Perl 6 language and compiler development | Logs at colabti.org/irclogger/irclogger_log/perl6-dev | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
Set by Zoffix on 27 July 2018.
lizmat .tell scimon what you told about yesterday will continue to work until at least 5.e 08:10
yoleaux lizmat: I'll pass your message to scimon.
lizmat .tell scimon and judging from your slides, you didn't mention the internals apart from $*USAGE, and that will stay 08:14
yoleaux lizmat: I'll pass your message to scimon.
lizmat .tell scimon all changes are under the hood for all of your viewers
yoleaux lizmat: I'll pass your message to scimon.
Geth roast: a0af8bf031 | (Elizabeth Mattijsen)++ | S06-other/main-refactored.t
Add tests for MAIN_HELPER not getting called

If a new API "ARGS-TO-CAPTURE" is available.
08:21
lizmat m: sub MAIN { say "main" }; LEAVE say "goodbye" # according to docs, this is wrong 08:26
camelia main
goodbye
lizmat docs.perl6.org/language/functions#sub_MAIN states "The sub with the special name MAIN is executed after all relevant phasers."
either "LEAVE" is not relevant, or the doc is wrong or there's something else amiss 08:27
geekosaur seems badly phrased to me; I think it wants to say that it's run after all the appropriate phasers for entering the scope in which it's contained. 08:31
which leaves two questions: (1) it's specific abotu entering, why not exiting that scope? (2) when does it run relative to runnable code in that scope? 08:32
((2) might be answered in a different part, I haven't looked) 08:33
then again, if it means actually after all the ophasers fr that scope, then the behavior does nto match the docs. and someone needs tothink about thatbehavor, because I'd expect a LEAVE phaser to be used for e.g. cleaning up resources --- but I'd expect those resources to be available when MAIN is run 08:35
Geth rakudo: 24413b5f28 | (Elizabeth Mattijsen)++ | src/core/Main.pm6
Point to GENERATE-USAGE instead of USAGE

USAGE is going to be deprecated. Also unplan to deprecate $*USAGE for now, although I still think we should have a better way.
09:16
lizmat m: sub MAIN(Int:D $number) { dd $number }; @*ARGS = "foo" # shouldn't that mention that it should be an Int ? 11:58
camelia Usage:
<tmp> <number>
timotimo is it the :D that makes it not work any more?
lizmat m: sub MAIN(Int $number) { dd $number }; @*ARGS = "foo"
camelia Usage:
<tmp> <number>
lizmat bisectable6: sub MAIN(Int $number) { dd $number }; @*ARGS = "foo"
bisectable6 lizmat, On both starting points (old=2015.12 new=24413b5) the exit code is 2 and the output is identical as well
lizmat, Output on both points: «Usage:␤ /tmp/UF6goIiW74 <number> ␤»
lizmat hmmm... maybe room for improvement then 11:59
m: multi sub MAIN(Int:D :$name) { dd $name }; @*ARGS = "foo" # apparently we do this only for named parameters 12:07
camelia Usage:
<tmp> [--name=<Int>]
[Tux] Rakudo version 2018.09-410-g24413b5f2 - MoarVM version 2018.09-124-g588873f7a
csv-ip5xs0.907 - 0.909
csv-ip5xs-207.586 - 7.618
csv-parser22.950 - 24.222
csv-test-xs-200.425 - 0.427
test8.109 - 8.187
test-t1.698 - 1.818
test-t --race0.768 - 0.834
test-t-2031.333 - 31.500
test-t-20 --race10.613 - 12.420
12:08
the big gap might be disk caching. I just rebooted with e new kernel
diakopter is that good 12:33
every time you post that I have to ask for the link to the historical
timotimo buggable: speed 12:34
buggable timotimo, ▄▅▄▅▇▆▆▇▅▅▆▃▆▇↑█▆▆▆▆▃▁▃▃▃▄▂▅▄▂▂▃▄▃▃▃▅↑▇▃▂▄▂▂▂▁▂▁█▂ dates: 2018-09-21–2018-10-12; range: 10.398s–13.179s; speed: 11% faster (widths: 9/3) 12:35
diakopter yeah but url
timotimo tux.nl/Talks/CSV6/speed4.html
hum, it's not loading 12:36
[Tux] 1.698 is pretty ok
[Tux] checks site …
something broke https (again). still works as tux.nl/Talks/CSV6/speed4.html 12:40
timotimo OK
[Tux] fixed. https should work again 12:41
diakopter I'm looking forward to 0.01 seconds 12:42
lizmat diakopter: if you can get startup below 0.001, then that should be absolutely doable :-) 12:48
hmmm.. looks like the 0.768 for test-t --race is the lowest ever 12:50
afk&
Geth rakudo: 54c560e010 | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-notes.md
keep useful notes for others
13:01
rakudo: a245bc13a7 | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-notes.md
fix typo
13:15
travis-ci Rakudo build errored. Tom Browder 'keep useful notes for others' 14:02
travis-ci.org/rakudo/rakudo/builds/440623180 github.com/rakudo/rakudo/compare/2...c560e0103b
buggable [travis build above] ✓ All failures are due to: timeout (1 failure). 14:02
Geth rakudo: 243a2f01db | (Elizabeth Mattijsen)++ | src/core/Main.pm6
Tell if there's no useful documentation created

Which can happen if the only MAIN as a "hidden-from-USAGE"
14:11
tripleclonesrg Now it's running. :) 14:27
well, doesn't matter it seems
wolognE it's when you get a sawzall and replace the blade with a dildo 14:28
wolognE dude you helped me with a bash script there's no need to be all high and mighty lmao 14:28
derpingit: for a single strand of fiber, they're called bi-di sfp modules
pedahzur: It's strictly a hybrid ISO kind of thing
pingfloyd: Richard Stallman always uses he/him for himself
timotimo oh look the spam's back 14:29
hacostaUh Now it's running. :) 14:32
well, doesn't matter it seems
timotimo ah, nice, p6bannerbot is back, we can set +m again, then 14:33
.. if someone could :)
timotimo TYVM 14:37
Geth nqp/master: 4 commits pushed by (Paweł Murias)++ 14:49
Geth rakudo: b781479ea5 | (Paweł Murias)++ | tools/build/NQP_REVISION
Bump NQP for rakudo.js REPL fixes
14:55
¦ rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....g9bfc26ed7
travis-ci Rakudo build failed. Tom Browder 'fix typo' 14:55
travis-ci.org/rakudo/rakudo/builds/440630326 github.com/rakudo/rakudo/compare/5...45bc13a754
buggable [travis build above] ☠ All failures are due to: failed make test (1 failure). Across all jobs, only t/02-rakudo/repl.t test file failed. 14:55
Zoffix .ask pmurias would you have any guesses on this release blocker github.com/rakudo/rakudo/issues/23...-427702563 ? It bisects to js branch merge, yet is fine on both parents. Seems to be something with precomp. 14:56
yoleaux Zoffix: I'll pass your message to pmurias.
Zoffix releasable6: status 14:57
releasable6 Zoffix, Next release in ≈8 days and ≈4 hours. 4 blockers. 0 out of 413 commits logged 14:58
Zoffix, Details: gist.github.com/f5cf28dc4fe9976a7c...bb0e7e8525
Zoffix .tell jnthn as we're approaching the "sometime in October" date for giving 6.d changelog to users to prep for 6.d release, a friendly reminder about de-dynamification of $_ ( github.com/perl6/6.d-prep/issues/10 ) that needs to be done; the rework of start-in-sink-context you wanted to do ( github.com/perl6/doc/commit/df0b71d8d ) as well as @LARRY labeled Issues for 6.d stuff: 15:04
yoleaux Zoffix: I'll pass your message to jnthn.
Zoffix github.com/rakudo/rakudo/issues?q=...3A%40LARRY
.tell jnthn github.com/rakudo/rakudo/issues?q=...3A%40LARRY
yoleaux Zoffix: I'll pass your message to jnthn.
Geth nqp: 3816e22113 | (Paweł Murias)++ | Configure.pl
Cleanup indentation.
15:12
brrt .tell Zoffix the JIT blockers are not, in honesty, blockers 15:17
yoleaux brrt: I'll pass your message to Zoffix.
brrt .tell Zoffix my plan was to silence them unless MVM_JIT_DEBUG is passed as a flag
yoleaux brrt: I'll pass your message to Zoffix.
travis-ci Rakudo build failed. Elizabeth Mattijsen 'Tell if there's no useful documentation created 15:33
travis-ci.org/rakudo/rakudo/builds/440654193 github.com/rakudo/rakudo/compare/a...3a2f01dba0
buggable [travis build above] ☠ All failures are due to: failed make test (1 failure). Across all jobs, only t/02-rakudo/repl.t test file failed. 15:33
Geth rakudo: rbt++ created pull request #2363:
throws-like perform exact match for non-strings
16:43
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Tell if there's no useful documentation created 17:31
travis-ci.org/rakudo/rakudo/builds/440654193 github.com/rakudo/rakudo/compare/a...3a2f01dba0
Geth rakudo: 7d02b949ef | (Elizabeth Mattijsen)++ | src/core/Main.pm6
Remove trailing whitespace if there is no pod to add
18:43
Geth roast: 8da39b89e6 | (Elizabeth Mattijsen)++ | S06-other/main-refactored.t
Add test for "is hidden-from-USAGE"
18:50
lizmat sometimes I wish we had a .discard method, that would be the same as .grep, but with reversed conditional 19:27
".grep: !*.foo" is not very readable, ".discard: *.foo" so much better 19:28
since we have if vs unless, and split vs comb this feels like an omission 19:29
travis-ci Rakudo build failed. Elizabeth Mattijsen 'Remove trailing whitespace if there is no pod to add' 19:30
travis-ci.org/rakudo/rakudo/builds/440766703 github.com/rakudo/rakudo/compare/b...02b949ef9d
buggable [travis build above] ☠ All failures are due to: failed make test (6 failures). Across all jobs, 2 unique test files failed. 19:30
Geth rakudo: 9894eda5b5 | (Elizabeth Mattijsen)++ | t/05-messages/01-errors.t
Remove expected trailing whitespace

Looks like the problem fixed by 7d02b949ef9d69deff was actually tested for
19:35
lizmat that should make Travis happy again
dogbert11 lizmat: I see a few failing tests in t/05-messages/01-errors.t 19:37
lizmat dogbert11: I just fixed that ? ^^ 19:38
dogbert11 incredible, must have missed your fix by mere seconds. lizmat++
lizmat :-) 19:39
dogbert11 there are a couple of other files that fail. I suspect it's stuff that samcv is working on 19:40
to be more specific, there are a couple of passing TODO's 19:41
lizmat yeah, but I can't make heads or tails of those 19:45
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Remove expected trailing whitespace 20:23
travis-ci.org/rakudo/rakudo/builds/440784734 github.com/rakudo/rakudo/compare/7...94eda5b568
Geth rakudo: 2c43d2bf5f | (Elizabeth Mattijsen)++ | src/core/Main.pm6
Introduce &*ARGS-TO-CAPTURE and &*GENERATE-USAGE

Two dynamic variables representing the default handlers that can be used inside a ARGS-TO-CAPTURE or GENERATE-USAGE respectively.
20:45