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.
AlexDaniel .tell dogbert2 commit title could've been better here: github.com/rakudo/rakudo/commit/8f807fc4 00:30
yoleaux AlexDaniel: I'll pass your message to dogbert2.
AlexDaniel .tell dogbert2 especially given that there are two commits with the same title :) github.com/rakudo/rakudo/commit/ecd899c9 00:32
yoleaux AlexDaniel: I'll pass your message to dogbert2.
releasable6 Next release in ≈1 day and ≈15 hours. 2 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 03:00
Geth rakudo: 07cfbeb2cd | (Paweł Murias)++ | 5 files
Turn p6reprname into a desugar
09:49
Geth nqp/hllbool: 0527d90895 | (Stefan Seifert)++ | src/vm/moar/QAST/QASTOperationsMAST.nqp
Map hllbool and hllboolfor ops
10:56
rakudo/hllbool: 4 commits pushed by (Stefan Seifert)++
pmurias nine: should I write tests for hllbool/hllboolfor or do you have ones uncommited/plan to? 11:52
pmurias starts writing those tests 11:54
pmurias nine: turns out the MoarVM op addition is not yet public :/ 12:41
timotimo yeah, it's in a branch, just like the nqp code is 12:47
nine pmurias: I wouldn't mind at all if you wrote some tests :) 13:03
Geth nqp/hllbool: 7517ed04b1 | (Paweł Murias)++ | 2 files
Add tests for nqp::hllboolfor and nqp::hllbool
13:57
pmurias nine: I'll implement it on js (and jvm later) 13:58
pmurias do we actually need nqp::hllboolfor for anything? 14:00
pmurias
.oO(it would be usefull in nqp code)
14:02
jnthn I think for in the MOP code 14:02
timotimo yeah, i imagine you'd put it at the very least into BOOTSTRAP.nqp
pmurias I forgot nqp::p6bool was registered in nqp land too 14:04
pmurias also shouldn't the hll be the first argument to nqp::hllboolfor (it's a bit bikeshedding) 14:05
Geth nqp/hllbool: f1ce0029a5 | (Paweł Murias)++ | 2 files
[js] Implement nqp::hllbool and nqp::hllboolfor
14:16
pmurias hlllizefor has the same API so my bikeshad is incorrect ;) 14:19
timotimo what does a bikeshed need to do wrong to be considered "incorrect", i wonder 14:20
pmurias timotimo: the order of arguments I was arguing as inconsistent is already used by a different op so it's consistent 14:30
timotimo oh, but i was refering to an actual bikeshed for real-life bikes 14:35
pmurias timotimo: too small that bikes don't fit it, or insecure that your bikes get stolen ;) 14:40
timotimo what is this, a bikeshed for ants?
Geth nqp/hllbool: c246068226 | (Paweł Murias)++ | 3 files
[jvm] Implement nqp::hllbool and nqp::hllboolfor
14:42
jnthn We should probably stop doing the "anonymous class embedded in a method" thing anywhere that's a hot path 14:56
I know that I'm guilty here :)
I'd also thought we might trivially optimize that out, but it turns out...not so easily
Geth rakudo: 06073c5b55 | (Jonathan Worthington)++ | src/core/Rakudo/Iterator.pm6
Move some classes out of enclosing methods

We should probably make a pass through and do this for any code that may be a hot path. I'd hoped we would be able to optimize the closure captures out easily; alas, we can't, and it can make some otherwise small and potentially inlineable methods into things that `capturelex` or similar and so cannot be. For some code that obtained many iterators of the types involved here, this gave a fairly notable improvement.
15:20
timotimo honestly, i'm glad that they now have names 15:21
they show up as <anon|34> in the profiler otherwise
jnthn That's I guess another benefit :)
A run of a benchmark I'm looking at took 2.993 before and 2.784 after that 15:22
timotimo wow, that's worth a lot 15:23
jnthn yes, on that particular benchmark it was the difference between being slower than Ruby and being as fast as Ruby :P
timotimo cool 15:25
but ruby is pretty slow, isn't it?
jnthn Depends on the benchmark
timotimo i guess that's always true :)
jnthn I've a utf-8 million lines reading benchmark where we beat Perl 5 and Python but Ruby wins on that one. :) 15:26
timotimo interesting! 15:27
Geth rakudo: 34468f11dc | (Jonathan Worthington)++ | src/core/Hash.pm6
Slight tweak to ASSIGN-KEY

Generates simpler code, so is faster unspecialized, and specializes a bit better too.
16:11
pmurias m 18:07
sorry
releasable6 Next release in ≈19 hours. 2 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 23:00