Geth roast/absolute-lib-paths-take2: dfe7430ded | (Nick Logan)++ | 265 files
Allow tests to be run outside of rakudo dir

Absolutify lib paths and file paths using $?FILE.IO.parent(...), since a test can know the file layout of its own repo but not the cwd the user is invoking a test from.
This allows you to install perl6, clone the roast anywhere you ... (11 more lines)		 01:18
roast/absolute-lib-paths-take2: fb83b01d95 | (Nick Logan)++ | S11-modules/require.t
Revert "[v6.d REVIEW] Use preciser test routines"

This reverts commit 6fa5f6481ffe2f6458e32087c1c9d209b056234c.
  `use "t/spec/packages"`, and hence Test::Util, cannot be used
here because the 'can change directory and require a module' test relies on t/spec/packages not being loaded.
roast: ugexe++ created pull request #390:
Allow tests to be run outside of rakudo dir		 01:20
rakudo/release/2018.02.1: badec48118 | (Aleks-Daniel Jakimenko-Aleksejev)++ | 3 files
Docs for the point release
 02:55
Zoffix makes a wild guess that the problem is in JSON::Fast 03:07
AlexDaniel sounds reasonable github.com/timo/json_fast/commit/d...c663e710ef 03:12
Zoffix 2018.02.24 zoffix@VirtualBox~$ perl6 -MJSON::Fast -e 'say from-json q|"\/"|' 03:13
\\/
2018.02.24 zoffix@VirtualBox~$ perl -MMojo::JSON=from_json -wlE 'say from_json q|"\/"|'
/
bah: github.com/timo/json_fast/issues/36
.tell timotimo FWIW there's a possible regression in JSON::Fast vis-a-vis unescaping slashes: github.com/timo/json_fast/issues/36 03:15
yoleaux Zoffix: I'll pass your message to timotimo.
AlexDaniel what's the most appropriate place for the sakefile that I'm using to cut releases? Doesn't really fit into whateverable project because it's not related to the bot framework, I also hardly see it in the rakudo repo itself 03:18
Zoffix huggable: shrug 03:19
huggable Zoffix, ĀÆ\_(惄)_/ĀÆ
AlexDaniel should I make an App::Releaser module or something?
Zoffix If it's for cutting rakudo releases, why not stick it into rakudo repo? 03:20
AlexDaniel hehā€¦ I guess that's a good point 03:21
Geth rakudo: MasterDuke17++ created pull request #1558:
Speedup Rational.Str		 05:41
rakudo/release/2018.02.1: b656f6b1b1 | (Aleks-Daniel Jakimenko-Aleksejev)++ | VERSION
[release] Bump VERSION to 2018.02.1
 06:06
rakudo: dc0f06b500 | (Aleks-Daniel Jakimenko-Aleksejev)++ | 3 files
Docs for the point release
 06:08
AlexDaniel alright, that seems to be correct 06:09
p6lert: help 06:21
p6lert AlexDaniel, github.com/perl6/alerts P6lert commands: [insta]?add ALERT, update ID ALERT, delete ID; ALERT format: ['severity:'\S+]? ['affects:['<-[\]]>+']']? ALERT_TEXT
AlexDaniel p6lert: add severity:normal affects:[2018.02] Rakudo 2018.02 (note: no .1) was discovered to contain a bug which caused segfaults when using Whatever curries in topics of for/given statement modifiers. This is fixed in 2018.02.1 point release. 06:27
p6lert, add severity:normal affects:[2018.02] Rakudo 2018.02 (note: no .1) was discovered to contain a bug which caused segfaults when using Whatever curries in topics of for/given statement modifiers. This is fixed in 2018.02.1 point release.
well, I guess that doesn't work?
or I just don't have the rights :)
Geth nqp/master: 4 commits pushed by pmurias++ 09:50
Zoffix AlexDaniel: right, no rights. If you have a stable IRC host I can add you to the list. 11:46
AlexDaniel Zoffix: is the one AlexDaniel` has good enough?
yea it should be
AlexDaniel` uhā€¦ am I even here? 11:47
AlexDaniel yea
Zoffix: oh, but that's not the username you're talking about
then no I think 11:48
(like the one you register)
Zoffix Right, I meant IRC *host*. Like perl6/zoffix 11:49
p6lert: add Rakudo compiler-only release 2018.02 (note: no .1) was discovered to contain a bug which caused segfaults when using Whatever curries in topics of for/given statement modifiers. This is fixed in 2018.02.1 point release. 11:50
p6lert Zoffix, Added alert ID 4: alerts.perl6.org/alert/4
Zoffix AlexDaniel++ # for cutting point release 11:51
|Tux| Rakudo version 2018.02-25-gdc0f06b50 - MoarVM version 2018.02-6-g1849ae6d6
csv-ip5xs0.807 - 0.833
csv-ip5xs-207.207 - 7.509
csv-parser11.855 - 12.810
csv-test-xs-200.436 - 0.441
test9.349 - 9.607
test-t2.619 - 2.657
test-t --race1.089 - 1.101
test-t-2046.417 - 46.681
test-t-20 --race16.361 - 16.716
 12:13
Geth rakudo: 008b9279ad | (Daniel Green)++ | src/core/Rational.pm
Speedup Rational.Str

By generating the fractional part two characters at a time instead of one.		 12:21
rakudo: 6006e2f633 | (Zoffix Znet)++ (committed using GitHub Web editor) | src/core/Rational.pm
Merge pull request #1558 from MasterDuke17/speedup_Rational.Str

Speedup Rational.Str
AlexDaniel mst: can I have perl6 dev cloak? 12:38
timotimo i'm having trouble building a test case for \/ that currently fails; what do i have to do? 14:29
yoleaux 03:15Z <Zoffix> timotimo: FWIW there's a possible regression in JSON::Fast vis-a-vis unescaping slashes: github.com/timo/json_fast/issues/36
timotimo what the heck, the code looks like it's been trying to replace \n with n and \b with b and all that for a while, how did this ever work 15:23
Zoffix m: use JSON::Fast; use Test; is-deeply ļ½¢"\/"ļ½£.&from-json, ļ½¢/ļ½£, "backslashed slashes get unescaped" 15:31
camelia ===SORRY!===
Could not find JSON::Fast at line 1 in:
/home/camelia/.perl6
/home/camelia/rakudo-m-inst-1/share/perl6/site
/home/camelia/rakudo-m-inst-1/share/perl6/vendor
/home/camelia/rakudo-m-inst-1/share/perl6
CompUniā€¦
Zoffix staar: use JSON::Fast; use Test; is-deeply ļ½¢"\/"ļ½£.&from-json, ļ½¢/ļ½£, "backslashed slashes get unescaped"
star: use JSON::Fast; use Test; is-deeply ļ½¢"\/"ļ½£.&from-json, ļ½¢/ļ½£, "backslashed slashes get unescaped"
camelia ok 1 - backslashed slashes get unescaped
Zoffix timotimo: ^
timotimo i get "ok" locally now 15:33
but i already got that with what's "master" on my github
so i should just have released?
Zoffix timotimo: my test fails for me even with master 15:44
This is the repo, right? github.com/timo/json_fast
timotimo OK, let's see, i'ma push a branch in a tiny bit
that's the one
Zoffix Fails for me on fresh clone: gist.github.com/zoffixznet/47f686c...d3569b9e9c 15:46
timotimo check the "escape-unescape-fixes" branch please 15:48
Zoffix timotimo: passes in that branch
timotimo good
i believe i need even more tests
and i need to stop using "" and '' in my tests to write out json literals 15:52
especially in the escaping tests 15:54
Zoffix: you think this is good for release perhaps?
Zoffix I guess :) It does fix a critical bug :)P 15:55
Geth nqp: 6269c85dcc | pmurias++ | src/vm/jvm/runtime/org/perl6/nqp/sixmodel/SerializationReader.java
[jvm] Fix typo in comment
 17:39
nqp: ca9da81087 | pmurias++ | 2 files
[js] Resolve object repossession conflicts
nqp: a30610a7a7 | pmurias++ | src/vm/js/nqp-runtime/io.js
[js] Fix nqp::rmdir

Make it ignore missing directories and throw a proper exception for non-empty ones
travis-ci NQP build failed. pmurias '[js] Fix nqp::rmdir 17:57
travis-ci.org/perl6/nqp/builds/345354504 github.com/perl6/nqp/compare/02628...0610a7a75d
stmuk is 2018.02.1 tagged? I don't see it 18:27
Zoffix stmuk: yeah, it is. And I see the tag 18:29
japhb Are the fixes on both the mainline and the branch? 18:30
Zoffix "12 hours ago": i.imgur.com/IIMLnd4.png
stmuk Zoffix: the VERSION file in rakudo master still says 2018.02 18:33
jnthn japhb: The fix was committed in mainline, and was cherry-picked to a release branch, I believe. So yes. 18:38
japhb jnthn: Ah good. 18:41
stmuk apart from b656f6 18:43
Zoffix stmuk: yeah. I it was realeased off a branch. I guess it also needs to be merged to master?
AlexDaniel`: ping ^
japhb .tell retupmoca Saw this while `zef install`ing my usual packages: ===> Extraction: Failed to find a META6.json file for JSON::JWT:ver<1.0>:auth<github:retupmoca> -- failure is likely 18:44
yoleaux japhb: I'll pass your message to retupmoca.
Zoffix m: dd :{:$_} 18:48
camelia -> ;; $_? is raw { #`(Block|59930128) ... }
Zoffix This is a bug, innit? That shouldn't ever be a Block
m: dd :{;}
camelia -> ;; $_? is raw { #`(Block|63574576) ... }
TimToady looks kinda like a bug to me; the $_ criterion was only supposed to be for disambiguating bare {} in cases where it's obviously intened as a block, like in: map { $_ => 1 }, ... 18:52
yoleaux 21 Feb 2018 23:13Z <Zoffix> TimToady: would you like to weigh in on this Issue: github.com/rakudo/rakudo/issues/1552 [@LARRY][RFC] [@LARRY] Booliness of Zero-Denominator Rationals
Zoffix Filed as R#1559 18:53
synopsebot R#1559 [open]: github.com/rakudo/rakudo/issues/1559 :{ } sometimes constructs a Block
TimToady on the truthiness of nan rats, I can argue it both ways, but I'm inclined to think that the cost of checking the denominator is only when the numerator is already 0, which is usually not going to be the case, so I think the efficiency argument doesn't for making it True doesn't carry a lot of weight 18:57
so maybe this morning I'm more in favor of Least Astonishment
(assuming here that we want infinities to be true) 18:58
Zoffix TimToady: Least Astonishment is zero-denominator Rats to be True?
TimToady is 0/0 False, 1/0 True 18:59
Zoffix oh
TimToady if num is non-zero, it's always True, otherwise, actually, I guess we don't need to check denom if we want NaN to be falsey 19:00
japhb Least astonishment would have ?NaN == ?(NaN.Rat), no?
TimToady well, there are probably conflicting astonishments here 19:01
japhb "conflicting astonishments" -- I like that phrase
Zoffix +1 on False being "denominator is 0" and the rest is true 19:02
TimToady anyway, I'm okay with a bit of inconsistency between num and rat
there are no perfect number representations in computers anyway... 19:03
japhb s/in computers//
TimToady :D 19:04
.oO(the first uninteresting integer)
japhb Apropos of which, I find my brain enjoys numerically interesting clock times. 12:34, 12:36, 12:48, 11:11, 22:22, 10:01, that sort of thing. But I suspect the clock times that I will find interesting are rather dense, because dangit, there's so many interesting numbers. 19:07
Speaking of which, 11:11 19:11
stmuk 16:20 (dude!) 19:12
Zoffix err @Zoffix[*-1] ~~ s/denominator/numerator/;
TimToady nod, I actually read it that way :) 19:13
japhb Latest Rakudo builds and zef installs all my usual stuff. (This is an improvement, because a week ago it was segfaulting) 19:16
TimToady has to borrow a power snake this morning, sigh... 19:17
.oO("Power tools, power tools, it's my favorite thing to use...")
 19:18
Zoffix m: dd <-42/-1> 19:29
camelia RatStr.new(42.0, "-42/-1")
Zoffix Would've thought that to make a regular Rat
japhb m: dd < -42/-1 > 19:32
camelia RatStr.new(42.0, "-42/-1")
japhb m: dd <-42/1> 19:33
camelia -42.0
Geth nqp: b5da011768 | usev6++ | src/vm/jvm/runtime/org/perl6/nqp/io/FileHandle.java
[JVM] Reset eof to false after seeking

Fixes tests in S16-io/eof.t (github.com/rakudo/rakudo/issues/1322).
After a call to seek the state of eof might very well have changed. Better reset it to false so that it is properly checked the next time eof is tested.		 19:34
gfldex m: my @a = (<a b>, (1, 2)); dd @a[*;1..*]; 19:39
should that terminate?
camelia (timeout)
Zoffix wonders if .Bool of allomorphs is actually explicitly specced 19:43
Don't see anything in S02-literals/allomorphic.t and they're using .Bool of numeric portion, but seemingly by accident
m: dd ?<0.0>, ?<0>, ?<0e0>, ?<0+0i >
camelia Bool::False
Bool::False
Bool::False
Bool::False
TimToady I suspect preferring the numeric test is more betterer 19:44
Zoffix nods
Geth rakudo: 748d1a57b9 | (Zoffix Znet)++ | src/core/Rational.pm
Implement Rational.Bool

Fixes R#1552 github.com/rakudo/rakudo/issues/1552
Changes[^1] behaviour of zero-denominator rationals, making <0/0> be False. The rest remain as they were, except now likely perform slightly faster.
  [1] irclog.perlgeek.de/perl6-dev/2018-...i_15851526		 19:45
synopsebot R#1552 [open]: github.com/rakudo/rakudo/issues/1552 [@LARRY][RFC] Booliness of Zero-Denominator Rationals
roast: f82c984b43 | (Zoffix Znet)++ | S32-num/rat.t
Spec Rational.Bool

Closes github.com/rakudo/rakudo/issues/1552 R#1552
Rakudo impl: github.com/rakudo/rakudo/commit/748d1a57b9 Discussion: irclog.perlgeek.de/perl6-dev/2018-...i_15851526
travis-ci NQP build failed. usev6 '[JVM] Reset eof to false after seeking 19:56
travis-ci.org/perl6/nqp/builds/345400559 github.com/perl6/nqp/compare/a3061...da01176878
Geth roast: 4fab1d3d7c | (Zoffix Znet)++ | S02-literals/allomorphic.t
Spec .Bool on allomorphs

Uses .Bool of numeric portion.
  irclog.perlgeek.de/perl6-dev/2018-...i_15851777		 19:57
Zoffix ZOFVM: Files=1289, Tests=153234, 145 wallclock secs (20.17 usr 3.10 sys + 3126.66 cusr 152.58 csys = 3302.51 CPU) 20:21
Also the "slightly faster" on that Rational.Bool commit turns out to be 11.5x faster :P MasterDuke++ for bringing that one up. 20:22
dogbert11 m: Seq.Bool 20:36
camelia Invocant of method 'Bool' must be an object instance of type 'Seq', not a type object of type 'Seq'. Did you forget a '.new'?
in block <unit> at <tmp> line 1
Zoffix
.oO( I thought we fixed all this stuff )
 20:37
I see more 20:38
m: Seq.Int
camelia Cannot look up attributes in a Seq type object
in block <unit> at <tmp> line 1
dogbert11 RT #130867
synopsebot RT#130867 [new]: rt.perl.org/Ticket/Display.html?id=130867 [BUG] `only` methods are used where `multies` should be, preventing dispatch to Mu.* candidates
Zoffix Ah, OK, yeah, that's the ticket I was thinking about
Seq.Numeric and Seq.elems are in the same pile 20:39
dogbert11 wasn't lizmat working on stuff like this the other day? 20:41
lizmat dogbert11: I started doing some subs and then got distracted :-( 20:42
dogbert11 I get distracted all the time :-)
Zoffix dogbert11: hm, good point. I wonder if we'll then have the opposite problem: methods that were declared non-multi on purpose, to block off the multi from higher up :) 20:44
(the fix would be to just stick a proto in the class, but still)
dogbert11 so how do we figure out if 'multi' was left out on purpose then 20:45
Zoffix Well, I know some of that stuff (like .new) is covered by roast. 20:51
The rest, we'd just need to think about each one, I guess.
lizmat fwiw, I intend to look further 20:52
Zoffix lizmat++
Zoffix &
lizmat but I appear to tire easily still, so I try to maximize time on recovery atm
dogbert11 clever move
lizmat like calling it a night at 10pm :-) 20:57
dogbert11 sleep well 21:00
lucasb I submitted 2 new suggestions in 6.d-prep. I appreciate any comments there. Thanks. 22:46
tbrowder .tell AlexDaniel NQP#424 has an iteresting mod to .travis.yml to help get past stallevjobs 22:47
yoleaux tbrowder: I'll pass your message to AlexDaniel.
synopsebot NQP#424 [open]: github.com/perl6/nqp/pull/424 add a new built-in sub: run-command
lucasb I see both proposals as related, but it seemed better to submit them separated. (In other words, if there was a .clear method, the purpose of the undefine() function would diminish.) 22:48
jnthn Not sure Arrays need a .clear method, just assign () or Empty 22:49
tbrowder s/stallevjobs/stalled jobs/
lucasb jnthn: Correct. On the other hand, assignment can't be chained like a method call chain, right? like @array.something.clear.something 22:51
jnthn Do you have a use case? 22:52
I mean, it's not like many methods on @array return the array itself
lucasb idk, maybe a use case would be to push to a empty array? if $something { @array.clear.push(...) } else { @array.push(...) } 22:54
*if* the clear method returns self. I suggested other options for the return value in the issue
jnthn That feels so cluttery compared to @array = ...; though? 23:02
timotimo Zoffix, you don't need to declare something non-multi to block multis from up above, you can just define a new proto 23:08
Zoffix timotimo: yeah, I know. But declaring an only works too and I was saying what if some of the methods are onlies because the multies weren't meant to be available. 23:09
multies up the food chain I mean
m: Rat.new: 1, 2, 3 23:10
camelia Too many positionals passed; expected 1 to 3 arguments but got 4
in block <unit> at <tmp> line 1
Zoffix If you blindlign just add `multi` to everything, taht ^ would start saying "Default .new takes only named args" thing is what I meant
timotimo oh, huh. 23:14
Zoffix m: my @array = <a b c>; (@array &&= ()).push: 42; dd @array 23:15
camelia Array @array = [42]
Zoffix Not sure why I'd be working with an array that got stuff up in it if I want an empty array tho 23:16
m: my @array = <a b c>; (@array = ()).push: 42; dd @array 23:17
camelia Array @array = [42]
Zoffix even just this works
.tell samcv the other day I said that executing code per lang version was blocked by R#1289, but I forgot to mention that that's basically for classes. For just subs, you can stick 6.d subs into files in src/core.d/ dir 23:35
yoleaux Zoffix: I'll pass your message to samcv.
synopsebot R#1289 [open]: github.com/rakudo/rakudo/issues/1289 [6.d][severe] [6.d BLOCKER] Implement a Way to Know Caller's Language
Zoffix .tell samcv there's already stuff for &await you could use as an example 23:36
yoleaux Zoffix: I'll pass your message to samcv.
samcv hmm
yoleaux 23:35Z <Zoffix> samcv: the other day I said that executing code per lang version was blocked by R#1289, but I forgot to mention that that's basically for classes. For just subs, you can stick 6.d subs into files in src/core.d/ dir
23:36Z <Zoffix> samcv: there's already stuff for &await you could use as an example
samcv ok
thank you
:)

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!