AlexDaniel unidump: 05:28
unicodable6 AlexDaniel, gist.github.com/1009cb702c7f752ecb...8158eda652
AlexDaniel /o\
unidump: a  b 05:29
unicodable6 AlexDaniel, gist.github.com/8cffe8e155ffc07ad0...af327d5784
AlexDaniel a: a  b 05:34
u: a  b
unicodable6 AlexDaniel, U+0061 LATIN SMALL LETTER A [Ll] (a)
AlexDaniel, U+0020 SPACE [Zs] ( )
AlexDaniel, 4 characters in total (a  b): gist.github.com/a3cb10c1f85a3dd72c...e4763addf2
AlexDaniel greppable6: « .* » 07:12
greppable6 AlexDaniel, 70 lines, 18 modules: gist.github.com/576bc17fe1eb2c0818...e13e4a7f26
AlexDaniel sad :(
Geth rakudo: a4f9090e24 | (Zoffix Znet)++ | 2 files
Fix X::TypeCheck::Binding::Parameter with subsets…

  …show the name of the subset rather than the base type.
Fixes github.com/rakudo/rakudo/issues/1560 R#1560		 07:46
synopsebot R#1560 [closed]: github.com/rakudo/rakudo/issues/1560 subset type check error points to the original type, not the subset name
Geth rakudo: 235d3f1ce7 | (Zoffix Znet)++ | src/Perl6/Metamodel/BOOTSTRAP.nqp
Remove unused code

Accidental left-over from
  github.com/rakudo/rakudo/commit/a4...4f121bd66c		 07:51
AlexDaniel releasable6: are you alive? 10:08
releasable6 AlexDaniel, I cannot recognize this command. See wiki for some examples: github.com/perl6/whateverable/wiki/Releasable
[Tux] Rakudo version 2018.02.1-38-g235d3f1ce - MoarVM version 2018.02-6-g1849ae6d6
csv-ip5xs0.816 - 0.843
csv-ip5xs-207.566 - 7.567
csv-parser12.289 - 12.338
csv-test-xs-200.455 - 0.477
test9.348 - 9.517
test-t2.595 - 2.645
test-t --race1.098 - 1.098
test-t-2046.498 - 48.937
test-t-20 --race16.190 - 17.422
 13:56
MasterDuke anyone mind merging github.com/rakudo/rakudo/pull/1563 ? 14:11
travis-ci NQP build passed. Samantha McVey 'Only run the new decode/encode conf tests on moarvm for now' 14:26
travis-ci.org/tbrowder/nqp/builds/345931388 github.com/tbrowder/nqp/compare/a8...f90977a16d
MasterDuke does anyone have any optimization idea for List.STORE? github.com/rakudo/rakudo/blob/mast...#L741-L806 14:29
pmurias do we have any sort of style document describing how we put in whitespace in the core 14:31
(somebody needs to write a working p6tidy)
MasterDuke i think it's on DrForr's to-do list, now that's he has a 1.0 of Perl6::Parser 14:33
pmurias question: like do we want to have nqp::bindattr(...,...,...,...) or nqp::bindattr(..., ..., ..., ...)? 14:36
the pull request seems to be aliging stuff 14:37
pmurias feels like asshat reviewing code formating with a code formatting style guide 14:38
MasterDuke i think most nqp:: calls in the core don't put spaces after the commas
pmurias * an asshat
MasterDuke i think it's because nqp:: calls usually create longer lines and this is a way to keep the size down 14:39
eh, i don't mind a reasonable formatting discussion 14:40
pmurias s/with/without/ 14:45
Geth rakudo: 36caa18a56 | MasterDuke17++ | src/core/Rational.pm
Whitespace fixes for Rational.pm
 14:47
rakudo: 2a8b17ae9c | (Paweł Murias)++ (committed using GitHub Web editor) | src/core/Rational.pm
Merge pull request #1563 from MasterDuke17/whitespace_fixes_for_Rational.pm

Whitespace fixes for Rational.pm
MasterDuke thanks 14:48
[Coke] yawns 17:04
AlexDaniel [Coke]: are you bored? There are 10 issues assigned to you :) github.com/issues?utf8=%E2%9C%93&a...d%3Afalse+ 17:07
skids what do we do these days to performance test a change locally? 17:21
japhb skids: What kind of change is it? It depends on whether you are optimizing a particular method, or a common construction, or underlying infrastructure ... 17:25
gfldex skids: I got a sizable JSON chunck that I read, modify, and store. The script in question likes threads. 17:26
skids Static optiizer tweak.
*optimizer
japhb Tweaking what exactly? 17:27
skids Eliminating unnecessary $_ ensconses on inlined blocks.
japhb (And FWIW, Zoffix and lizmat have done a lot of "optimize a particular construction and test the results")
Oh, that's gonna be a biggie.
skids Yeah more a general test than specific cases. 17:28
japhb One possibility for something that pervasive is to test the difference on the rakudo, spectest, and stresstest suites, plus some well-known benchmarks like [Tux]'s suite of CVS parsers. 17:29
Oh wait -- are you looking to confirm that there *is* an effect, to quantify what that effect is, or confirm that there are no regressions from it? 17:30
skids Yes? :-) 17:31
japhb Heh
Start with [Tux]'s main test-t script, because it runs quickly and fairly stably, and shows many optimizations well, and work out from there. 17:32
skids Ok, thanks.
gfldex skids: if you look for regressions you could setup your own meta-module and offload the testing to travis. 17:33
skids What's a meta-module? 17:34
gfldex skids: like so: gfldex.wordpress.com/2017/01/20/once-a-week/
skids Will check it out.
gfldex skids: and stuff the most used modules into the depends-section 17:35
Zoffix might be able to point to usage statistics on modules.perl6.org 17:36
japhb Looking at the travis screenshots in that post makes me wonder what culture sees switch to the left as on and switch to the right as off. 17:37
skids Reverse tinder? 17:39
(I've always hated those android style switches.)
gfldex the site might have been build be a legt handed person 17:40
if you use your left thunb to move a switch, moving it the the left is more natural
japhb Oh that's an interesting possibility 17:43
skids Well, I know everyone is acclimated, but I've always found the idea of touching something that says "off" to turn it "on" to be a bit dicey. 17:44
gfldex .oO( 5 hours of sleep is not good for your speeling )
japhb skids: FWIW I agree with you still 17:46
timotimo steevn spellburg
japhb Recently I saw a two-state button that read "Turn on" and "Turn off", which is I suppose mildly better, but because you then have to mentally compute the *current* state each time, is still far less than ideal.
skids Keeping both "on" and "off" visible and just moving the highlight/grayout makes those types of switches a bit better. 17:48
japhb wishes for the tactility and dimensionality of audio equipment
skids: Apple had at least one iteration in which they did something like that, except to my eyes the "highlight" appeared to *de-*emphasize the current setting 17:49
skids Yeah I remember that. 17:50
Hrm. A good number of missing meta deps installing Text::CSV... and internally withing the Test::META and JSON package stacks. 18:07
*within
Or maybe my zef is old or something. 18:09
MasterDuke skids: what frequently do is have a clone of moritz's perl6-all-modules and just add paths to there in a -I 18:14
*what i
jdv79 shouldn't 'make t/spec/S24-testing/1-basic.t' work? 18:15
gist.github.com/anonymous/0f389a0b...8b28d65caa 18:17
timotimo well, it's not in the spectest.data it seems like 18:28
for what reason? i don't know
jdv79 whatever. i'll just put it as a PR 18:37
do bugs go in rt or gh now?
MasterDuke GH preferred, but could use either 18:40
Geth rakudo: jdv++ created pull request #1568:
Unbusts like and unlike coercion.		 18:48
Kaiepi 6 18:53
agh i need to disable my touchpad 18:54
Zoffix MasterDuke: "i think most nqp:: calls in the core don't put spaces after the commas". lizmat doesn't put spaces after commas in nqp code 19:25
And I find overall that style makes it look a lot cleaner, but it irks me to write that way so in my nqp code I put spaces after commas. 19:26
MasterDuke yeah, i'm not sure that i'm 100% consistent. i tend to prefer spaces after commas in general, but generally go with what the majority of the surrounding code does (if there is pre-existing code) 19:30
Zoffix And I know closing parens placement is where lizmat-nqp and Zoffix-nqp differ on 19:31
Mine: github.com/rakudo/rakudo/blob/2a8b...pm#L25-L82 19:32
Lizmat's: github.com/rakudo/rakudo/blob/mast...3200-L3298
MasterDuke i believe yours is more common in lisp circles 19:33
Zoffix jdv79: I left some comments on your PR. If you want, I can just commit a fix myse;f/ 20:51
Geth rakudo: d5baa036fa | (Zoffix Znet)++ | src/core/IO/CatHandle.pm
Implement IO::CatHandle.handles

Lets the user work on source handles individually, which is especially handy when working with $*ARGFILES. e.g.:
   # print at most the first 2 lines of each file in $*ARGFILES:
   .say for flat $*ARGFILES.handles.map: *.lines: 2
Addresses R#1546 github.com/rakudo/rakudo/issues/1546		 21:41
synopsebot R#1546 [open]: github.com/rakudo/rakudo/issues/1546 [IO] Implement IO::CatHandle.handles
roast: 1a89a1e8f3 | (Zoffix Znet)++ | S32-io/io-cathandle.t
Spec IO::CatHandle.handles

Covers R#1546 github.com/rakudo/rakudo/issues/1546 Rakudo impl: github.com/rakudo/rakudo/commit/d5baa036fa
rakudo: 639c6da0ce | (Zoffix Znet)++ | src/core/IO/CatHandle.pm
Privatize the cat
 21:48
roast: f22a783617 | (Zoffix Znet)++ | README.md
Document we got Test::Utils

And how to write `use lib` for them		 21:57
roast: c13f9b1d18 | (Zoffix Znet)++ | README.md
Remove trailing whgitespace
roast: 98a2004409 | (Zoffix Znet)++ | README.md
Fix urls to module source
 22:02
jnthn
.oO( Privatize the cat? This small government thing is really getting out of hand... )
 22:12
timotimo who Handles the Cat? 22:16
MasterDuke and why is the cat not in a hat instead? 22:18
japhb These jokes are a catastrophe
MasterDuke don't fly off your handle! 22:19
japhb Stop, please, I'll go catatonic 22:21
;-)
Geth roast: 4adee51f32 | (Zoffix Znet)++ | packages/Test/Util.pm
Implement Test::Util::test-iter-opt

Tests iterator produces correct values and tests the values of
  .count-only/.bool-only—if implemented—are correct at each
step of reification.		 22:30
roast: 95a7aca30c | (Zoffix Znet)++ | S32-str/ords.t
Fix assumption Str.ords.iterator always…

  …offers .count-only/.bool-only optimization methods.		 22:31
jdv79 .tell zoffix yeah; just do it 22:38
yoleaux jdv79: I'll pass your message to zoffix.
jdv79 .tell zoffix note though that timo changed it in that commit i referenced in the issue for a reason.
yoleaux jdv79: I'll pass your message to zoffix.
Geth roast: 99504f7991 | (Zoffix Znet)++ | packages/Test/Util.pm
Add test-iter-opt variant that doesn't test values

For testing opts on iterators for which order of values isn't guaranteed, such as iterators from Hashes.		 22:47
roast: 2c021d7a1b | (Zoffix Znet)++ | MISC/bug-coverage.t
Simplify test code

Use the now-available `test-iter-opt` utility test routine.
roast: 1d33b3e421 | (Zoffix Znet)++ | S24-testing/14-like-unlike.t
Test like/unlike coerce to Str

Covers github.com/rakudo/rakudo/issues/1567		 23:50
rakudo: 7c1a6cac77 | (Zoffix Znet)++ | 3 files
Improve Test.pm6's like/unlike

  - Coerce arguments to Str
   Fixes github.com/rakudo/rakudo/issues/1567 R#1567
   Closes github.com/rakudo/rakudo/pull/1568 R#1568
  - Improve/fix failure diagnostics
   - Don't quote the $expected, it's a regex, not a string
   - .perl()ify the $got so user sees all potential whitespace weirdness
   - Say we wanted to [not] match against a regex, not just "expected"
   - `unlike` used to incorrectly say "expected: a Str **that matches**"
synopsebot R#1567 [closed]: github.com/rakudo/rakudo/issues/1567 Test::like() and Test::unlike() don't coerce anymore
synopsebot R#1568 [closed]: github.com/rakudo/rakudo/pull/1568 Unbusts like and unlike coercion.

Powered by Cro and the Raku® Programming Language.

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

Thank you!