»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by masak on 28 November 2015.
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: e102823 | (Zoffix Znet)++ | lib/ModulesPerl6.pm:
Purge old packed assets
00:00
RabidGravy stmuk_, probably won't be going 00:01
not enough fun
stmuk_ plenty of perl 6 talks (including my own) and beer! that's fun! for CVO anyway 00:04
Quom_ I think *funds (I know I'm in the same boat)
psch ...i count 4 Perl 6 talks, one being the keynote and another being criticism 00:10
masak 'night, #perl6 00:12
timotimo criticism? cool!
RabidGravy I couldn't think of a valid contribution to make to be honest 00:13
psch well, afair two points of that criticism where "if i inject IterationEnd into an Iterator it stops iterating" and "dispatching on Mu is weird"
s/where/were/
stmuk_ that's 4 more than LPW last year 00:15
RabidGravy psch, aka "it doesn't do what I want so it's all shit"
concrete contribution is key 00:16
RabidGravy you've got to have skin in the game 00:17
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 8f07327 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/PostProcessor.pm:
Add docs for ModulesPerl6::DbBuilder::Dist::PostProcessor baseclass
00:19
stmuk_ act.yapc.eu/lpw2015/talks/tag/whining 00:20
:> 00:21
psch ...i don't know if that's a good idea
but then i'm keeping my community involved inside this channel so eh vOv
RabidGravy I don't do community anymore 00:25
RabidGravy nighty night 00:27
uruwi Any way to get the bytes of an Int (capital I)?
TimToady m: say 17128347832173841723849712839478219037480932.polymod(256 xx *) 00:28
camelia rakudo-moar e0a403: OUTPUT«(228 23 40 12 144 143 142 64 125 138 15 49 80 201 182 176 159 196)␤»
uruwi Was hoping for a more efficient method, but that suffices. 00:29
TimToady nap & 00:30
csd_ This page seems to have out of date info on it: doc.perl6.org/type/Supply ; the code snippets aren't working for me 00:31
timotimo huh
i didn't know there was no coherent design behind perl6
timotimo no benefits of lessons learned from the old language? clashing features? 00:32
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 3965833 | (Zoffix Znet)++ | t/02-db-builder/08-postprocessor-base.t:
Improve test/pod coverage of ModulesPerl6::DbBuilder::Dist::PostProcessor to 100%
Zoffix csd_, how old is your Rakudo? 00:32
csd_ Zoffix: i installed it today
Zoffix csd_, installed from where? 00:33
csd_ rakudobrew
Zoffix csd_, there've been some changes to Supply stuff very recently. The docs likely haven't been updated yet
csd_ figure that someone here is probably the right person to let know 00:34
Zoffix csd_, it would be helpful if you either submit an Issue with which examples are wrong, or a PR with what they should be :) github.com/perl6/doc 00:34
jnthn is the one who made the changes, but I doubt they have the time to work on docs ATM :) Christmas release prep and all :)
psch rakudo commit a8231f14b2d5400e0653aac453496e00318142c5 explains a lot, i gather 00:35
i mean, the commit message
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: ea6dbc2 | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/PostProcessor.pm:
Reword what _builder is for
00:42
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 8ac60dc | (Zoffix Znet)++ | t/02-db-builder/06-postprocessor-travis.t:
Improve test coverage for ModulesPerl6::DbBuilder::Dist::PostProcessor::TravisCI
01:00
lucs_ Re. "short names" on the command line for MAIN: I expected something like -d=foo to work if I had declared sub MAIN (:$dir) , but nope. 02:04
What's the deal with short names?
psch lucs: that's not automatic, you have to declare aliases yourself 02:10
lucs psch: Okay, um, can you point me to what I need to read please?
psch m: sub f(:f(:foo($foo))) { say $foo }; f :f(2); f :foo(2) # like this 02:11
camelia rakudo-moar e0a403: OUTPUT«2␤2␤»
lucs Oho. Hmm...
psch lucs: i guess Signatures on doc.perl6.org?
lucs Okie doke. Thanks!
timotimo i'm still looking for suggestions for an advent calendar post i could write :S 02:13
github.com/perl6/mu/blob/master/mi...instorming - doesn't inspire me enough :(
psch yeah, kinda similar here... 02:15
also somewhat tied up in figuring r-j out more to get it actually working again
but that's lots of "do i actually understand how that works..?" :P
diakopter timotimo: maybe inspiring is to take some RosettaCode entries that TimToady wrote and pontificate about them 02:16
like, what does idiomaticity mean in some cases vs. performance and learnability 02:17
timotimo github.com/perl6/ye-olde/issues/2
diakopter rosettacode.org/wiki/Category:Perl_6
timotimo diakopter: i don't think i understood why you asked me to explain why the thing is bad 02:25
diakopter which content is out of date
timotimo but i put an exhaustive answer together to what i think you have meant
ah. right
diakopter (how much of it)
timotimo i don't know what other files are in there, because the front page doesn't link to them 02:26
diakopter I mean, approx how much of that page
timotimo hm. i'd say between 50% and 75% 02:27
diakopter maybe just request him to post a prominent notice that it's vastly out of date?
timotimo yeah, something like that would be good 02:28
diakopter I mentioned his ghid there
timotimo i saw. thanks!
didn't occur to me :S
so ... bedtime for me!
diakopter so unless he's unwatched/silenced that repo
timotimo good night!
diakopter he'll get an email
NITE! 02:29
TimToady m: sub f(:f(:$foo)) { say $foo }; f :f(2); f :foo(2) 02:55
camelia rakudo-moar e0a403: OUTPUT«2␤2␤»
TimToady you can simplify it a bit
diakopter blinks* 02:56
lucs TimToady: Yep, thanks. 02:57
diakopter m: sub f(:f(:$foo)) { say $foo }; f :f(2); f :foo(f); f :f((f::)==f(:f(f))) 03:01
camelia rakudo-moar e0a403: OUTPUT«2␤(Any)␤True␤(Any)␤(Any)␤True␤True␤»
diakopter speaking of ffs... 03:03
[Coke] gist.github.com/coke/f0a4c2da754b24bef2a7 - what am I missing, isn't that a match? 03:19
Quom_ You don't have a second path? 03:21
[Coke] bother. I was expecting "path".IO.chdir to DWIW, but you're right, that's wrong. 03:23
docs: The -e and -f file test operators do not exist in Perl 6. Use instead :e and :f. 03:26
Zoffix I actually see 2 Brian D Foy's Twitter accounts :s @briandfoy_perl and @briandfoy
[Coke] docs.perl6.org/type/IO::Path (shouldn't refer to the p5 way at all)
jdv79 why can't Test see an op i create? 03:28
Zoffix fwiw: twitter.com/zoffix/status/670806203077931009 though my last interaction with Brian D Foy was by far not a pleasant one...
jdv79 m: use Test; cmp-ok(4,"+",5); sub infix:<a>($a,$b){$a+$b}; say 3 a 4; cmp-ok(2,"a",3);
camelia rakudo-moar e0a403: OUTPUT«ok 1 - ␤7␤not ok 2 - ␤␤# Failed test at /tmp/GRbqKZffIw line 1␤# Could not use 'a' as a comparator␤»
jdv79 hardcore scoping 03:31
what could be some workarounds? 03:33
Zoffix m: ༤ / ༨ + ༪
camelia rakudo-moar e0a403: OUTPUT«WARNINGS:␤Useless use of "+" in expression "/ ༨ + ༪" in sink context (line 1)␤»
Zoffix m: say ༤ / ༨ + ༪
camelia rakudo-moar e0a403: OUTPUT«1␤»
Zoffix This is sick. Those chars don't even display for me right, yet Perl 6 does its thing :P 03:34
Zoffix wonders how many languages can do that
jdv79 i think i can see them physically 03:35
no idea what some are
Zoffix They just display as boxes with numbers for me: i.imgur.com/tc26vTL.png
jdv79 imgur.com/9tdMvVO 03:38
they look better here ^H 03:39
Quom_ m: use Test; cmp-ok(4,"+",5); sub infix:<a>($a,$b){$a+$b}; say 3 a 4; cmp-ok(2,&[a],3) 03:40
camelia rakudo-moar e0a403: OUTPUT«ok 1 - ␤7␤ok 2 - ␤»
jdv79 yeah, its the string arg i wsa interested in
jdv79 but that's closer. thanks.. 03:41
Quom_ I don't think that's possible with how cmp-ok is implemented right now 03:42
jdv79 its only one extra char so its what i wanted basically
Zoffix What were the names of built-in JSON conversion functions? 03:44
m: say '["foo":"bar"]'.json-decode
camelia rakudo-moar e0a403: OUTPUT«Method 'json-decode' not found for invocant of class 'Str'␤ in block <unit> at /tmp/EoEOxl8jvJ:1␤␤»
jdv79 from and to json i think 03:45
Zoffix m: say '["foo":"bar"]'.from-json
camelia rakudo-moar e0a403: OUTPUT«Method 'from-json' not found for invocant of class 'Str'␤ in block <unit> at /tmp/4MPv3ba4K0:1␤␤»
jdv79 subs i think
Zoffix m: say from-json '["foo":"bar"]' 03:46
camelia rakudo-moar e0a403: OUTPUT«Method 'ast' not found for invocant of class 'Any'␤ in block <unit> at /tmp/wVzuDIrhiF:1␤␤»
Zoffix which is weird, 'cause: 03:47
m: say to-json ['42']
camelia rakudo-moar e0a403: OUTPUT«[␤ "42"␤]␤»
Zoffix And I see a bunch of mention of from-json here: github.com/rakudo/rakudo/search?ut...amp;q=json 03:47
A module it is! :P
Zoffix is still bothered native to-json and some JSON modules have all that extra whitespace in generated JSON 03:48
Quom_ Zoffix: Does JSON::Fast have extra ws?
Zoffix Yup. Does same thing as above 03:49
3 useless bytes
jdv79 use JSON::XS 03:51
Zoffix no such thing modules.perl6.org/repo/JSON::XS
Zoffix pets the /repo/ feature 03:52
jdv79 :| 03:53
grondilu trying to implement exponentiation by squaring. Could not get make it faster than the current &[**] : gist.github.com/grondilu/ba900191b018daf5cded 03:54
lucs How can I capture stderr from a run() command? 03:56
Zoffix lucs, perl6 -e 'my $proc = run "perl", "-wlE", "warn q|Hallo world|", :err; my $captured-output = $proc.err.slurp-rest; say $captured-output;' 04:02
lucs, this doc might be helpful: doc.perl6.org/type/Proc
lucs Zoffix: Hmm... Tried something similar, unsuccessfully (yep, was reading that .../Proc doc). 04:04
Trying again...
lucs Aha. I was building the command to run in @cmd, with ":err" pushed at the end. Like that, "run @cmd" fails. But if I don't push it and do "run @cmd, :err", it works. 04:08
lucs Thanks Zoffix. 04:08
Zoffix \o/ 04:08
lucs You bet :) I'm really loving Perl6.
Zoffix Yeah, it's neat. 04:09
lucs Understatement my friend :)
dalek kudo/nom: e3c591a | TimToady++ | src/core/ (3 files):
make tolerance relative to operands when practical

Also, rename $*SIGNIFICANCE to $*TOLERANCE.
04:12
Zoffix m: my $z = [42]; say $z<42> 04:15
camelia rakudo-moar e0a403: OUTPUT«Type Array does not support associative indexing.␤ in block <unit> at /tmp/882iGcZajP:1␤␤Actually thrown at:␤ in block <unit> at /tmp/882iGcZajP:1␤␤»
ShimmerFairy TimToady: we have a ≅ operator now? I can't help but wonder how much more will find its way into the core setting :P
Zoffix ^ would be nice if that error showed file/line number
dalek ast: 190ed77 | TimToady++ | S (3 files):
test scaled tolerance on =~= and <=>
04:16
grondilu TimToady: I like TOLERANCE better, though I would have expected the name to refer to numbers somehow, like $*NUMERIC-TOLERANCE
ShimmerFairy Zoffix: it does, the "/tmp/garble:1" bit 04:17
TimToady TOLERANCE is a very common term in physical measurements, and always refers to numbers there
ShimmerFairy I guess the mathiest name would be EPSILON or Ε , but I'm not sure if that would fit as well here :) 04:18
grondilu true
Zoffix ShimmerFairy, not really: i.imgur.com/80lJfpL.png
flussence
.oO( what's the texas version of ≅? :)
oh, =~=
ShimmerFairy Zoffix: oh, maybe camelia has --ll-exception enabled?
Zoffix If I had a 1,000,000-line script, that would've been fun :)
ShimmerFairy Zoffix: errors that are thrown from NQP or the underlying VM don't display diagnostic info unless you use --ll-exception 04:19
grondilu m: $*TOLERANCE = 1e-6; say 10 =~= 10 + 1e-7
camelia rakudo-moar e0a403: OUTPUT«Dynamic variable $*TOLERANCE not found␤ in block <unit> at /tmp/hicMWVsRfq:1␤␤Actually thrown at:␤ in block <unit> at /tmp/hicMWVsRfq:1␤␤»
TimToady floating-point errors are a lot bigger than ε
Zoffix Maybe. I get a ton of output with that option i.imgur.com/U8iOfqU.png 04:20
grondilu m: $*SIGNIFICANCE = 1e-6; say 10 =~= 10 + 1e-7
camelia rakudo-moar e0a403: OUTPUT«Cannot modify an immutable Num␤ in block <unit> at /tmp/NhZFUVTtQ9:1␤␤»
TimToady wait till :31 or so
can't override PROCESS:: vars, like that, needs a 'my'
otherwise one interpreter could modify the PROCESS value for a different interpreter 04:21
grondilu just realises that =~= is NYI
TimToady sure it is
grondilu oh, I'm confused then
TimToady you need a 'my'
ShimmerFairy TimToady: sure, but I understand "epsilon" to be a name you see sometimes for a limit on how precise you want to be (so I guess in that context an "epsilon" argument means "pretend epsilon is this")
grondilu m: my $*SIGNIFICANCE = 1e-6; say 10 =~= 10 + 1e-7
camelia rakudo-moar e0a403: OUTPUT«True␤»
Zoffix ShimmerFairy, bummer... I, as an end user, would rather not even know what "NQP" is and get a proper error from the get-go. Stuff like my Int Int; does provide useful feedback after-all: i.imgur.com/EScHZtU.png
TimToady to me, ε is an infinitesimal
grondilu m: my $*SIGNIFICANCE = 1e-6; say 10 =~= 10 + 9e-7 04:21
camelia rakudo-moar e0a403: OUTPUT«True␤»
ShimmerFairy Zoffix: yeah, I think we all agree that cases requiring --ll-exception are an issue in error reporting, or else in one of the lower layers :) 04:23
grondilu shouldn't it be a relative value (dimensionless) ? sub infix:<=~=>($a, $b) { 2*abs($a - $b) / sqrt($a**2 + $b**2) } or something?
Zoffix :)
grondilu sorry
sub infix:<=~=>($a, $b) { 2*abs($a - $b) / sqrt($a**2 + $b**2) < $*TOLERANCE } 04:24
ShimmerFairy TimToady: yeah, so I think in those places I've seen they're really saying "pretend we're in a world where ε (infinitesimal) is known to be 1e-6, because I really don't care to be more precise than that"
grondilu m: my $*SIGNIFICANCE = 1e-6; say 1e8 =~= 1e8 + 9e-7 04:25
camelia rakudo-moar e0a403: OUTPUT«True␤»
grondilu m: my $*SIGNIFICANCE = 1e-6; say 1e8 =~= 1e8 + 9e-6
camelia rakudo-moar e0a403: OUTPUT«False␤»
Quom_ m: enum A (B=>10, C=11); # LTA
camelia rakudo-moar e0a403: OUTPUT«===SORRY!===␤Cannot invoke this object␤»
TimToady was trying to avoid anything so geometric, so it just scales the larger of the two arguments (in abs)
ShimmerFairy Quom_: you also can't do enum A «B :C<B>» (alias enum values) :< 04:26
Quom_ ShimmerFairy: Can't one just declare constant C = B ?
grondilu absolute tolerance does make some sense as well though, so I don't know if a relative one really is better. 04:27
TimToady if they're close enough to be inside the tolerance, it won't make a lot of difference
ShimmerFairy Mouq: not if you want to avoid the scourge of unscoped enum values, which P6 also doesn't let you avoid :P (P6 takes a hybrid approach when compared to, say, C++)
TimToady you can easily get absolute by $a - $b ≅ 0 04:28
TimToady .tell masak see irclog.perlgeek.de/perl6/2015-11-29#i_11620750 04:29
yoleaux TimToady: I'll pass your message to masak.
grondilu just checked the code for &[=~=] and thinks it's fine indeed. 04:30
(allowing relative and absolute comparison) 04:31
Mouq Do we have mashalling from Array[num] to CArray[num]? 04:34
TimToady hmm, Complex.narrow should probably also pay attention to the tolerance
jdv79 why ≅ and not ≈? 04:35
flussence too confusable with smartmatch, I guess 04:36
Zoffix m: say 3.1 ≈ 3; say 3.000000000000001 ≈ 3;
camelia rakudo-moar e3c591: OUTPUT«5===SORRY!5=== Error while compiling /tmp/6QNuKA1xZs␤Confused␤at /tmp/6QNuKA1xZs:1␤------> 3say 3.17⏏5 ≈ 3; say 3.000000000000001 ≈ 3;␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement…»
Zoffix There's a whole bunch of them en.wikipedia.org/wiki/Equals_sign#...tely_equal
ShimmerFairy jdv79: I'm guessing among other things, 1) What's the texas version gonna be? Can't be ~~ , and 2) it's named APPROXIMATELY EQUAL TO :) 04:37
m: say "≈".uniname; say "≅".uniname
camelia rakudo-moar e3c591: OUTPUT«ALMOST EQUAL TO␤APPROXIMATELY EQUAL TO␤»
grondilu is not sure what's the difference between "almost equal" and "approximately equal" 04:41
jdv79 why can a routine be called ± but not a param? 04:42
grondilu m: sub prefix:<±>($n) { $n | -$n }; say -pi == ±pi 04:43
camelia rakudo-moar e3c591: OUTPUT«any(False, True)␤»
grondilu m: sub ±($n) { $n | -$n }; say -pi == ±(pi) 04:44
camelia rakudo-moar e3c591: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Wrwwbahwh6␤Missing block␤at /tmp/Wrwwbahwh6:1␤------> 3sub7⏏5 ±($n) { $n | -$n }; say -pi == ±(pi)␤ expecting any of:␤ new name to be defined␤»
grondilu I dn't think a routine can be called ±
unless you make it a term or something
jdv79 i meant an op ^H 04:45
grondilu well, identifier rules for ops are quite different than for variables, aren't they?
you can pretty much use anything for an op 04:46
(even spaces IIRC)
TimToady m: sub infix:<±>($mean, $n) { $mean + (-$n .. $n) }; say 22/7 == pi ± .1
camelia rakudo-moar e3c591: OUTPUT«False␤»
TimToady m: sub infix:<±>($mean, $n) { $mean + (-$n .. $n) }; say 22/7 == pi ± .5
camelia rakudo-moar e3c591: OUTPUT«False␤»
TimToady m: sub infix:<±>($mean, $n) { $mean + (-$n .. $n) }; say pi ± .5 04:47
camelia rakudo-moar e3c591: OUTPUT«2.64159265358979e0..3.64159265358979e0␤»
TimToady m: sub infix:<±>($mean, $n) { $mean + (-$n .. $n) }; say 22/7 ~~ pi ± .1
camelia rakudo-moar e3c591: OUTPUT«True␤»
TimToady dur
grondilu did not know that semantic for $num + $range, is that new?
TimToady several weeks old 04:48
[Coke] SHould IO::Path.chdir impact the cwd for shell? (right now I have to use $*CWD = ....) 04:50
ShimmerFairy jdv79: you're modifying the parser when you declare an op, so you're basically just writing a rule that checks for ± (and the reason for that is pretty much because operators typically want symbols over traditional identifiers :P) 04:51
MadcapJake Check out this GIF of the latest work I've done for linter-perl6 twitter.com/MadcapJake/status/6708...6390053889 05:09
jdv79 can an op not be multi? 05:11
zostay_ m: say |(1,2,3).race.map(* + 1) 05:14
camelia rakudo-moar e3c591: OUTPUT«234␤»
zostay m: say |(1,2,3).race.grep(* == 3) 05:14
camelia rakudo-moar e3c591: OUTPUT«␤»
zostay m: say |(1,2,3).race.map(* + 1).grep(* == 3) 05:15
camelia rakudo-moar e3c591: OUTPUT«3␤»
zostay i don't understand why i can map a HyperSeq, but I can't seem to grep it
grondilu MadcapJake: that's pretty cool. I wonder if that'd be possible in Vim 05:18
zostay or at least, i can't grep until i've mapped it once...
TimToady zostay: that's a known bug 05:19
zostay oh, good... i thought i was just confused ;)
TimToady at least, I think it is... 05:20
geekosaur found Atom troublesome... left it open overnight and it made everything else swap out
jdv79 gist.github.com/anonymous/0838392ab4b6f943e7ad 05:21
neat that that is possible and not hard
jdv79 how come i seem to lose my precedence change when i put that into a module and export them ? 05:55
skids $ perl6 -e '"/tmp/ffffjjjj".IO.unlink; my $fh = open("/tmp/ffffjjjj", :create); say so $fh.path ~~ :f, so "/tmp/ffffjjjj".IO ~~ :f;' # FalseTrue ... :-/ 06:00
MadcapJake grondilu: no idea how to program vim plugins 06:07
geekosaur: not sure what you mean by "swap out" or "everything else" 06:08
jdv79 m: sub infix:<o1> ($a, $b) is export { say 1 }; sub infix:<o2> ($a, $b) is tighter(&infix:<o1>) is export { say 2 }; 1 o1 1 o2 1; # ok 06:09
camelia rakudo-moar e3c591: OUTPUT«2␤1␤»
jdv79 m: module m { sub infix:<o1> ($a, $b) is export { say 1 }; sub infix:<o2> ($a, $b) is tighter(&infix:<o1>) is export { say 2 } }; import m; 1 o1 1 o2 1; # what happened?
camelia rakudo-moar e3c591: OUTPUT«1␤2␤»
TimToady looks like a bug to me 06:19
TimToady m: module m { sub infix:<o1> ($a, $b) is export { say 1 }; sub infix:<o2> ($a, $b) is tighter(&infix:<o1>) is export { say 2 } }; import m; say &infix:<o1>.prec,' ',&infix:<o2>.prec 06:20
camelia rakudo-moar e3c591: OUTPUT«prec => t= prec => t=␤»
TimToady yeah definitely some kind of bug 06:21
is the export aliasing more indirect than I thought?
m: module m { sub infix:<o1> ($a, $b) is export { say 1 }; sub infix:<o2> ($a, $b) is tighter(&infix:<o1>) is export { say 2 }; say &infix:<o1>.prec }; import m; say &infix:<o1>.prec,' ',&infix:<o2>.prec 06:22
camelia rakudo-moar e3c591: OUTPUT«prec => t=␤prec => t= prec => t=␤»
TimToady m: module m { sub infix:<o1> ($a, $b) is export { say 1 }; sub infix:<o2> ($a, $b) is tighter(&infix:<o1>) is export { say 2 }; say &infix:<o2>.prec }; import m; say &infix:<o1>.prec,' ',&infix:<o2>.prec
camelia rakudo-moar e3c591: OUTPUT«prec => t=␤prec => t= prec => t=␤»
TimToady looks like tighter isn't working 06:23
even inside
m: module m { sub infix:<o1> ($a, $b) is export { say 1 }; sub infix:<o2> ($a, $b) is tighter(&[+]) is export { say 2 }; say &infix:<o2>.prec }; import m; say &infix:<o1>.prec,' ',&infix:<o2>.prec
camelia rakudo-moar e3c591: OUTPUT«prec => t=␤prec => t= prec => t=␤»
TimToady huh
m: module m { sub infix:<o1> ($a, $b) is export { say 1 }; sub infix:<o2> ($a, $b) is equiv(&[*]) is export { say 2 }; say &infix:<o2>.prec }; import m; say &infix:<o1>.prec,' ',&infix:<o2>.prec 06:24
camelia rakudo-moar e3c591: OUTPUT«assoc => left, prec => t=␤prec => t= assoc => left, prec => t=␤»
TimToady m: module m { sub infix:<o1> ($a, $b) is export { say 1 }; sub infix:<o2> ($a, $b) is prec("u=") is export { say 2 }; say &infix:<o2>.prec }; import m; say &infix:<o1>.prec,' ',&infix:<o2>.prec 06:25
camelia rakudo-moar e3c591: OUTPUT«5===SORRY!5=== Error while compiling /tmp/1rZR8Dos5N␤Can't use unknown trait 'is prec' in a sub+{precedence} declaration.␤at /tmp/1rZR8Dos5N:1␤ expecting any of:␤ rw raw hidden-from-backtrace hidden-from-USAGE␤ cached pure de…»
TimToady well, feel free to RT it
Mouq Does Panda not precomp modules anymore? 06:42
ShimmerFairy Mouq: it does not, in fact. (It's supposed to be rakudo's job anyhow) 06:47
Mouq Oh, wait, is that the CURLI branch you reference? 07:00
ShimmerFairy Mouq: yeah, I think the work is going in there, but precomp was always supposed to be rakudo's business. The support was removed from panda to get a push going in implementing it (as well because it was problematic to maintain) :P 07:15
dalek kudo/nom: bc027e2 | TimToady++ | src/core/Complex.pm:
use 0e0 instead of 0 when approximating 0
07:25
kudo/nom: 3807f26 | TimToady++ | src/core/Num.pm:
.narrow is approximate on Complex and Num
ast: cb4dae7 | TimToady++ | S32-num/narrow.t:
test that .narrow approximates on Complex and Num
TimToady m: say exp(i * pi).narrow 07:30
camelia rakudo-moar e3c591: OUTPUT«-1+1.22464679914735e-16i␤»
TimToady m: say exp(i * pi).narrow 07:31
camelia rakudo-moar 3807f2: OUTPUT«-1␤»
TimToady m: say (.1 + .2) * 10 07:33
camelia rakudo-moar 3807f2: OUTPUT«3␤»
TimToady m: say (.1 + .2) * 10 == 3 07:34
camelia rakudo-moar 3807f2: OUTPUT«True␤»
TimToady m: say (.1 + .2 - .3) * 10 == 0
camelia rakudo-moar 3807f2: OUTPUT«True␤»
TimToady oops
m: say (.1e0 + .2) * 10 == 3
camelia rakudo-moar 3807f2: OUTPUT«False␤»
TimToady m: say ((.1e0 + .2e0) * 10).narrow == 3
camelia rakudo-moar 3807f2: OUTPUT«True␤»
TimToady I guess I did the test right, anyway... 07:35
llfourn_ what's the specific goal for the december release? 17 is in my mind but cannot find citation.. 07:36
s/goal/date/ s/17/17th/ 07:37
TimToady we're aiming for the 17th so that we'll be sure to hit the 25th :) 07:38
llfourn @TimToady: cheers :) 07:40
HoloIRCUser1 Hello 07:52
grondilu m: say i**2 == -1; # seriously this is bad 08:26
camelia rakudo-moar 3807f2: OUTPUT«False␤»
grondilu writes a patch
zaucker_ Goof morning ... I just thought I could try teaching programming to my 9 your old daughter. And while at it could teach myself Perl6. Are there any online resources about teaching kids? Preferably Perl6 or 5, but I guess others would do as well. 08:28
stmuk_ perl6intro.com
zaucker_ Well, yes. I was wondering if there is something specifically targeted at children/girls. 08:29
Mouq zaucker_: I think most likely nowadays you'll find the most kid-friendly tutorials/programming introductions for javascript
stmuk_ there isn't yet 08:32
Hotkeys grondilu: isn't that bad in the same way that 08:36
m: say 0.1e0 + 0.2e0 == 0.3e0
camelia rakudo-moar 3807f2: OUTPUT«False␤»
Hotkeys is bad
i vote for a ComplexRat type
JimmyZ m: say 0.1e0 + 0.2e0 =~= 0.3e0 08:37
camelia rakudo-moar 3807f2: OUTPUT«True␤»
grondilu I still think it's as bad as if 2e0**2 !== 4 08:39
m: say 2e0**2 == 4 08:40
camelia rakudo-moar 3807f2: OUTPUT«True␤»
grondilu I once whined about it and I forgot why you convinced me it was a bad idea, but I still don't get why we have to base Complex on num. 08:41
adding a ComplexRat would make things even more confusing imho. Maybe a parametric role though? 08:42
but truly if it was me I'd define Complex as : class Complex does Numeric { has Real ($.re, $.im); ... } 08:43
skids Hotkeys: floating points are an established mechanism that needs to be supported in a modern language, warts and all. Or you cannot interface to things that use them.
JimmyZ m: say 0.1e0 + 0.2e0 08:44
camelia rakudo-moar 3807f2: OUTPUT«0.3␤»
Hotkeys m: say 0.1e0 + 0.2e0 - 0.3e0
camelia rakudo-moar 3807f2: OUTPUT«5.55111512312578e-17␤»
Hotkeys skids: I know
skids (but I'm indifferent on Complex) 08:46
grondilu or maybe doing something similar than for Rational? Les seuls films que j'ai vus où les billets était durs à avoir étaient le premier Harry Potter et le premier Seigneur des Anneaux. 08:48
ooos
oops
wrong paste
my role Complex[::R] does Numeric { method re returns ::T {...} }; 08:49
moritz grondilu: the design docs want it that way, it just isn't implemented 08:50
grondilu oh ok
masak good antenoon, #perl6 09:09
yoleaux 04:29Z <TimToady> masak: see irclog.perlgeek.de/perl6/2015-11-29#i_11620750
masak TimTolerance++ 09:09
nine_ I think, I'm gonna merge curli today 09:11
moritz \o/ 09:12
nine_ I'd feel much more comfortable if it weren't for those ephemeral spectest failures. But since I have them no nom, too, I just have no idea if curli makes them worse or not. 09:17
masak curli has something to do with module, no? 09:19
modules*
nine_ yes
masak what will be different after curli is merged? 09:22
moritz stuff will precompile! 09:23
masak ooh 09:24
at no risk to life or limb? that's good news indeed!
nine_ That's the intention, yes ;) 09:25
And we can install as many modules as we like without affecting startup time
I've already tried it with > 100 modules 09:26
grondilu installation will still happen through panda?
JimmyZ or zef
nine_ Package managers become a bit simpler though since CompUnit::Repository::Installation manages all precompilation and the actual copying of files. 09:28
grondilu will it allow for uninstallation of modules? 09:29
nine_ NYI but certainly possible. 09:30
grondilu install linenoise and regrets it as it is painfully slow to start 09:31
lizmat good *, #perl6!
grondilu *installed
(though maybe now with precompilation it will be quicker)
nine_ grondilu: not just maybe :)
dalek kudo/nom: b464aca | LLFourn++ | src/Perl6/ (2 files):
avoid installing %?LANG too early

  %*LANG was copied to %?LANG too early causing EVAL not work sometimes
09:32
kudo/nom: 6153f1c | lizmat++ | src/Perl6/ (2 files):
Merge pull request #603 from LLFourn/simple_lang_eval_fix

avoid installing %?LANG too early
lizmat wonders whether we should remove all macro support before 6.c 09:48
JimmyZ well, some modules depends on macro 09:49
lizmat but I think we agree that the current macro's won't survive, right? 09:56
so better to stop them short now before we get more modules depending on them?
lizmat BTW, is anybody on channel going to the CCC between Xmas and NewYear this year ?? 09:57
moritz oo-monitors uses a macro 10:01
seems to be the only use in the ecosystem 10:02
lizmat: not me (re CCC)
lizmat would be nice if someone could give a Perl 6 presentation there
JimmyZ m: use MONKEY-TYPING; my $d = ""; augment class Str { method some_meth_1 {$d = $d ~ self;}}; for '0' .. '5' { .WHAT.say; sub; }; 10:05
camelia rakudo-moar 6153f1: OUTPUT«5===SORRY!5=== Error while compiling /tmp/kD4DPX0_SP␤A unit-scoped sub definition is not allowed except on a MAIN sub;␤Please use the block form.␤at /tmp/kD4DPX0_SP:1␤------> 3elf;}}; for '0' .. '5' { .WHAT.say; sub;7⏏5 };␤»
masak is there any compelling reason I should put `use v6;` at the top of my modules? 10:16
dalek kudo/nom: ff81e96 | lizmat++ | src/core/Exception.pm:
Make sure unexpected/nogo are always sorted
ast: ffd8df7 | lizmat++ | S32-str/split.t:
Fix test now that they're always sorted
10:17
JimmyZ masak: hightlighting? 10:18
lizmat masak: sane error message when run with perl 5 ? 10:19
dalek ast: e597ff3 | lizmat++ | S06-signature/closure-parameters.t:
Unfudge now passing test
10:22
nine_ Should this worry me? gist.github.com/niner/50bd135b0747cf2f36fe 10:23
dalek ast/curli: 2ba7c8a | (Stefan Seifert)++ | / (2 files):
Fix integration/precompiled.t for real

These tests for precompilation issues passed because we didn't actually precompile anything. Fixed now by using some trickery to force precompilation despite adding a new repo after loading the first module.
10:27
pmurias hi 10:33
RabidGravy masak, also potentially enable e.g. "file" to recognise Perl 6 modules 10:35
dalek kudo/nom: 2897239 | lizmat++ | src/core/Str.pm:
Add Str.NFC/NFD/NFKC/NFKD as NYI stubs for JVM

Should at least improve error message on RT #126678
10:37
[Tux] test 50000 24.509 24.396 10:45
test-t 50000 21.298 21.185
csv-parser 50000 29.335 29.221
dalek ast: 6af57be | peschwa++ | S06-signature/closure-parameters.t:
Add test for RT #125988.
10:54
psch hmm 10:55
that test file has a few skipped tests which are intriguing
bartolin_ psch: do you think github.com/rakudo/rakudo/pull/598 is a good way to avoid the new NPE on rakudo.jvm? 10:57
psch bartolin_: well, yes. it seems a good way to avoid the NPE 11:00
bartolin_: the problem is that underlying behavior seems different, though
bartolin_: as in, moar is probably right in not-dying with an atkey with a non-existing key 11:01
psch although there's probably more involved than that... nqp-j can deal with that as well 11:02
i'll have a closer look what exactly NPEs there
bartolin_ psch: ok, so it would be more like a bandaid for r-j.
psch++
lizmat fwiw, I think all nqp code should always do an nqp::existkey first before being able to do an nqp::atkey 11:03
psch lizmat: it's definitely safer, and probably even cheaper, yeah 11:04
dalek p/js-cps: 175c6ed | (Pawel Murias)++ | src/vm/js/nqp-runtime/core.js:
[js] Make nqp::continuationcontrol return the result of the callback passed to nqp::continuationinject.
11:07
p/js-cps: 3cbb936 | (Pawel Murias)++ | src/vm/js/nqp-runtime/ (3 files):
[js] Work on passing the continuations test.

Continuationreset returns the right value. Continuations are true for nqp::istrue.
psch i just don't know if we can allow ourselves the luxury of (1) having a "standard" backend that is permissive and (slightly) DWIM-y but (2) still expect people to do stuff the hard and rigorous way
see also the getattr vs getattr_{i,n,s} hiccup
bartolin_ psch: fwiw, for three days there are about 5000 test failures on r-j due to aborted tests with that NPE: github.com/usev6/perl6-roast-data/..._rates.csv 11:11
psch r: class A { has int $.a }; say nqp::getattr(A.new, A, '$!a').WHAT # this is the getattr hiccup i'm refering to
camelia rakudo-jvm e3c591: OUTPUT«===SORRY!===␤java.lang.NullPointerException␤»
..rakudo-moar 289723: OUTPUT«(int)␤»
psch lo and behold, another NPE :| 11:12
psch r: my constant 𝑒 := e; 11:13
camelia ( no output )
..rakudo-jvm e3c591: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Missing initializer on constant declaration␤at /tmp/tmpfile:1␤------> 3my constant7⏏5 𝑒 := e;␤»
psch grrr
bartolin_ psch: (different behaviour of backends) I wondered about that, too. would be interesting to hear jnthn++'s opinion
psch well, the NPE for getattr was a "No such Attribute" somewhen earlier... :/ 11:14
and i suspect the misparse for 𝑒 is something about uniprop..?
bartolin_: well, in general i think we want backends to behave identically. C just can do a few things java can't easily is what jnthn said when i asked him about the getattr mismatch 11:16
i didn't yet look closer at that
that 𝑒 misparse prevents r-j build, btw :/
i guess i'll #?if moar that locally for now... 11:17
jnthn nine_: (merge day) yay :) 11:19
nine_ jnthn: gist.github.com/niner/50bd135b0747cf2f36fe worries me...a lot :( 11:22
jnthn: it's the only thing preventing me from merging
jnthn Is that a hang?
nine_ yes
jnthn urgh 11:23
Can you get a backtrace with line numbers?
nine_ Seems to happen only on curli and just when I try to run spectests with TEST_JOBS=1 11:23
Running with TEST_JOBS=12 seems to be _more_ stable which is totally insane
jnthn o.O 11:24
nine_ I'll recompiled with -g
psch r: my \f := nqp::create(Int); say so f 11:27
camelia rakudo-jvm e3c591: OUTPUT«===SORRY!===␤java.lang.NullPointerException␤»
..rakudo-moar 289723: OUTPUT«False␤»
jnthn That should really work out...
psch bartolin_: ^^^ that's kind of the reason behind needing the existskey
as in, it's the same failure mode for a related reason, i suppose 11:28
bartolin_ j: my $f := nqp::create(Int); say so f ## guess, that works
camelia rakudo-jvm e3c591: OUTPUT«5===SORRY!5=== Error while compiling /tmp/CLSxIDC97C␤Undeclared routine:␤ f used at line 1␤␤»
bartolin_ j: my $f := nqp::create(Int); say so $f ## guess, that works
camelia rakudo-jvm e3c591: OUTPUT«===SORRY!===␤java.lang.NullPointerException␤»
psch r: my %h; say so nqp::atkey(%h, 'foo') # hah 11:29
camelia rakudo-moar 289723: OUTPUT«(signal SEGV)»
..rakudo-jvm e3c591: OUTPUT«===SORRY!===␤java.lang.NullPointerException␤»
psch they agree \o/
:P
bartolin_ *g*
psch r: my %h; say so nqp::atkey(nqp::decont(%h), 'foo') # at least it wasn't the missing decont... 11:32
camelia rakudo-jvm e3c591: OUTPUT«===SORRY!===␤java.lang.NullPointerException␤»
..rakudo-moar 289723: OUTPUT«(signal SEGV)»
psch bartolin_: in any case, i think that means your PR should probably be merged
dalek kudo/nom: 0718179 | lizmat++ | src/core/io_operators.pm:
Make MMD setup of sub print same as say/note/put

This should fix RT #126725, but causes some spectest fallout. However, I do agree that this is the right way to go!
11:34
hankache hi 11:36
psch r: my \h = {}; say so nqp::atkey(h, 'foo') # this is more like the case in gistseen, i suppose... 11:37
camelia rakudo-moar 289723: OUTPUT«False␤»
..rakudo-jvm e3c591: OUTPUT«===SORRY!===␤java.lang.NullPointerException␤»
nine_ jnthn: gist.github.com/niner/39a4084d7e1de469d730 with MoarVM 623eadf4f67ffc0f73ce30da4ed0d7c8b5bb0296
dalek ast: d1d79f7 | LLFourn++ | S (3 files):
tests EVAL with %*LANG modifications
11:38
ast: 600579f | LLFourn++ | S (3 files):
Merge pull request #80 from LLFourn/eval_tests

tests EVAL with %*LANG modifications
jnthn oops 11:39
Wait, huh...there's actually code in there that's meant to prevent this exact bug. 11:41
dalek ast: 78d7b2f | lizmat++ | S16-io/print.t:
Bring basic print tests inline with new MMD setup
11:42
ast: abba1e3 | lizmat++ | S19-command-line/dash-e.t:
Bring basic -e tests inline with new MMD setup
lizmat that should fix the roast fallout github.com/rakudo/rakudo/commit/07181796e4
jnthn nine_: Can you try applying gist.github.com/jnthn/c42ed1caa3f7ff0ac073 and see if we hit that oops? 11:44
nine_ jnthn: gladly 11:45
psch hrm, nqp::create doesn't work out on r-j because we don't have initialize() for any REPR 11:51
psch that's nice grunt work, if anyone wants to keep busy :P 11:53
nine_ jnthn: no success :/ gist.github.com/niner/dc58ac9573d2eacdce9a 11:55
jnthn Eek 11:57
psch ...i'm not sure i understand create in interp.c correctly 11:58
as in, MVMHash for example has initialize(tc), but no initialize(tc, st, root, data) 11:59
jnthn nine_: It's really bizzare. It's clearly allocating in the nursery.
psch but with what's returned from initialize(tc) it should still call the second function..?
jnthn And clearly doing it from initialize.
jnthn But: 12:00
MVM_gc_allocate_gen2_default_set(tc);
MVM_repr_init(tc, (MVMObject *)sc);
MVM_gc_allocate_gen2_default_clear(tc);
That should be forcing allocation into gen2
Does the debugger let you display tc->allocate_in_gen2?
Ah, hang on... 12:03
jnthn nine_: Please try this patch: gist.github.com/jnthn/cb85357d35af59d1d004 12:04
psch oh, yes, i misunderstood the create op in interp.c... 12:06
nine_ allocate_in_gen2 is 0 12:07
trying the second patch now 12:08
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 942344c | (Zoffix Znet)++ | bin/build-project-list.pl:
Add --db-file option for easier testing
12:09
RabidGravy m: class B {}; class C { has B @.c }; my @bs = (B.new, B.new); my %h; %h<c> = @bs; C.new(|%h) 12:15
camelia rakudo-moar 071817: OUTPUT«Type check failed in assignment to @!c; expected B but got Array␤ in block <unit> at /tmp/hecYyPnOaX:1␤␤»
RabidGravy has that ^ changed in the last week or so? It's broken JSON::Unmarshal 12:16
jnthn Yes; in the fix to align attribute initialization semantics with variable assignment semantics.
m: class B {}; my B @c; my @bs = (B.new, B.new); my %h; %h<c> = @bs; @c = %h<c>; 12:17
camelia rakudo-moar 071817: OUTPUT«Type check failed in assignment to @c; expected B but got Array␤ in block <unit> at /tmp/BbApNSrwbV:1␤␤»
jnthn It's 'cus a Hash is a bunch of Scalar containers
m: class B {}; my B @c; my @bs = (B.new, B.new); my %h; %h<c> = @bs; @c = @(%h<c>);
camelia ( no output )
jnthn m: class B {}; my B @c; my @bs = (B.new, B.new); my %h; %h<c> := @bs; @c = %h<c>; 12:18
camelia ( no output )
jnthn Those are two ways to deal with it
RabidGravy cool, let's see, I think it's going to be the bind 12:19
RabidGravy yep that works 12:22
psch well, the nqp::create issue probably can be solved by generating an initialize method in P6Opaque.generateJVMClass, checking for that in Ops.create and calling it if it exists 12:25
except if the REPR itself has to know whether there are attrs that have to be initialized...
RabidGravy cheers jnthn
masak so, if I've already settled syntax highlighting to my satisfaction, and I don't tend to accidentally run `perl` instead of `perl6` -- rather the reverse, actually -- I can safely leave out `use v6;` from my modules? 12:27
jnthn psch: Is there more to do in rt.perl.org/Ticket/Display.html?id=125988 ?
masak I wrote a little coverage tool, and that tool is (correctly) telling me that I can omit those lines.
jnthn masak: A Perl 6 script that doesn't declare what version its targeting will automatically be targeting the latest the compiler supports. 12:28
jnthn masak: use v6; is equivalent to use v6.c; 12:29
psch jnthn: i don't think so, no. i don't have bug admin rights, fwiw
(which i'm fine with, for the record :P )
jnthn psch: OK, I can mark it resolved :) 12:31
Down to 24 xmas RTs. :) 12:33
jnthn m: say "{24 / 18} a day if we want them done by Dec 17th :)" 12:34
camelia rakudo-moar 071817: OUTPUT«1.333333 a day if we want them done by Dec 17th :)␤»
dalek kudo/nom: 2923369 | lizmat++ | src/core/Cool.pm:
Make any bigint for pos a failure for index/rindex

This further evades the difference between MoarVM and JVM for nqp::unbox_i of bigints (fails on MoarVM, silently truncates on JVM)
12:35
jnthn nine_: I pushed the two patches; I think the first one is a good sanity check to have anyway, and the second is necesary even if not sufficient. 12:38
masak jnthn: I can see the benefit of pinning down the exact Perl 6 version, yes. but (as you point out) `use v6;` alone doesn't do that, so it's no more specific than leaving it out. 12:39
jnthn masak: I said it's equivalent to use v6.c; :) 12:40
dalek ast: db02e90 | lizmat++ | S32-str/ (2 files):
Fix index/rindex tests
12:41
jnthn masak: So it is pinning it down.
masak jnthn: oh! 12:42
jnthn: this is.. some kind of special case for v6.c, then?
jnthn masak: Mostly just identifying that 6.c is the first stable version of the language 12:44
masak *nod* 12:48
jnthn lunch &
masak ok, so leaving it out has the semantics of "I implicitly expect to run on whatever's the latest Perl 6 version"
nine_ jnthn: \o/ 12:50
RabidGravy tadzik, when you have a minute be a love and merge the github.com/tadzik/JSON-Unmarshal/pull/13 it fixes the above ;-) 12:51
nine_ jnthn: spectest ran without a single hang :) Only see the failures that I have with nom, too. 12:52
diakopter nine_: test jobs 1 ?
nine_ diakopter: yes 12:53
lucs Pod/MAIN question: fpaste.scsys.co.uk/502136 12:56
tadzik Type check failed in assignment to @!dogs; expected Dog but got Array 12:57
*cracks up*
where can I read up on the change? 12:58
lucs (oops, "pod1.pl6" should be "mainpod.pl6") 13:00
ShimmerFairy lucs: I believe #= is a trailing declarator block, and that you'd want #| 13:03
lucs Ah, thanks, trying...
Hmm... No luck, the message is shorter: "Usage \n pod1.pl6 <foo> <bar>" 13:04
diakopter from where can I install Debugger::UI::CommandLine 13:05
I try to run perl6-debug and it says I need this library installed
do I need a working panda to get this library? 13:06
ShimmerFairy lucs: I think the USAGE prints MAIN's declarator block if given, and not the arguments, but I wouldn't know about that for sure (which is why it printed the first message before, it attached to MAIN) 13:07
lucs ShimmerFairy: Yeah, makes sense.
But unfortunate, eh.
ShimmerFairy diakopter: yep, your rakudo repo is actually incomplete unless you install some external modules. yaaaaay 13:07
diakopter ShimmerFairy: I didn't know that; are you serious? 13:08
ShimmerFairy diakopter: yep, last I checked you need Debugger::UI::CommandLine, Inline::P5, and Linenoise to use everything in the rakudo repo.
I've complained about this horrendous flaw before :) 13:09
diakopter do you know where I can get Debugger::UI::CommandLine ?
ShimmerFairy via panda, AFAIK. Have to check modules.perl6.org to see where the repo is if you need to install it manually. 13:10
diakopter: github.com/jnthn/rakudo-debugger/ 13:11
diakopter the makefile there is for parrot 13:12
ShimmerFairy I'm afraid I'm only aware of ex-parrots :P (but in more seriousness, that's what "Debugger::UI::CommandLine links to from the modules page, so I guess ignore the makefile??) 13:13
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: c5dd6f1 | (Zoffix Znet)++ | bin/build-project-list.pl:
Fix bug with db-file option being ignored
13:14
href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 8bc3ecc | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder.pm:
Output log info on what logotype dir is used for the build
diakopter nine_: is your new panda ready? 13:15
Zoffix RE: "Have to check modules.perl6.org to see where the repo is" FWIW this works: modules.perl6.org/repo/Debugger::UI::CommandLine 13:17
Zoffix
.oO( will git clone modules.perl6.org/repo/Debugger::UI::CommandLine work if I add a correct modules.perl6.org/repo/Debugger::UI.../info/refs redirect as well )
13:17
diakopter I dunno, but I don't know how to build that module. seems I need Terminal::ANSIColor too, which I also don't know how to install [without panda?] 13:18
ShimmerFairy diakopter: without panda, I'd try requiring a fake module (e.g. perl6 -Mxyzzy ) and see where Perl 6 is looking, and just install them in one of those locations 13:20
diakopter sure, but what does "install them" mean
some kind of pre-compile?
ShimmerFairy nah, just copy them over (you could precomp, but it's not required) 13:21
diakopter well, I copied that library and its dependency, but trying to use perl6-debug resulted in this: 13:25
Unhandled exception: ctxlexpad needs an MVMContext at gen/moar/m-CORE.setting:17440 (././CORE.setting.moarvm::24) 13:26
diakopter adding --ll-exception gives Too many positionals passed; expected 1 argument but got 2 at gen/moar/m-Metamodel.nqp:2779 (./blib/Perl6/Metamodel.moarvm:new:0) 13:28
jnthn wonders if it's bitrotten somehow
diakopter jnthn: in case you're curious gist.github.com/diakopter/8471d586ad73e3516f0f 13:31
jdv79 if once were so inclined where would be a good spot to starting into rt.perl.org/Public/Bug/Display.html?id=126761 ? 13:32
diakopter also, I don't get why --ll-exception makes a less-low-level exception appear
jdv79 s/starting/start looking/ 13:33
dalek kudo/nom: 849dfcf | lizmat++ | src/core/Buf.pm:
Allow quoted words as input for Buf.unpack

I think that actually makes more sense than having a string that we need to parse. And that would make a call to Buf.unpack a lot lighter, not having to do a .comb with a Regex.
13:36
dalek ast: fd7d309 | lizmat++ | S32-str/unpack.t:
Add tests for Buf.unpack(<t em plate*>)
13:36
dalek p/js-cps: b980c6d | (Pawel Murias)++ | src/vm/js/ (2 files):
[js] Avoid emitting CPS-less versions of blocks that only work in CPS-mode.
13:42
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 571e579 | (Zoffix Znet)++ | / (2 files):
Fix incorrect repo URLs
13:51
AlexDaniel html form got rendered: examples.perl6.org/categories/cookb...cript.html 14:16
dalek kudo/nom: 235d085 | lizmat++ | src/core/ (2 files):
Move sprintf handler init to Rakudo::Internals
14:27
ast: cf41db9 | lizmat++ | S02-types/WHICH.t:
SprintfHandler is no longer public

Nor should it ever have been, I guess
14:28
timotimo AlexDaniel: that's pretty hilarious, TBH 14:31
stmuk_ I'm intending to fix examples.perl6.org today 14:32
timotimo at first i wondered why perl6.org or *something* doesn't prominently advertise the LineNoise module ... then i remembered the startup time thing and no uninstall function %) 14:35
dalek kudo/nom: 6d9d0f1 | lizmat++ | src/core/Str.pm:
Fix regression caused by removal of .split(:all)
15:06
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 514298e | (Zoffix Znet)++ | t/02-db-builder/05-build-project-list.pl.t:
Add output tests for build-project-list.pl script
15:07
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 45700df | (Zoffix Znet)++ | t/02-db-builder/05-build-project-list.pl.t:
Test on fewer dists to run faster
15:13
nine_ diakopter: yes, panda's curli branch works with rakudo/curli 15:14
diakopter nine_: cool :) 15:18
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: cbc06bc | (Zoffix Znet)++ | lib/ModulesPerl6/DbBuilder/Dist/ (2 files):
Add docs for ModulesPerl6::DbBuilder::Dist::PostProcessor::TravisCI
15:45
[TuxCM] I want 44 random order numers out of the range 10101..99998, so I typed: 15:50
[TuxCM] perl6 -e'(10101..99998).pick(44).sort>>.say' 15:50
and got 44 numers, but they were all the same
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 7ff7bd0 | (Zoffix Znet)++ | t/02-db-builder/02-github-source-and-source-base-class.t:
Improve test coverage
15:51
psch [TuxCM]: seems like it's a mis-optimization on Range.pick
[TuxCM]: you can work around that by reifying the Range with "eager" inside the parens 15:52
m: (eager 10101..99998).pick(44).sort>>.say # as such
camelia rakudo-moar 6d9d0f: OUTPUT«11333␤13214␤13649␤18426␤18534␤19614␤23206␤24780␤25847␤27280␤30241␤37977␤38675␤40013␤40078␤40151␤43761␤46868␤48396␤48507␤48531␤49852␤50702␤51191␤51301␤54160␤54484␤55260␤57537␤59749␤63173␤71499␤…»
[TuxCM] so it is a bug?
psch i hope so, it'd be weird design otherwise :)
timotimo m: say (10101..99998).pick(10) 15:53
camelia rakudo-moar 6d9d0f: OUTPUT«(31845 63838 61319 60171 35864 96036 65385 51705 32185 33410)␤»
timotimo m: say (10101..99998).pick(10).sort
camelia rakudo-moar 6d9d0f: OUTPUT«(59599 59599 59599 59599 59599 59599 59599 59599 59599 59599)␤»
timotimo um ...
sort is the culprit here. wtf?
psch huh
that's even weirder..?
m: (2,3,1).sort.say
camelia rakudo-moar 6d9d0f: OUTPUT«(1 2 3)␤»
lucasb locally I get 44 times the number 35252 :)
timotimo m: my @a = (10101..99998).pick(10); say @a.perl; say @a.sort.perl
camelia rakudo-moar 6d9d0f: OUTPUT«[18593, 32665, 33280, 92199, 96658, 28720, 32128, 53865, 30880, 59247]␤(18593, 28720, 30880, 32128, 32665, 33280, 53865, 59247, 92199, 96658)␤»
psch lucasb: the number changes with each run i bet ;)
timotimo perhaps .sort on a seq is problematic? 15:54
[TuxCM] I leave this up to you lot
lucasb psch: oh, yeah, of course, only tried once :)
psch m: Seq.new([1,2,3]).sort
camelia rakudo-moar 6d9d0f: OUTPUT«Type check failed in binding $iter; expected Iterator but got Array␤ in block <unit> at /tmp/4H9NMeD6Eq:1␤␤»
psch uhm
timotimo m: Seq.new([1,2,3].iterator).sort 15:55
camelia ( no output )
timotimo m: Seq.new([1,2,3].iterator).sort.say
camelia rakudo-moar 6d9d0f: OUTPUT«(1 2 3)␤»
lucasb about the examples site not escaping html when rendering without syntax highligthing, I think the line is this one: github.com/perl6/perl6-examples/bl...y.pm6#L205 15:57
"<pre>" ~ $node.contents.join ~ "</pre>" 15:58
I think Pod::To::HTML would have done the job, but since a code callback was added, it uses that
psch m: (1..10).pick(5).WHAT.say 15:59
camelia rakudo-moar 6d9d0f: OUTPUT«(Seq)␤»
psch m: (1..10).pick(5).sort.say
camelia rakudo-moar 6d9d0f: OUTPUT«(1 3 4 9 10)␤»
psch m: (10000..99999).pick(5).sort.say 16:00
camelia rakudo-moar 6d9d0f: OUTPUT«(39558 39558 39558 39558 39558)␤»
psch m: (2..99999).pick(5).sort.say
camelia rakudo-moar 6d9d0f: OUTPUT«(6575 6575 6575 6575 6575)␤»
psch m: (1..99999).pick(5).sort.say
camelia rakudo-moar 6d9d0f: OUTPUT«(59693 59693 59693 59693 59693)␤»
psch m: (1..1000).pick(5).sort.say
camelia rakudo-moar 6d9d0f: OUTPUT«(385 385 385 385 385)␤»
psch m: (1..256).pick(5).sort.say
camelia rakudo-moar 6d9d0f: OUTPUT«(230 230 230 230 230)␤»
psch m: (1..255).pick(5).sort.say
camelia rakudo-moar 6d9d0f: OUTPUT«(231 231 231 231 231)␤»
RabidGravy :-\
psch m: (0..14).pick(5).sort.say; (0..15).pick(5).sort.say 16:01
camelia rakudo-moar 6d9d0f: OUTPUT«(0 5 6 13 14)␤(11 11 11 11 11)␤»
psch found another magic number \o/
Skarsnik lol
psch m: ^2 .pick(2).sort.say 16:04
camelia rakudo-moar 6d9d0f: OUTPUT«Cannot find method 'say': no method cache and no .^find_method␤ in block <unit> at /tmp/87uWfoPQXb:1␤␤»
psch ...?
lizmat I think it's an artefact from .sort
lizmat is looking into it 16:06
jnthn Note that .sort>>.say is a bit useless, given >> does not promise ordering of side-effects :) 16:08
Zoffix Seems to be bugged on multiples of 3
m: (0..6).pick(2).sort.say; (0..12).pick(4).sort.say; (0..18).pick(6).sort.say;
camelia rakudo-moar 6d9d0f: OUTPUT«(2 2)␤(2 2 2 2)␤(11 11 11 11 11 11)␤»
psch m: (0..15).pick(5).sort.say 16:09
camelia rakudo-moar 6d9d0f: OUTPUT«(13 13 13 13 13)␤»
psch 0..6 is 7 elements
geekosaur not just multiples of 3; (1..256) and (1..255) both did it
Zoffix m: (0..15).pick(5).sort.say; my @z = (0..15).pick(5); @z.sort.say; 16:11
camelia rakudo-moar 6d9d0f: OUTPUT«(0 0 0 0 0)␤(0 3 8 14 15)␤»
geekosaur o.O 16:12
Zoffix m: (0..15).pick(5).sort.say; my @z = (0..15); @z.pick(5).sort.say; 16:13
camelia rakudo-moar 6d9d0f: OUTPUT«(4 4 4 4 4)␤(6 7 9 11 14)␤»
Zoffix m: (eager 0..15).pick(5).sort.say;
camelia rakudo-moar 6d9d0f: OUTPUT«(0 6 8 10 14)␤»
Zoffix oh that was already mentioned 16:14
zengargoyle good * #perl6 16:18
Zoffix \o 16:19
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: d482d0d | (Zoffix Znet)++ | Build.PL:
Remove unused prereq
lucasb would it make sense to make .pick warn when you try to get more than the number of elements? (with the exception of whatever)
psch m: 5.pick(10)
camelia ( no output )
lucasb m: say (1..5).pick(42)
camelia rakudo-moar 6d9d0f: OUTPUT«(1 3 5 2 4)␤»
psch m: 5.pick(10).say
camelia rakudo-moar 6d9d0f: OUTPUT«(5)␤»
lucasb I mean, something like the warn/error: "You can't pick that many distinct elements", something like that 16:20
zengargoyle dunno, seems you'd sometimes want to .pick a maximum number of somethings no matter how many things you have to pick from.
Zoffix lucasb, I wouldn't want that. I rather it pick the maximum number. It's not all the time that I'll have a constant as an argument and I wouldn't want to manually check it's not above the range just to avoid a warning
zengargoyle @contestants.pick($winners)
[TuxCM] I was just using perl6 in Sinterklaas-preparation :) 16:27
timotimo note there's also .pick(*) 16:29
MadcapJake jnthn: As far as how bright/conspicuous twigils are, that's down to the theme but if you're planning on sticking with the One Dark theme, you could add this to your stylesheet (Edit -> Open Your Stylesheet): atom-text-editor::shadow { .support.class.twigil { color: #be5046; } }
MadcapJake that will work for any theme but the color chosen is a dark red that is already used for other elements of the One Dark theme. 16:30
Zoffix Seen that editor that looks like shooting stuff when you're typing? twitter.com/JoelBesada/status/6703...5655293952 16:33
Pretty need.
MadcapJake yeah he wrote an atom plugin that does it too :)
Zoffix \o/ 16:33
MadcapJake atom.io/packages/activate-power-mode
Zoffix doesn't have Atom on this box :'( 16:34
MadcapJake 32bit right?
Zoffix yea
timotimo as soon as "THIS VERSION IS VERY BUGGY RIGHT NOW" is no longer true, i'll totally switch to atom for everything
Skarsnik lol I need this xD
This look fun
MadcapJake timotimo: what version? it's already at 1.2 and 1.3 is almost out of beta 16:35
Zoffix My Perl 6 looks like this right now: i.imgur.com/NteCQnH.png :'( 16:35
timotimo it says exactly that on that site you've linked to
Zoffix timotimo, that's the notice for that weird plugin, not Atom :)
timotimo yeah 16:36
Zoffix k
MadcapJake oh lol :D
timotimo but "that weird plugin" is the thing that'd convince me to switch to atom
Zoffix heh
MadcapJake Zoffix: that is a sad monochrome screen :( 16:37
Zoffix MadcapJake, mst codes like that all the time :P
Skarsnik I use kate, the perl6 coloration is meh, but it's not that bad 16:38
[TuxCM] jnthn, do you need a shorter test case for that core dump? 16:39
MadcapJake check this theme out: atom.io/themes/hacking-the-kernel
Skarsnik it look neat x) 16:40
dalek href="https://modules.perl6.org/new-db-builder:">modules.perl6.org/new-db-builder: 3b0cc57 | (Zoffix Znet)++ | / (11 files):
Prettify code
16:42
Zoffix That's it. Merging
MadcapJake looks pretty good with Perl 6: imgur.com/SZZOjU4 16:43
Zoffix look like a Fallout4 Pip Boy
MadcapJake haha yeah! 16:44
dalek Heuristic branch merge: pushed 114 commits to modules.perl6.org by zoffixznet
iH2O looks like you folks have been investing a lot of work in this new potentially-cult language 16:45
Zoffix iH2O, yup 16:46
iH2O lol
jnthn [TuxCM]: The utf8-c8 one?
[TuxCM]: In which case no, I just need time and to not have a headache, and I'll be able to figure it out :) 16:47
Zoffix m: say ༤ / ༨ + ༪ * ༤;
camelia rakudo-moar 6d9d0f: OUTPUT«2.5␤»
[TuxCM] ok
Zoffix iH2O, ^ and now we can do Tibetan math! The goal is complete: P
[TuxCM] and yes, that was the one I meant
raiph m: my Stringy $foo = "27"; say 1 + $foo # I was hoping that use of `Stringy` rather than `Str` would avoid Cool. Is it realistic to anticipate the ability to write a module, or use a later Perl 6 version, that allows one to avoid Cool / implicit coercion and instead force a compile-time error in this scenario? 16:57
camelia rakudo-moar 6d9d0f: OUTPUT«28␤»
dalek href="https://modules.perl6.org:">modules.perl6.org: c7c5e5d | (Zoffix Znet)++ | bin/build-project-list.pl:
Fix default app path
16:59
psch m: say Stringy ~~ Cool 17:04
camelia rakudo-moar 6d9d0f: OUTPUT«True␤»
dalek rl6-roast-data: 4e66910 | coke++ | / (9 files):
today (automated commit)
psch raiph: it depends on how general you want that 17:13
raiph: what already works (but might be too specific) is just but-ing a role with a Numeric that dies
raiph m: say Numeric ~~ Cool
camelia rakudo-moar 6d9d0f: OUTPUT«True␤»
psch m: my $x = "26" but role { method Numeric { die "nope" } }; say 1 + $x # as here for example 17:14
camelia rakudo-moar 6d9d0f: OUTPUT«nope␤ in method Numeric at /tmp/4bVQobqNRy:1␤ in block <unit> at /tmp/4bVQobqNRy:1␤␤»
psch +"onto the Str"
RabidGravy psch++ # I was just about to paste exactly the same thing 17:16
psch i guess another approach could be exporting a ClassHOW that skips Cool in MRO based type checks or something like 17:21
+that
m: say &infix:<+>.candidates>>.signature 17:22
camelia rakudo-moar 6d9d0f: OUTPUT«No appropriate parametric role variant available for 'Rational'␤ in any specialize at gen/moar/m-Metamodel.nqp:2511␤ in any compose at gen/moar/m-Metamodel.nqp:2821␤ in any make_pun at gen/moar/m-Metamodel.nqp:1658␤ in any find_method at gen/mo…»
psch m: say &infix:<+>.candidates
camelia rakudo-moar 6d9d0f: OUTPUT«No appropriate parametric role variant available for 'Rational'␤ in any specialize at gen/moar/m-Metamodel.nqp:2511␤ in any compose at gen/moar/m-Metamodel.nqp:2821␤ in any make_pun at gen/moar/m-Metamodel.nqp:1658␤ in any find_method at gen/mo…»
[Coke] (macros) I think if we mark it as experimental at least in the docs we'll be ok. if we want, we could make it need "use MONKEY_MACROS" or some such. 17:24
raiph psch: Thanks! 17:27
dalek href="https://modules.perl6.org:">modules.perl6.org: 208cbe3 | (Zoffix Znet)++ | bin/build-project-list.pl:
Exit with non-success so we could log properly
17:30
Zoffix .tell moritz I've updated modules.perl6.org builder so now it no longer uses /var/www/modules.perl6.org dir. I've updated the cron job and all and cleaned the /var/www/modules.perl6.org dir but left the dir itself in place, as I see it's specified in Apache's config as document root and it'll whine if it's missing. The build script now pauses for 5 seconds after each dist, so we don't hit GitHub's rate limit when we get more dists in the ecosystem. 17:36
yoleaux Zoffix: I'll pass your message to moritz.
Zoffix .tell moritz The cron still runs it on 20,50 minutes. That's fine. If the script is still running, attempt to run a second copy will just exit. Since logging that from cron line like before would overwrite the log, I moved it into the update bash script to log into a temp file and only a copy the file to update.log when build script finishes a build.
yoleaux Zoffix: I'll pass your message to moritz.
Zoffix
.oO( this channel needs a Markov Chain bot... )
17:38
moritz Zoffix: ok, thanks
yoleaux 17:36Z <Zoffix> moritz: I've updated modules.perl6.org builder so now it no longer uses /var/www/modules.perl6.org dir. I've updated the cron job and all and cleaned the /var/www/modules.perl6.org dir but left the dir itself in place, as I see it's specified in Apache's config as document root and it'll whine if it's missing. The build script now pauses for 5 seconds after each dist, so we don't hit GitHub's rate limit when we get more dists in
the ecosystem.
17:36Z <Zoffix> moritz: The cron still runs it on 20,50 minutes. That's fine. If the script is still running, attempt to run a second copy will just exit. Since logging that from cron line like before would overwrite the log, I moved it into the update bash script to log into a temp file and only a copy the file to update.log when build script finishes a build.
dalek href="https://modules.perl6.org:">modules.perl6.org: 68274e7 | (Zoffix Znet)++ | TODO-IDEAS.md:
Remove "Cleanup db build script" from TODO; that"s already done
17:58
lucs In some roast test files, I see use lib 't/spec/packages'; What/where is that? 18:45
dalek osystem: 21f8d11 | RabidGravy++ | META.list:
Add WebService::Soundcloud

  github.com/jonathanstowe/WebService-Soundcloud
18:46
Zoffix w00t 18:48
lucasb lucs: iiuc, it expects to be run from the rakudo root dir, so t/spec and the roast dir are the same
Zoffix RabidGravy++
lucasb lucs: so t/spec/packages is the directory roast/packages in the roast repo 18:49
RabidGravy won't do downloads properly until I have made a big change to HTTP::UserAgent
lucs lucasb: How does that work? Are you saying that perl knows it should replace "t/spec" with "roast"? 18:51
RabidGravy it just checks the files out there 18:53
lucasb lucs: when you say 'make spectest' in the rakudo root dir, it git clones roast into t/spec 18:54
dalek href="https://modules.perl6.org:">modules.perl6.org: 7124208 | (Zoffix Znet)++ | bin/build-project-list.pl:
Use shorter dist build interval; 3secs is still safe
18:58
lucs lucasb: Hmm... I'd like to run (after modification) individual tests using a freshly modified/recompiled (via .../rakudobrew build moar) perl6, something like this: perl6 .../roast/S42-foo/bar.t 18:59
lucs I'm guessing 'make spectest' runs the whole test suite, which I'd like to avoid for now. 19:01
brrt good * #perl6
psch lucs: there are make target for every spectest file iff you have roast checked out to t/spec 19:02
lucs: in general for CORE hacking i strongly recommend checking out rakudo manually
lucs psch: Um, any instructions online somewhere to do that? 19:03
psch git clone github.com/rakudo/rakudo rakudo && cd rakudo && perl Configure.pl --backends=moar --gen-nqp --gen-moar --make-install
lucs Oh cool, thanks. 19:04
I'll let you know how it goes...
[Coke] rakudo-jvm build failures are known? 19:11
dalek ast/curli: 9a3c980 | lizmat++ | S02-types/WHICH.t:
Add test for new X::Pragma::CannotPrecomp class
19:12
roast/curli: 7dc6365 | lizmat++ | S02-types/WHICH.t:
roast/curli: Add X::Package::UseLib
dalek Heuristic branch merge: pushed 111 commits to rakudo/curli by niner 19:12
mst I am really unable to fathom who thought X for exceptions was a remotely good idea 19:13
I foresee another 10 years of "no, you can't call your examples X, Y and Z" to go with the "no, you can't call your examples A, B and C" we do over in #perl
mst I suppose I should've noticed and complained about this several years ago but still: ew :) 19:14
nine_ I guess next time we should give ourselves a decade or two more to solicit enough feedback ;) 19:15
user3 rakudo 2015.11 won't' compile on my system pastebin.com/6pCq2HgP
stmuk_ nine_++
dalek ast: 9a3c980 | lizmat++ | S02-types/WHICH.t:
Add test for new X::Pragma::CannotPrecomp class
19:16
roast: 7dc6365 | lizmat++ | S02-types/WHICH.t:
roast: Add X::Package::UseLib
brrt X is short though 19:16
stmuk_ is there a bool version of ~$x +$x etc? 19:17
stmuk_ ? 19:17
brrt i auppose the idea was that exceptions are common and should bw huffma-coded
i think ?$x 19:18
Ben_Goldberg stmuk_, so $x
stmuk_ oh yeah thanks I forgot that
FROGGS_ user3: a workaround might be to install libatomic-ops-dev
brrt m: my $x = 12; say ?$x; 19:19
camelia rakudo-moar 6d9d0f: OUTPUT«True␤»
user3 FROGGS_: noted
stmuk_ stmuk-- # it's fairly obvious from the cheatsheet too
dalek p: 9559a26 | (Stefan Seifert)++ | tools/build/MOAR_REVISION:
Bump MOAR_REVISION for deadlock fixes
brrt there are no dumb questions stmuk 19:20
dalek kudo/nom: 82f1062 | (Stefan Seifert)++ | tools/build/NQP_REVISION:
Bump NQP revision to get MoarVM deadlock fixes
nine_ curli is merged!
FROGGS_ O.o
\o/
nine++
diakopter nine_++
brrt \o/ nine++
stmuk_ precomp in nom?
FROGGS_ nine_: now I need to switch branches :o) 19:21
psch [Coke]: yeah, i think nqp-j doesn't handle the cclass for 𝑒 correctly
[Coke]: which causes the misparse and breaks the build
nine_ Please note that I do expect a couple of rough days ahead. I'd have liked to work a little more on a smooth transition but we're just too close to the release
FROGGS_ true 19:22
MadcapJake n00b question: what did the curli branch do?
FROGGS_ MadcapJake: you install into a database (repository) now 19:23
nine_ Merged curli in panda, too.
MadcapJake as in modules are installed into a db?
FROGGS_ MadcapJake: I mean, the modules get installed into a repository
aye
hoelzro timotimo: working on the docker image right now 19:24
nine_ MadcapJake: you get all the benefits of precompilation pretty much for free and may install as many modules as you like without having it affect rakudo's startup speed.
brrt it advancifies module installation for robustness and features
hoelzro as far as security goes, it's one of the official docker images, so I think that's in their hands. I would hope that they would update their base image and rebuild our image on top of that in face of critical security issues
MadcapJake oh very nice! thanks for the explanations everyone! 19:24
moritz the main benefit speed-wise is precompilation that curli enables again 19:25
brrt the converse, installing modules having am effect on startup time was bonkers
moritz nine_: rakudo star manually precompiles its modules right now
moritz nine_: I guess that's not recommended anymore after curli, right? 19:26
timotimo hoelzro: awesome 19:26
nine_ moritz: the code that precompiles won't work anymore anyway
nine_ moritz: the interfaces changed drastically. 19:26
moritz nine_: so what should star do instead? 19:26
FROGGS_ nine_: it just calls `perl6 --target=mbc --output=..`
moritz: what if it used panda? 19:27
nine_ moritz: install with something like $*REPO.repo-chain.first(CompUnit::Repository::Installable).install(...)
jnthn nine_++ \o/
moritz FROGGS_: that would be awesome 19:29
zengargoyle nine_++ woo hoo 19:30
hoelzro nine_++ 19:31
jnthn mst: Um, but we already fixed declaring a class X { } breaking reporting of compile-time errors, and it doesn't break the built-ins reporting errors either, and presumably in something small enough to use X as an example class name you're probably not going to be declaring/using typed exceptions yourself. :)
nine_ Seems like t/spec/S10-packages/precompilation.t takes some attempts to pass (but then does consistently). I'll have a closer look tomorrow. Need a bit of relaxation now :)
moritz nine_++ 19:32
FROGGS_ nine_: rest well :o)
moritz FROGGS_: will you look into bootstrapping star with panda?
jnthn mst: So it's not that likely to be much trouble any more. :)
nine_: Enjoy some well-earned rest!
FROGGS_ moritz: can do that, yes
MadcapJake will rakudobrew upgrade correctly? 19:33
mst jnthn: mmmm ... colour me "I don't have an argument nearly sufficient to justify changing your mind, but I'm still suspicious" at this point then 19:35
lucs psch: Worked like a charm, as did make spectest . 19:42
dalek : c2b3110 | moritz++ | misc/perl6advent-2015/schedule:
p6advent: Claim day 1, since nobody else seems to do it :(

also add author information to day 2
19:46
flussence «Stage parse : 96.158» -- it grew another 20% :(
«Could not find Panda:ver<True>:auth<True>:api<True> in:» -- I guess I need to update that too... 19:48
colomon woof, not a lot of days claimed yet.
moritz aye :( 19:49
moritz .uni cominbing ced 19:55
brrt moritz: i may be able to write some day, if there is interest in my stories
moritz .u cedil
yoleaux U+00B8 CEDILLA [Sk] (¸)
U+00C7 LATIN CAPITAL LETTER C WITH CEDILLA [Lu] (Ç)
U+00E7 LATIN SMALL LETTER C WITH CEDILLA [Ll] (ç)
moritz brrt: oh yes, please
[Coke] m: say 32:wq;
camelia rakudo-moar 82f106: OUTPUT«5===SORRY!5=== Error while compiling /tmp/rIUURywoHQ␤You can't adverb that␤at /tmp/rIUURywoHQ:1␤------> 3say 32:wq7⏏5;␤ expecting any of:␤ pair value␤»
MadcapJake i'll try and come up with something, anyone have a good beginner/intermediate advent idea that I could do?
moritz MadcapJake: since it's your area of work these past days/weeks, why not write about Perl 6 support in Atom? 19:56
MadcapJake ok! I didn't know that was a good area!
moritz MadcapJake: everything that's related to Perl 6, and hasn't been beaten to death, is a good topic :-)
[Coke] m: my $a = 32; say $a:wq; 19:57
camelia rakudo-moar 82f106: OUTPUT«5===SORRY!5=== Error while compiling /tmp/10vdAypzrW␤Variable '$a:wq' is not declared␤at /tmp/10vdAypzrW:1␤------> 3my $a = 32; say 7⏏5$a:wq;␤»
MadcapJake hehe, i'll look over the brainstorming list and see if I can find another one I could do.
[Coke] I swear I just had a stray :wq in my code that compiled.
RabidGravy is there a reason that modules.perl.org says "222 known Perl 6 modules" when the META.list has 462 20:00
moritz RabidGravy: likely a bug. Zoffix, something for you to look at :-) 20:01
brrt moritz: ok. will do some day :-) 20:03
moritz m: my int @a[4][5] 20:04
camelia rakudo-moar 82f106: OUTPUT«5===SORRY!5=== Error while compiling /tmp/3sRP5DnoKM␤Multiple shapes not yet understood␤at /tmp/3sRP5DnoKM:1␤------> 3my int @a[4][5]7⏏5<EOL>␤ expecting any of:␤ constraint␤»
moritz m: my int @a[4;5]
camelia ( no output )
moritz m: my int32 @a[4;5]
camelia ( no output )
dalek kudo/nom: dbfb5e2 | coke++ | src/core/Num.pm:
avoid jvm build issue

  psch++ for pointing out the issue.
20:07
[Coke] psch++ ^^ there's a workaround.
BenGoldberg .u e 20:08
yoleaux U+0065 LATIN SMALL LETTER E [Ll] (e)
BenGoldberg .u 𝑒 20:09
yoleaux U+1D452 MATHEMATICAL ITALIC SMALL E [Ll] (𝑒)
BenGoldberg Just a WAG, but I suppose jvm doesn't yet grok unicode above u+ffff. 20:11
stmuk_ can I request anyone likely to attend my LPW talk next month do the following? 20:14
stmuk_ disable javascript to reveal the SUBMIT on act.yapc.eu/lpw2015/talk/6470 and add to personal schedule? 20:14
dalek : c515940 | (Jake Russo)++ | misc/perl6advent-2015/schedule:
Update schedule
20:16
: ab17807 | RabidGravy++ | misc/perl6advent-2015/schedule:
Merge pull request #11 from MadcapJake/master

Update schedule
stmuk_ and also upvote the other perl6 talks :) 20:19
moritz MadcapJake: please /msg me your email address, so that I can send you an invitation for the advent blog 20:26
dalek pan style="color: #395be5">perl6-examples: c1f959c | (Steve Mynott)++ | lib/Pod/Htmlify.pm6:
fix conditional breakage due to require return value change - this should reenable colour syntax and fix current website issues
20:30
timotimo o/ 20:56
hankache hi timotimo
timotimo hey hankache
moritz: can you suggest an advent topic for me, too? 20:57
Mouq Is there already an understood way to get enum values from C code using NativeCall? 20:58
hankache the link to Advent calendars in perl6.org/documentation/ is dead 20:59
RabidGravy I think you'd have to make a small C wrapper that did it for you
hankache what is it btw? 21:00
Mouq nine_++
FROGGS_ Mouq: I just declare a Perl 6 enum using the documented names/values
hankache: perl6advent.wordpress.com/ 21:01
FROGGS_ hankache: the links do not seem dead 21:02
Zoffix RabidGravy, where did it say it? On the top of the home page?
hankache FROGGS_ I tried opening it a minute ago it didn't. now it works 21:03
RabidGravy yeah, it says 459 now. weird
Mouq FROGGS_: That works for some enums, but others, the library insists I have the wrong value… I haven't tried since last night, so maybe there's something obvious I didn't notice that I'm doing wrong.
FROGGS_ Mouq: how does it tell you that the value is wrong? 21:05
Zoffix RabidGravy, I know the approximate location of the error, so I'll take a look at it 21:06
Zoffix The log says "Removed 1 dists that are no longer in the ecosystem", but we didn't remove anything from the ecosystem, so there's glitching with the error catcher when a particular dist fails to rebuild 21:07
RabidGravy software's shit like that ;-)
Zoffix I'm kinda disappointed, because I have pretty good test coverage lol :P 21:08
Mouq FROGGS_: " ** On entry to STRMM parameter number 1 had an illegal value". (I've tried every value from 0 to 20000, and the enum values are 141 and 142) 21:10
FROGGS_ Mouq: I think the problem is not about the enum, but either about how you call functions or layout structures 21:16
lucasb the irc logger bot didn't get back :( 21:19
Mouq moritz: github.com/moritz/json/pull/20 21:23
stmuk_ shouldn't panda really be in travis? 21:23
Zoffix k, identified the problem modules.perl6.org 21:26
stmuk_ I'm seeing Could not find Panda:ver<True>:auth<True>:api<True> in panda as well 21:26
lucasb m: @*INC # would a short lived deprecation for this be helpful in these following days? 21:27
camelia rakudo-moar dbfb5e: OUTPUT«Dynamic variable @*INC not found␤ in block <unit> at /tmp/qlyfOvuLaF:1␤␤Actually thrown at:␤ in block <unit> at /tmp/qlyfOvuLaF:1␤␤»
Zoffix There's protection if fetching commits/dist info fails, but if fetching META file itself fails (e.g. due to network error), the dist gets chucked. 21:27
Gonna fix it after I grab some beer and food.
Mouq lucasb: At the very least, a more helpful message would probably be good 21:27
Zoffix lucasb, yes, there are modules on ecosystem that rely on @*INC entirely for their functionality. modules.perl6.org/repo/Pluggable is one example 21:32
RabidGravy m: @*INC || warn "bo inc" 21:32
camelia rakudo-moar dbfb5e: OUTPUT«bo inc in block <unit> at /tmp/UFpY2FQHgw:1␤»
flussence how do I... get «prove -e perl6» to run now? Even the simplest no-deps modules I have don't work any more. 21:36
stmuk_ is there currently anyway of building panda under rakudobrew? 21:38
moritz does PERL6LIB=lib work?
stmuk_ or should I just wait 21:39
flussence trying JSON::Tiny gives me @*INC errors, most things give me a stacktrace with True xx 3, and in workflows that *used* to work perfectly fine I get «Type check failed in binding; expected Str but got IO::Path» 21:39
moritz: that works, PERL6LIB=. with a correct META6.json no longer does
stmuk_ rakudobrew build-panda [panda-ver] 21:39
ah
moritz flussence: ah well, that's probably part of the rough days after the merge that nine_++ warned us about 21:40
flussence as long as there's *one* thing that works, I can live with it... 21:41
moritz my p6 advent post for Dec 01 is up, and can be reviewed by anybody who has a p6advent login
nine_ flussence: replace BEGIN { @*INC.push('lib') }; with a simple use lib 'lib'; 21:42
flussence: or even remove it altogether and just run prove -e 'perl6 -Ilib' t
jnthn Seems there's already a PR for JSON::Tiny: github.com/moritz/json/pull/20 21:42
nine_ I guess that's just really old code that got copy&pasted a lot and is easily replaced 21:43
jnthn oh, seems I even can merge it :)
flussence nine_: is that for JSON::Tiny? not my module, someone else'll have to fix that :)
dalek on: ef41b31 | Mouq++ | t/0 (5 files):
@*INC is dead, long live CURLI
on: c75a04c | jnthn++ | t/0 (5 files):
Merge pull request #20 from Mouq/master

Fix tests on latest Rakudo
FROGGS_ moritz: I made some progress with star using panda... however, I wont be able to push something sensible today
moritz: will continue tomorrow
moritz FROGGS_: ok, great
nine_ FROGGS_, moritz: thanks for thanking this on :) 21:44
jnthn Sleep time...and tomorrow is a travel day
So won't be about a lot
'night o/
FROGGS_ gnight
nine_ Good nichgt!
moritz does the list form of lib work? i.e. "use lib 'blib', 'lib';" 21:46
skids not use lib <blib/lib lib>? 21:50
autarch folks here might enjoy this - blog.urth.org/2015/11/29/perl-6-is-fun/ 21:59
autarch moritz: I know for sure that the list form works - I'm using it in a number of places 22:00
unless you mean as of a very recent patch? 22:01
flussence recent would be in the past few hours...
Mouq autarch++
autarch well, I see a patch related to @*INC just above so who knows! 22:02
lizmat CORE:ver<Bend-281-gdbfb5e2>:auth<perl>:api<> already installed 22:02
in block <unit> at tools/build/install-core-dist.pl:14
nine: does that ring a bell ? ^^^
flussence all of curli just got merged, but the commit bot died without reporting it
lizmat I can't build enymore 22:03
zapping install dir and trying again
Mouq autarch: Btw, I got your fork of P2H working on doc. It's a fair bit slower, so I was trying to see what I could do to cache the renderer objects
autarch Mouq: I'm working on threading everything - github.com/autarch/doc/tree/autarc...org-design - very much in progress work to essentially refactorwite the whole doc generation system 22:04
dalek pan style="color: #395be5">perl6-examples: 7477382 | (Steve Mynott)++ | util/update-and-sync:
attempt to fix examples.perl6.org by using older versions of rakudo/panda until the dust settles from curli changes
autarch but if you look in lib/DocSite/Generator.pm you can see that the generation it's doing so far is all threaded 22:05
lizmat nine++ # for all the curli work
lizmat nine: as a datapoint: zapping the install dir solved it!\ 22:05
moritz autarch++ # blog post
moritz autarch: might I lure you into writing a post for the perl 6 advent calendar? 22:06
[Tux]: would you like to write about your Text::CSV experience for the Perl 6 advent calendar?
autarch Mouq: btw, that branch is pretty poorly named at this point - I started just trying to change the header & footer of the generated docs but it's gradually become a _much_ bigger effort - first I had to rewrite Pod::To::HTML
moritz (or other topics... )
autarch moritz: you're making me cry - I just finished my first draft for the Perl 5 advent calendar and I thought I was done ;)
but I could probably be talked into it as long as it can be a little later in the month 22:07
autarch Mouq: then I started looking at htmlify.p6 and I wanted to make it OO, and make it thread, and make it less of a god process/object, and make it spit out nice output as it goes, and ... 22:08
RabidGravy I'm going for approx 11th if that's cool
Mouq autarch: Oh, ok, cool! I didn't realize you had already started working on that :) 22:08
autarch Mouq: it'll probably take a little while to be done, but once it's done I think it'll be in much better shape for doing other things I really want to do, like reorganize the language section per github.com/perl6/doc/issues/114 22:09
masak general question: if I `use B;` in a module and my B.pm happens to also `use A;`, I get all A.pm's global names "for free", and I can choose freely whether to also `use A;` in my module. 22:11
autarch TimToady: btw, any chance of adding something like DateTime.posix-hi-res to get an epoch with fractional seconds? writing "my $secs = $now.posix + ($now.second - $now.whole-second)" is a bit grotesque!
dalek kudo/nom: af4d62e | lizmat++ | / (3 files):
Remove "use cur"

Now that the curli branch has been merged, this now *really* has become a fossil
masak what do people think about that in general? is it good practice to still `use A;`, even though leaving it out is harmless? 22:12
autarch moritz: should I move my branches for various things over to the perl6 org now? it might make them more findable, but I don't want to step on any toes. 22:12
timotimo moritz: i think it was actually the november release that first shipped with NSA, no? 22:12
masak (so far, my thinking has been that it's beneficial, because the module becomes slightly more resilient to changing module loading order.)
dalek pan style="color: #395be5">perl6-examples: 122e637 | (Steve Mynott)++ | Makefile:
explicit panda path
22:12
masak (also, that it's good documentation, if the module actually uses stuff from A.pm)
masak 'night, #perl6 22:13
timotimo moritz: also in the last sentence "the excitement" instead of "the exciting"
Mouq masak: That, and maybe readability would be better too :) 22:13
night!
dalek : 814ec24 | RabidGravy++ | misc/perl6advent-2015/schedule:
Update schedule

Add my pennies worth
22:14
lucasb is ilbot3 the irc logger, isn't it? why did it stopped recording again? it seems it's still online in the channel 22:17
lucasb oh, that explains... 22:19
RabidGravy probably because of that ^ 22:19
autarch hmm, any idea how I can fix this travis failure - travis-ci.org/autarch/perl6-Pod-No...s/93830482 ? 22:21
did rakudobrew change in some way that I need to account for in my .travis.yml? I just cargo-culted that travis file from some other repo, I think. 22:22
MadcapJake masak: if `B.pm` is a top-level module sort-of-thing then don't add it because the inner modules are all expected to be available but if A is a separate entity (external or a seperate domain), then I would say add it to show that you are specifically looking to use A's functionality.
timotimo does $supply.schedule-on actually cause things on the supply to be scheduled? i thought it returns a new supply that will do scheduling for you?
diakopter autarch: I think rakudobrew needs some updating for latest panda rewrite (merged hours ago) 22:23
autarch ah, ok
so I guess I just wait
timotimo ah, i'm wrong 22:23
MadcapJake `rakudobrew build moar` says it successfully installed panda but then I got a bunch of errors when it tried ot install some modules 22:26
timotimo MadcapJake: i nuked my install/ folder before i updated to latest master commits 22:28
timotimo and with that i was able to bootstrap a panda successfully 22:30
neilb hi, i've been running the pull request challenge this year, for p5 CPAN modules. Participants are assigned a random module/dist and have a month to do a pull request on it. If they do at least one PR, then get an assignment the following month. 22:31
Zoffix waves to neilb 22:31
timotimo ugh
i have to update my $PATH every single time i build a new rakudo?
neilb I've done a survey of participants, to get data for a talk I'm giving, and one of them was "would you be interested in a similar challenge for Perl 6"
(Hi Zoffix) 22:32
Mouq timotimo: I saw that too… I think some linking is in order 22:32
neilb so far nearly 65% of respondents have said they'd be interested in a simular Perl 6 challenge
timotimo alternatively run perl6 --version on start of the shell to get the thing >_>
chansen_ autarch: $secs = $now.posix + ($now.whole-second % 1); # slightly less grotesque 22:33
Zoffix neilb, 65% is how many humans?
neilb 118 22:34
Zoffix Not bad
m: 461/118
camelia ( no output )
Zoffix m: say 461/118
camelia rakudo-moar af4d62: OUTPUT«3.906780␤»
Zoffix 4 dists per human :D
neilb So if someone wants to organise a similar challenge for Perl 6, there's some appetite
Zoffix neilb, thanks for letting us know. 22:35
Zoffix Maybe would be a nice thing to start right after a New Years. 22:35
neilb happy to share experiences and lessons learned, on what worked and what didn't.
I announced it around this time last year -- most people signed up on new year's eve, to get an assignment on 1st Jan :-) 22:36
RabidGravy neilb, as an author who received two PRs as a result of that, they were somewhat half hearted :-\
Zoffix sent several half-hearted PRs 22:37
timotimo i sent half a human heart as a PR. it was not accepted.
Zoffix Most because the subject matter of the module I got assigned was beyond my knowledge.
s/Most/Mostly/'
chansen_ autarch: typo, should be .second not .whole-second
neilb RabidGravy: plenty of people joined the challenge expressly to learn more about Perl, CPAN, the CPAN toolchain, github and doing PRs. 22:39
As a result there have been plenty of less-than-perfect PRs, I know.
Mouq
.oO( "All PR is good PR" )
22:41
neilb A number of people have said they'd like a similar challenge for Perl 6 to help them learn Perl 6. So yes, the PRs wouldn't be perfect. But growing a community generally means adding people who are going to be less than perfect :-)
RabidGravy wasn't a problem though
neilb Mouq: :-) 22:41
autarch chansen_: I still think it should be part of the core API 22:43
RabidGravy neilb, given the youth of the ecosystem perhaps a "Most Wanted Challenge" might be more effective 22:43
Zoffix +1 22:44
Mouq It seems CURLI can't handle libraries that reference other files in the library not listed in "provides" in META.info
neilb RabidGravy: aye, maybe so. 22:44
stmuk curli doesn't seem to allow reinstall either 22:44
neilb Anyway, there's definitely appetite for some kind of challenge, so maybe someone here might like to take up that baton? 22:45
chansen_ autarch: I agree 22:45
Zoffix shies away from the baton 22:45
timotimo Mouq: is that surprising? 22:46
Mouq timotimo: Well, it means that Pod::Coverage is broken, which means that Linenoise can't be installed
timotimo :o
timotimo why does Pod::Coverage even rely on something like that? 22:46
RabidGravy haven't finished half the modules I need yet 22:47
Mouq Pod::Coverage may also be broken for other reasons relating to CURLI, but that was the first thing I found 22:48
RabidGravy I'll look in the morning 22:49
lucasb timotimo: maybe it is the other way around, Linenoise requires LibraryMake
...that requires Pod::Coverage
leedo Mouq: i was just looking at the Pod::Coverage trouble myself
leedo github.com/retupmoca/P6-Find-Bundled/pull/2 possibly related as well? 22:50
Zoffix huh 22:51
Mouq leedo: Interesting 22:51
Zoffix thought () were not allowed in ifs
Zoffix leedo, is that even supposed to work at all? github.com/retupmoca/P6-Find-Bundl...b1e6a986L9 22:52
@*INC is gone IIUC
timotimo if( ... ) will try to call a sub called "if"
if ( ... ) is just unnecessary, but not wrong
leedo timotimo: well it does avoid the RHS gobbling the block at least 22:52
Mouq m: say CompUnitRepo::Local::File 22:53
camelia rakudo-moar af4d62: OUTPUT«Could not find symbol '&File'␤ in block <unit> at /tmp/gdo8geIzxd:1␤␤Actually thrown at:␤ in block <unit> at /tmp/gdo8geIzxd:1␤␤»
dalek href="https://modules.perl6.org:">modules.perl6.org: f86a5ba | (Zoffix Znet)++ | / (5 files):
Add meta_url to Dists model

This will let us salvage distro builds regardless of what error happens during build
22:53
stmuk examples.perl6.org should now be in colour and have correct HTML escaping 22:54
Sgeo > "hi" 22:54
Sgeo m: "Foo"[0][0] 22:55
camelia ( no output )
Sgeo m: say "Foo"[0][0]
camelia rakudo-moar af4d62: OUTPUT«Foo␤»
Zoffix stmuk++
AlexDaniel m: say “Foo”[0][1]
camelia rakudo-moar af4d62: OUTPUT«Index out of range. Is: 1, should be in 0..0␤ in block <unit> at /tmp/XXYUUtkNJx:1␤␤Actually thrown at:␤ in block <unit> at /tmp/XXYUUtkNJx:1␤␤»
AlexDaniel m: “Foo”[0][1] = 20 22:56
camelia rakudo-moar af4d62: OUTPUT«Index out of range. Is: 1, should be in 0..0␤ in block <unit> at /tmp/EW3O96f0HB:1␤␤Actually thrown at:␤ in block <unit> at /tmp/EW3O96f0HB:1␤␤»
AlexDaniel m: “Foo”[0] = 20
camelia rakudo-moar af4d62: OUTPUT«Cannot modify an immutable Str␤ in block <unit> at /tmp/pfJ_UhYwDx:1␤␤»
AlexDaniel that's fair
m: “Foo”[0][* - 1] = 20 22:57
camelia rakudo-moar af4d62: OUTPUT«Cannot modify an immutable Str␤ in block <unit> at /tmp/gcZMCGJD9H:1␤␤»
AlexDaniel m: say “Foo”[0][* - 1]
camelia rakudo-moar af4d62: OUTPUT«Foo␤»
dalek pan style="color: #395be5">perl6-examples: 2a85dbb | (Steve Mynott)++ | categories/best-of-rosettacode/accumulator-factory.pl:
remove 2nd pod section since it's not useful and breaks the syntax colouring
23:03
MadcapJake timotimo: thanks, nuking it worked 23:05
dalek kudo/nom: 0ede335 | lizmat++ | src/core/Range.pm:
Fix Range.pick(N), push-all sparse version
23:06
lizmat .tell nine it looks like trying to build with uncommitted changes, gives the "already installed" error if you built on the latest commit before 23:08
yoleaux lizmat: I'll pass your message to nine.
lizmat .tell [Tux] say (10101..99999).pick(44).sort should now work 23:09
yoleaux lizmat: I'll pass your message to [Tux]. 23:09
AlexDaniel m: say “foobar”.substr-eq(“bar”, 3); 23:10
camelia rakudo-moar af4d62: OUTPUT«1␤»
AlexDaniel why is it 1 and not True?
docs say that it returns Bool
MadcapJake m: say “foobar”.substr-eq(“baz”, 3);
camelia rakudo-moar af4d62: OUTPUT«0␤»
MadcapJake m: say “foobarbar”.substr-eq(“bar”, 3); 23:11
camelia rakudo-moar 0ede33: OUTPUT«1␤»
Mouq AlexDaniel: Probably it calls into nqp and is never coerced back to Bool. LHF :)
AlexDaniel Mouq: willing to fix it right now? :)
RabidGravy it's just a ? somewhere surley 23:12
hartenfels m: subset Empty of Str where /^$/; sub s (Empty $e) {}; try s('a'); $!.^name; 23:15
camelia ( no output )
hartenfels subset Empty of Str where /^$/; sub s (Empty $e) {}; try s('a'); say $!.^name;
lizmat AlexDaniel: - $pos >= 0 && nqp::eqat($str, nqp::unbox_s($needle), $pos); 23:16
+ $pos >= 0 && nqp::eqat($str, nqp::unbox_s($needle), $pos) == 1;
hartenfels m: subset Empty of Str where /^$/; sub s (Empty $e) {}; try s('a'); say $!.^name;
camelia rakudo-moar 0ede33: OUTPUT«X::AdHoc␤»
lizmat line 158 in Cool
good night, #perl6
Mouq 'night lizmat! 23:17
bolangi p6: say 3 23:18
camelia rakudo-moar 0ede33: OUTPUT«3␤»
Mouq m: say $*REPO 23:20
camelia rakudo-moar 0ede33: OUTPUT«inst#/home/camelia/.perl6/2015.11-283-g0ede335␤»
Mouq m: say $*REPO.repo-chain # aha 23:22
camelia rakudo-moar 0ede33: OUTPUT«(inst#/home/camelia/.perl6/2015.11-283-g0ede335 inst#/home/camelia/rakudo-m-inst-2/share/perl6/site inst#/home/camelia/rakudo-m-inst-2/share/perl6/vendor inst#/home/camelia/rakudo-m-inst-2/share/perl6)␤»
Herby_ Good evening, everyone! 23:36
Mouq o/ 23:36
Herby_ I'm trying to install the LWP::SImple module, on Ubuntu using Panda 23:46
and it failed
Zoffix Herby_, pastebin errors fpaste.scsys.co.uk/
Herby_ pastebin.com/ZmvYa4C3
Zoffix Herby_, and how old your rakudo is
Herby_ This is perl6 version 2015.11 built on MoarVM version 2015.11 23:47
Zoffix No idea, sorry :) 23:47
Herby_ same conclusion I came to :) 23:48
my ubuntu installation is a mess
Zoffix There've been significant changes very recently, so maybe this is the fallout.
Skarsnik Herby_, try to clone the repository on github and run the test that fail 23:49
Herby_ progress is painful!
Herby_ ehhhhhhh, only thing I understood in that is "github" :) 23:49
Skarsnik you can try a panda --notest install 23:50
I think it's the right switch
Herby_ looks like its --notests 23:51
and it installed successfully. lets see if it works
Herby_ no such luck 23:53
Skarsnik hm
Herby_ Could not find symbol '&Simple' in block <unit> at test.pl6:3
i'm trying to run the example found under "Usage":github.com/tony-o/perl6-html-parser-xml/ 23:54
i've previously installed HTTP::Parser::XML
Skarsnik did you write use LWP::Simple ? x)
Herby_ doh! 23:59