lizmat commented on github.com/rakudo/rakudo/commit/25...33666a1264 01:15
lizmat gets a few hours of sleep
japhb .tell jnthn Sheesh, it just took *several hours* to show up. WTH, RT? 01:56
yoleaux2 japhb: I'll pass your message to jnthn.
geekosaur I swear they only run their mail queue every couple of hours; I get nothing for hours and then a bunch of ticket updates all at once 02:07
lizmat Files=1175, Tests=55878, 188 wallclock secs (11.49 usr 4.37 sys + 1123.80 cusr 110.80 csys = 1250.46 CPU) 05:45
lizmat starts commuting& 06:17
samcv .tell pmurias yeah it will happen in march sadly 06:32
yoleaux2 samcv: I'll pass your message to pmurias.
nine .tell lizmat github.com/niner/Inline-Perl5/commit/f461fabb67 is the solution :) 08:25
yoleaux2 nine: I'll pass your message to lizmat.
Geth rakudo/nom: 20614854da | (Samantha McVey)++ | 5 files
Add $*COLLATION var and implement coll operator

  $*COLLATION will allow setting of the collation level that the `col`
operator applies.		 11:35
timotimo oh, that's coll. 11:39
samcv so unicmp just does language and country inspecefic but coll will use the value of $*COLLATION 11:40
and we can have a .collate similar to .sort
jnthn Having just gone through a lengthy bisect to try and figure out what busted my $dayjob codebase, fd8df7f2ad is the commit that's apparently to blame 11:42
There is a use of .rotor in the code in question
samcv didn't seem too much changed in that commit other than moving locations and adding some #?if jvm 11:45
hm
jnthn Well, I do something like (gather some-func()).rotor(100, :partial) 11:46
And previously that would have called .list on the output of the gather prior to the .rotor stuff 11:47
Indeed, if I stick a .list call explicitly into my program, it seems to fix things.
m: sub foo() { for ^20 { take 'x' } }; for (gather foo()).rotor(3, :partial) { .say } 11:50
camelia rakudo-moar c0a907: OUTPUTĀ«(x x x)ā¤(x x x)ā¤(x x x)ā¤(x x x)ā¤(x x x)ā¤(x x x)ā¤(x x)ā¤Cannot invoke this object (REPR: Uninstantiable; Callable)ā¤ in block <unit> at <tmp> line 1ā¤ā¤Ā»
jnthn m: sub foo() { for ^20 { take 'x' } }; for (gather foo()).list.rotor(3, :partial) { .say }
camelia rakudo-moar c0a907: OUTPUTĀ«(x x x)ā¤(x x x)ā¤(x x x)ā¤(x x x)ā¤(x x x)ā¤(x x x)ā¤(x x)ā¤Ā»
jnthn Aha!
bisectable: sub foo() { for ^20 { take 'x' } }; for (gather foo()).rotor(3, :partial) { .say } 11:51
bisectable6 jnthn, Bisecting by exit code (old=2015.12 new=2061485). Old exit code: 0
pmurias jnthn: any quick hints why sub foo($foo, $bar = sub {$foo}) {...} works when using lowered signatures? It's get compiled to QAST::Block.new(my $value, QAST::Var.new(:name<$bar>, :default(...do p6capturelex on WVal with $block1 and invoke it...), $block1 = QAST::Block(do capturelex on WVal with $block2 and invoke it), $block2 = QAST::Block.new(...return $bar...))
yoleaux2 06:32Z <samcv> pmurias: yeah it will happen in march sadly
bisectable6 jnthn, bisect log: gist.github.com/4bfa4dc4835a8c9dd4...3eda3d3998
jnthn, (2017-01-18) github.com/rakudo/rakudo/commit/fd...622cf7486f
pmurias jnthn: I would expect the $block2 that for p6capturelex to work $block2 would need to be inside $block1, right? 11:52
jnthn pmurias: Yeah, I'm guessing we fail to nest stuff properly during the thunking 11:56
So it "works" (e.g. probably breaks if we recurse)
(or thread)
pmurias jnthn: why does it work at all? 11:57
jnthn pmurias: Probably some kind of autoclose
And that perl6_ops.c thing I pointed you at before where we ignore cases where the outer is wrong
Instead of blowing up on them 11:58
pmurias jnthn: should I make things nest properly? 12:02
jnthn pmurias: Yes 12:03
If you've time/motivation to do so :-)
They all want sorting out, it's not not always trivial to do so
But in this case should be
I think we might get it right for where clauses; if so can nab a solution from there 12:04
rotor bug reported as rt.perl.org/Ticket/Display.html?id=130725 12:05
Geth nqp: b5dea6d359 | (Samantha McVey)++ | tools/build/MOAR_REVISION
Bump MoarVM to get improvements to unicmp_s
 13:06
Ā¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...-gab225fa9
rakudo/nom: 4efcc29c80 | (Samantha McVey)++ | tools/build/NQP_REVISION
Bump MoarVM/NQP to get improvements for the coll collation operator
 13:08
Ā¦ rakudo/nom: version bump brought these changes: github.com/perl6/nqp/compare/2017....-gb5dea6d3
travis-ci NQP build passed. Samantha McVey 'Bump MoarVM to get improvements to unicmp_s' 13:10
travis-ci.org/perl6/nqp/builds/198830106 github.com/perl6/nqp/compare/439b5...dea6d35914
Geth nqp: 83080ffed0 | (Samantha McVey)++ | docs/ops.markdown
Update unicmp_s op documentation
 15:06
nqp: 2909825e5b | (Samantha McVey)++ | docs/ops.markdown
Add documentation for getstrfromname op and update unicmp_s desc.
 15:39
lizmat waves from Keflavik 15:47
yoleaux2 08:25Z <nine> lizmat: github.com/niner/Inline-Perl5/commit/f461fabb67 is the solution :)
jnthn o/ lizmat 15:48
lizmat jnthn o/
jnthn Been a long while since I was at Keflavik
Last time I went to Iceland I only ended up there half as many times as planned. :)
lizmat it's our first time.... and it's grey and raining :-) 15:49
jnthn heh, and thinking about it, that was the only time I went to Iceland,
Was meant to fly out of there. But...volcano :)
lizmat hehe
jnthn So ended up flying out of the airport on the north coast :)
lizmat is watching big purple plains with WOW on them
*planes 15:50
jnthn Much WOW.
lizmat .tell nine WOW :-) 15:59
yoleaux2 lizmat: I'll pass your message to nine.
lizmat news.ycombinator.com/item?id=13579838 #nice 16:03
Geth roast: ronaldxs++ created pull request #228:
add fudge-usage-stat.pl to show usage of fudge implementations, verbsā€¦		 16:05
lizmat commuting again& 16:07
IOninja hm, perl6.wtf builder is failing again to merge master into coverage branch: gist.github.com/zoffixznet/24bfd9e...08b01fa101 16:36
Something with ops. Maybe it'd be possible to merge that branch to master? 16:37
samcv :( 17:13
IOninja, also that may be the case, sorta. may need to add the op and then regenerate the stuff 17:16
with ./tools/update_ops.p6 17:17
that is what i'd do
[Coke] samcv: is your grant going to be submitted to the normal GC process in the next round? 17:30
yoleaux2 4 Feb 2017 17:01Z <brokenchicken> [Coke]: BTW is it just a typo on this grant approval page that it says "Test2 Manual" news.perlfoundation.org/2017/01/jan...votes.html ? I just thought... what if that vote was actually for Test2 Manual
samcv [Coke], i think it's already submitted
well he said he'd get back to me if they figure out a way to do it before March
but i assume it's already submitted for march becasue i emailed him the proposal
[Coke] brokenchicken: yes, it's a typo. I'll mention it to Makoto. 17:31
IOninja cool :) 17:44
[Coke] given that you have a grant manager, I think you're OK 17:54
TimToady nine: I'm having a problem with my braids branch where when make runs tools/build/install-core-dist.pl, it doesn't successfully re-Precompile CompUnit::Repository::Staging, and it's not clear what's going wrong 19:06
$core-dist.meta is just returning {ver => v0}
any kloo where I should look? 19:07
any amount of blowing things away doesn't help, so it's some difference in what I've done with the compiler, but I'm not seeing it 19:10
I don't think any of the REPO stuff depends closely on whether cursors use %*LANG directly, or indirectly via the cursor... 19:11
unless maybe something went wrong in an EVAL somewhere... 19:14
hmm, maybe EVAL $!serialized-spec; 19:21
no, that seems like it's used only for NativeCall stuff, and that all seems to precompile fine... 19:28
nine TimToady: that sounds....odd 19:32
yoleaux2 15:59Z <lizmat> nine: WOW :-)
nine Can hardly be the EVAL since CompUnit::Repository::Staging has no dependencies 19:33
TimToady maybe something's going wrong in compile_in_context, when has a weird comment:
# For now, EVALed code run during precomp will not get the 19:34
# outer lexical context's symbols as those may contain or
# reference unserializable objects leading to compilation
# failures. Needs a smarter approach as noted above.
nine Oh yeah, I remember that one
TimToady though I'm not seeing any compilation failures 19:35
well, a recompilaation failure, I suppose...
but I'm not really doing anything with outers
nine Maybe the JSON decoder? 19:36
TimToady I guess, if it uses cursors
a good lead, anyway
Geth roast: ronaldxs++ created pull request #229:
Fix typo misspellings of fudge implementation niezca		 19:38
mr_ron Commit message on pull amended to niecza 19:47
Geth roast: 629a459ecf | (Ronald Schmidt)++ | 4 files
Fix typo misspellings of fudge implementation niecza (#229)
 19:48
TimToady except it doesn't appear that the offending statement is actually parsing any JSON... 20:01
will think more about compile_in_context in the context of lunch... 20:10
MasterDuke anybody have comments/thoughts/objections on github.com/perl6/nqp/pull/345? if not i'll probably squash and merge later tonight 22:50
timotimo MasterDuke: i think the commit's fine :) 22:59
IOninja has no comments.
timotimo if it doesn't make the jvm unhappy
which i think you've checked it doesn't
TimToady nine: you were right, it was something to do with the JSON decoder, I wasn't setting up the grammar actions right 23:57

Powered by Cro and the RakuĀ® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!