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.
travis-ci NQP build passed. Vadim Belman 'Merge pull request #537 from vrurg/build-revamp 00:15
travis-ci.org/perl6/nqp/builds/533079195 github.com/perl6/nqp/compare/a3add...b0d87e6976
Geth nqp: vrurg++ created pull request #538:
Added detection of moar executable in --prefix
00:55
Kaiepi anyone on that can help me with some QAST stuff? 01:51
MasterDuke Kaiepi: i'm very much not an expert and make no promises, but what's the problem? 02:01
Kaiepi i can't figure out how to pass an array to a RI method
through a QAST::Op 02:02
from actions
MasterDuke what have you tried? 02:04
Kaiepi i've tried using a rakudo array, creating the sc for it, and passing it through a QAST::WVal, but since it's an array of blocks, i get this error: 02:05
The 'ForeignCode' class is a Rakudo-specific
implementation detail and has no serviceable parts inside
MasterDuke not using a QAST::Var? 02:07
Kaiepi haven't tried that 02:13
i'll see if it works
nope MasterDuke, same error 02:35
lizmat Files=1262, Tests=107948, 420 wallclock secs (30.13 usr 7.44 sys + 2953.36 cusr 270.04 csys = 3260.97 CPU) 08:31
Geth roast: Altai-man++ created pull request #537:
Add tests for #2862
10:46
Geth roast: 3b15fc12d3 | Altai-man++ | S32-list/head.t
Add tests for #2862
11:35
roast: b2ca325659 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | S32-list/head.t
Merge pull request #537 from Altai-man/test-head-intification

Add tests for #2862
|Tux| Rakudo version 2019.03.1-398-g6a60a99fc - MoarVM version 2019.05-7-gf8743c112
csv-ip5xs0.744 - 0.747
csv-ip5xs-205.753 - 6.112
csv-parser21.390 - 21.397
csv-test-xs-200.426 - 0.444
test6.749 - 7.128
test-t1.730 - 1.742
test-t --race0.769 - 0.797
test-t-2027.997 - 28.736
test-t-20 --race8.930 - 9.825
12:01
lizmat travelling& 13:26
Geth nqp/master: 4 commits pushed by (Vadim Belman)++ 14:35
Geth roast: f24ffc4e5c | usev6++ | 4 files
[JVM] Fudge some newly failing tests
17:52
Geth roast: b5c4ed2345 | usev6++ | 9 files
[JVM] Unfudge/Update some Unicode relates tests

  (after bumping Java version)
18:59
Kaiepi jnthn, you on? 21:29
or anyone good with QAST 21:30
timotimo i am decent at qast
what's the issue Kaiepi? 21:32
Kaiepi so for parallelizing feed operators, i have Rakudo::Internals.EVALUATE-FEED working, but i'm having trouble actually passing a list of callbacks to the method because i get an error about ForeignCode being a rakudo-specific implementation or something
timotimo interesting. did you get a backtrace with --ll-exception that got you close to the exact spot where that happens? 21:33
Kaiepi it doesn't help 21:34
bastille% ./perl6-m --ll-exception -e 'say ([1,2,3] ==> map { $_ ** 2 })'
The 'ForeignCode' class is a Rakudo-specific
implementation detail and has no serviceable parts inside
timotimo m: say Int.^attributes[0].^find_method("name").^methods
camelia {!set_name => The 'ForeignCode' class is a Rakudo-specific
implementation detail and has no serviceable parts inside, add_dispatchee => The 'ForeignCode' class is a Rakudo-specific
implementation detail and has no serviceable parts inside, clone =…
timotimo ^- this came from trying to .gist a code object 21:35
a code object that's too low-level to be friendly for introspection
ah, right, it's more a warning than an error
go to src/core/Rakudo/Internals.pm6 and on line 465 inside role ImplementationDetail make the gist method "die" with the text rather than just return it 21:36
then you can get the backtrace you want 21:38
Kaiepi ok 21:41
jnthn Kaiepi: Somewhere you'll usually want to take the QAST::Block and make a thunk from it or something; it's been a while since I did anything in actions, but search for things like create_code_object or make_topic_block or so on for inspiration 22:22
Note you'll need to make sure you're calling block_closure too so you don't get over-sharing
MasterDuke jnthn, timotimo, et al.: any thoughts on github.com/perl6/nqp/pull/532 after the most recent change? 23:07
timotimo the thing with comma is that comma already has profiling stuff in released versions 23:09
MasterDuke how does it do it? 23:10
jnthn It uses --profile-filename=... iirc
I think we'd probably have to support that for a migration period 23:11
MasterDuke so change --profile=, but leave --profile-filename= and --profile-kind= for now? 23:13
jnthn Yeah, though I'm pretty sure it's not passing --profile-kind 23:15
Maybe you could emit a warning if they're used?
And note they're going away at <some date>?
MasterDuke sure
jnthn We can adapt in Comma but we'll have to do it conditional on the Rakudo version we see.
So we don't break on things prior to this merge
So it'll need compat handling on both sides, just in different directions
MasterDuke do you want to wait until after the next release?
jnthn To merge this? 23:16
MasterDuke yeah
assuming i make the relevant changes before the next release 23:17
jnthn Not sure it helps; I mean, I'd want to have Comma's choice of what to pass predicated on a released Rakudo version
So it's chicken and egg.
So probably just merge it when it's ready
timotimo we didn't have --profile-kind yet
MasterDuke ok
jnthn timotimo: ah, yes :) 23:18
sleep time o/ 23:24