at install time. Nonetheless if you can test it thoroughly it would be awesome. Thanks 11:46
hi jjmerelo 11:47
12:02 hankache left
Geth nqp: 1e3ecca5d6 | (Tom Browder)++ (committed using GitHub Web editor) | docs/release_guide.pod
perl -> raku
12:18
12:26 sena_kun joined 12:28 Altai-man_ left 12:31 domidumont joined 13:00 bazzaar joined 13:30 travis-ci joined
travis-ci NQP build passed. Patrick Böker 'Document `--relocatable` flag 13:30
travis-ci.com/tbrowder/nqp/builds/166519486 github.com/tbrowder/nqp/compare/09...a1580a6d50
13:30 travis-ci left 13:41 domidumont left
tbrowder don't know why my fork is being announced! 14:11
14:25 Altai-man_ joined 14:28 sena_kun left
Geth nqp-configure: Prince213++ created pull request #21:
Use the 'Microsoft' keyword to detect nmake.
15:12
16:26 sena_kun joined 16:28 Altai-man_ left 16:45 bazzaar left
Geth rakudo: a56e62de52 | (Elizabeth Mattijsen)++ | src/core.c/Rakudo/Internals.pm6
Enhance R:I.NORMALIZE_ENCODING

Now also takes a type object to indicate 'utf8' encoding. This is handy when an :$enc parameter is not initialized to a default. This should also allow passing around $encoding values as is, without needing to provide a default vaklue everywhere.
17:17
rakudo: d03fa4b022 | (Elizabeth Mattijsen)++ | src/core.c/IO/Path.pm6
Make IO::Path.slurp upto 4.7x as fast

  - 4.7x as fast for :bin on a <4K file
  - 3.8x as fast for :enc(foo) on a <4K file
  - 2.8x as fast for :bin on a 2.4M file
  - 0.1% faster for :enc(foo) on a 2.4M file
This is achieved by going full native in IO::Path.slurp *without* creating an IO::Handle or any of the intermediate objects. This bypasses the costly creation of an IO::Handle object and calling the "open" function on it.
lizmat and the same goes for the slurp sub as well, of course :-) 17:18
and hopefully this should also help with modules that load a lot of precomped modules 17:20
17:37 lichtkind joined
Geth rakudo: f3baa389d9 | (Elizabeth Mattijsen)++ | src/core.c/io_operators.pm6
Make slurp() upto 40% faster

For a small file with :bin, bigger files and/or with encoding will see smaller improvements. Achieved by adding candidates for :bin,
  :enc and none, instead of slurping any nameds and passing them on.
17:39
nine lizmat++ # this doesn't just make slurp faster but also more readable as it's clear what argument it supports and this helps with introspection in IDEs 17:48
Xliff \o 18:00
Restarting work on parallel compilation and .precomp locking 18:01
18:04 jjmerelo left
Xliff nine: So I've looked at the latest changes to the precomp stuff and think another object is needed to handle locking and file access, since PrecompilationStore::File generally only deals with IO::Handles. 18:07
nine: I'm not set on the name, but something like PrecompilationStore::File::Item, which would then implement all of the file operations, including a .lock. 18:08
PrecompilationStore::File would still implement !lock() would would lock the entire store. 18:09
Does this sound like a suitable train of thought to run down, or do you see problems?
lizmat Would anybody complain if I would make IO::Handle.slurp not needing an .open anymore ? 18:17
18:22 Xliff left 18:25 Altai-man_ joined 18:28 sena_kun left
Geth rakudo: c1decf2a46 | (Elizabeth Mattijsen)++ | src/core.c/IO/Handle.pm6
Improve readability of IO::Handle.slurp

Still doesn't make a lot of sense to me.
18:36
rakudo: a69465a75d | (Elizabeth Mattijsen)++ | 2 files
Don't use slurp-rest in the core anymore
18:45
nine lizmat: would that make sense? What would be the difference to IO::Path.slurp? 19:51
lizmat it would bypass all of the rigamerole of IO::Handle.open 19:52
and it would call just that
nine Xliff: I thought you were already pretty close before you stopped? Can't remember a need for shuffling around anything
tellable6 nine, I'll pass your message to Xliff
lizmat but I've abandoned the idea
nine I just learned that we can have un-opened IO::Handles and that IO::Handle has a $!path. And am confused by this 19:53
lizmat yeah, it's a mess 19:54
20:26 sena_kun joined 20:28 Altai-man_ left
Geth rakudo: fc3e1aea7c | (Elizabeth Mattijsen)++ | src/core.c/io_operators.pm6
Make sure :bin / :enc are mandatory in sub slurp

Spotted by ugexe++
20:53
¦ problem-solving: vrurg assigned to jnthn Issue Use of literal named parameters syntax for multi-dispatching github.com/Raku/problem-solving/issues/195 20:56
21:03 sena_kun left
Geth rakudo: 7571f65a43 | (Elizabeth Mattijsen)++ | src/core.c/IO/Path.pm6
Make IO::Path.spurt upto 35% faster

For small files, larger files will see smaller improvements.
This is achieved by going full native in IO::Path.spurt *without* creating an IO::Handle or any of the intermediate objects. This bypasses the costly creation of an IO::Handle object and calling the "open" function on it.
21:46
lizmat and this concludes my hacking for today& 21:47
21:52 ufobat__ joined 21:56 ufobat_ left, Kaeipi left 21:57 Kaiepi joined 22:15 Kaiepi left 22:16 Kaiepi joined 23:28 rypervenche joined