»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org | UTF-8 is our friend! | feather will shut down permanently on 2015-03-31
Set by jnthn on 28 February 2015.
00:09 aborazmeh joined, aborazmeh left, aborazmeh joined
atweiden it seems the $Nightscape::conf seen by Actions.pm is independent of the one seen by bin/nightscape.pl 00:18
when i set conf to 3 in Actions.pm, it doesn't effect the conf in nightscape.pl
00:20 skids joined 00:28 laouji joined 00:30 aborazmeh left 00:38 laouji left 00:39 laouji joined 00:40 laouji left, laouji joined 00:47 aborazmeh joined, aborazmeh left, aborazmeh joined 00:52 aborazmeh left 00:54 aborazmeh joined, aborazmeh left, aborazmeh joined 00:57 spider-mario left 00:59 spider-mario joined 01:02 Zoffix joined 01:22 aborazmeh left 01:25 Darkc0de_ joined 01:26 Darkc0de_ left 01:43 telex left 01:44 telex joined 01:47 tinyblak joined 01:53 aborazmeh joined, aborazmeh left, aborazmeh joined 02:00 laouji_ joined 02:02 colomon left, laouji left 02:04 colomon joined 02:06 Foxcool_ left 02:09 Foxcool_ joined 02:11 noganex joined 02:14 noganex_ left 02:49 bjz joined 02:50 laouji joined 02:53 laouji_ left
ugexe so... anyone want to remove 'Parrot' from the github description? :) "Rakudo Perl -- Perl 6 on Parrot, JVM and MoarVM" 03:06
03:09 BenGoldberg left 03:12 mr-foobar left 03:37 laouji left 03:42 aborazmeh left
dalek c: e6e4d7e | skids++ | lib/Language/objects.pod:
Improve example for bare method disambiguity.
 03:44
03:45 mr-foobar joined 04:00 alini joined
dalek c: 1129a3f | skids++ | lib/Language/objects.pod:
Improve example. The showcased construct is redundant with the constraint.
 04:20
04:33 konsolebox joined 04:35 konsolebox left 04:38 laouji joined 04:39 raiph left
dalek c: ee17dc0 | skids++ | lib/Language/objects.pod:
Small elaboration on default constructor.
 04:40
c: 152bf49 | skids++ | lib/Language/objects.pod:
Simplify some circumlocuitous phrasing.
04:42 laouji left
skids finds it strangely appropriate that there are so many circumlocu.* words. 04:43
04:46 laouji joined, atweiden left 04:48 b2gills left, kaare_ joined
dalek p: 8398cee | paultcochrane++ | docs/ops.markdown:
Add links to opcode categories

This makes navigation within the docs easier.		 04:51
p: 1120f1f | paultcochrane++ | docs/ops.markdown:
Use consistent heading capitalisation
p: 468f1cb | paultcochrane++ | docs/ops.markdown:
Wrap long lines to improve readability
p: 0923187 | paultcochrane++ | docs/ops.markdown:
Fix typos
04:51 skids left 04:52 konsolebox joined 04:53 Khisanth joined 04:54 konsolebox left
moritz \o 04:59
yoleaux 14 Apr 2015 21:10Z <brrt> moritz: that doc.perl6.org is awesome. moritz++
04:59 b2gills joined
moritz now *that's* the kind of message you like to read when you get up in the morning. Thanks brrt! 04:59
[ptc] moritz: o/
05:11 konsolebox joined 05:12 konsolebox left 05:15 alini left 05:23 KCL_ joined 05:26 KCL left 05:28 konsolebox joined 05:29 konsolebox left 05:35 KimmoNO joined 05:36 espadrine joined 05:44 laouji left, laouji joined 06:09 konsolebox joined 06:14 konsolebox left 06:18 diana_olhovik joined 06:21 rindolf joined 06:23 yvan1 left
masak morning, #perl6 06:34
moritz++ # doc.perl6.org 06:35
06:41 RabidGravy joined, Rounin joined 06:43 FROGGS joined 06:48 Ven joined
moritz [ptc]++ also deserves some praise 06:55
as do Mouq++ and lots of other committers 06:56
masak is aiming to contribute more, but hasn't gotten the appropriately-shaped tuit yet 07:02
nwc10 I'm sure you're currently getting lots of gifts. Many of which are quite smelly.
masak it's been known to happen. 07:03
[Tux] No such method 'chars' for invocant of type 'Any' 07:14
(yes, I know about the change)
old and dead code
from last year
07:16 pecastro left
[Tux] That change broke internal class calls: 07:19
m: class C {method foo(Str $x){};method bar(Str$x){.foo($x)}};C.new.bar(""); 07:20
camelia rakudo-moar 9862a1: OUTPUT«No such method 'foo' for invocant of type 'Any'␤ in method bar at /tmp/6Nv2Pvn_CF:1␤ in block <unit> at /tmp/6Nv2Pvn_CF:1␤␤»
[Tux] that worked yesterday
now it needs a self.foo
moritz uhm, it didn't work
it just failed to give you the proper error
[Tux] ah :) 07:21
moritz there, finding bugs for you :-)
[Tux] again, just found that in old reference code
weird that it didn't fail, looking back at it
07:22 mr-foobar left
moritz $_ defaulted to Nil, and Nil swallows non-existing method calls on it 07:22
[Tux] I think I like that change 07:23
moritz too
07:24 rindolf left, rurban_ joined 07:25 pdcawley joined 07:28 vendethiel left 07:29 vendethiel joined 07:34 spider-mario left 07:37 FROGGS left
jnthn morning, #perl6 07:41
lizmat morning jnthn 07:42
I was just about to say:
nwc10 .tell jnthn you have no messages!
yoleaux nwc10: I'll pass your message to jnthn.
lizmat commute to Berlin&
so see you all (from) there 07:43
07:43 lizmat left 07:45 rindolf joined 07:46 yvan1 joined 08:02 espadrine left 08:10 cschwenz joined, alini joined 08:12 zakharyas joined 08:20 darutoko joined 08:22 abraxxa joined 08:28 laouji_ joined 08:29 rindolf left 08:32 laouji left
masak [Tux]: $_ and self were disambiguated in the language somewhere around 2005. it was a good decision, because they're really two very different "topic variables". 08:40
[Tux]: it does mean that some people got disappointed that they can't get away with writing `.bar()` for method calls. 08:41
08:41 pecastro joined
masak [Tux]: but actually, not all is lost. people who want to do `.bar()` just need to make sure that $_ contains self. like, make $_ the invocant in the method signature. or do `given self` somewhere inside the method. 08:42
in other words, it's not completely automatic, but (IMO) that's a good thing when you look at the big picture and the language as a whole.
jnthn Also, $.foo(...) 08:43
yoleaux 07:42Z <nwc10> jnthn: you have no messages!
jnthn m)
Or maybe :🌴) 08:44
nwc10 .u 🌴
yoleaux U+1F334 PALM TREE [So] (🌴)
08:45 KCL_ left
nwc10 [Tux] pasted the link www.informationweek.com/it-life/ntp...id/1319432 elsewhere. I found it interesting 08:45
08:46 KCL_ joined 08:49 espadrine joined
Ven m: class A{ method x($_:) { .bar }; method bar{5} }; say A.new.x # masak :P 08:56
camelia rakudo-moar 9862a1: OUTPUT«5␤»
jnthn Yes, that was masak's "make $_ the invocant in the sig" :) 08:57
Ven masak++ # teaching me cute tricks 08:59
masak yes, right. I forgot to mention `$.foo`. but that one comes with some strings attached, if you ask me. 09:00
in that it's much more natural to use it if the method foo is actually "property-like". if it does stuff, then `$.foo` is questionable.
dalek c: cd83ead | (Jonathan Stowe)++ | lib/Language/variables.pod:
Document $*HOME
 09:02
moritz fwiw I think it's quite funny in python
you have to distinguish attribute accesses from method calls on the caller side
and then on the callee side, you can cheat, and make a method appear as an attribute 09:03
(with @property)
so, software that's lying to itself about what it does
</rant>
09:04 FROGGS joined
masak moritz: I think that's some people's beef about get/set properties in JavaScript, too. 09:06
FROGGS o/ 09:07
psch \o
masak o/ 09:08
moritz masak: can you cheat in JS? write a method that's accessed as if it's an attribute?
dalek kudo/nom: 06dfca6 | TimToady++ | src/core/ (5 files):
grab backtraces before information is lost

plus various cleanups to deal with our multi-level exception system
kudo/nom: a7b45b6 | TimToady++ | src/core/ (2 files):
move some code that belongs in Backtrace
cognominal Perl6/Grammar.nqp is quite a haystack 09:10
.*ibbling in it. 09:11
masak moritz: yeah, that's what get/set methods do. 09:13
moritz: javascriptplayground.com/blog/2013/...s-setters/ 09:14
Ven cognominal: do you need some kind of $*NEEDLE? 09:15
from the 6guts blog post, I take that the grant was completed and started? 09:16
Ven doesn't know how the perl foundation works
FROGGS Ven: he has started but the grant still needs to be decided
Ven FROGGS: oke, thanks. 09:17
jnthn I've started under the hopeful assumption the community and TPF will be OK with the grant. :) 09:18
masak I got an email from Karen yesterday, asking if I approved of the grant proposal. (I do, of course.) so my guess is that it's just-about-to-be-approved. 09:19
FROGGS masak: I also got an email, but I don't tell about my reply :P
09:20 dakkar joined
masak (I don't have special approve-p6-grant powers, but it seems I ended up on a list of core p6 devs) 09:20
a list in TPF's possession, I should say.
FROGGS: always so secretive :P
FROGGS will start nagging jnthn about getting rid of the JSON parsing at startup as soon as his grant gets accepted 09:21
because then he has to unblock me! MUHAHHAHAHA
jnthn :P
FROGGS *g*
no really, "just replacing that json blurb with a .perl blurb" does not seem to be a LHF
at least, it is tricky 09:22
jnthn Indeed.
masak .perl is also not known to preserve values exactly
jnthn I'll have to have a look at the JSON blob.
And see how it's being used.
FROGGS because we do that when constructing @*INC, then precompile lib.pm which messes with @*INC, and precompile a module that uses the precompiled lib.pm 09:23
jnthn urgh
FROGGS masak: it is enough for that purpose, really
jnthn That needs a bit of thought, yes.
FROGGS aye
masak: the most exciting value in that blurb is a Version... the rest is about arrays, hashes, strings and numbers 09:24
09:24 Ven left
jnthn Given the current situation is that our startup time is now horrid when you install lots of modules, I should probably look at this relatively soon :) 09:24
If we don't get to it before this month's compiler release, this is probably one of the months we elect not to produce a R* though. 09:25
09:27 laouji_ left, laouji joined
nwc10 that would unfortunate, but I agree, it seems like the lesser of two evils 09:28
09:29 labster joined
jnthn Well, Star exists to provide something more stable; the original schedule for it was once per 3 months. 09:29
09:29 felher left
nwc10 that plan didn't seem to survive contact with the enemy 09:29
or at least, PR 09:30
moritz I guess that's partially my fault
for doing monthly R* releases when it was easy to do
masak well, we simply didn't have a process in place where we decide "should we release a Star"? 09:32
jnthn Well, a more useful conversatoin is "should we ahve a process in plae to decide should we relase a Star, and what should it look like if so?" 09:33
*have, *place
nwc10 will we have Rakudo yar, for pirates? "Is there a yar in the month? Yar! Ship it^W^Wmake it walk the plank" 09:35
jnthn Yes, but it's only released once per year, on Talk Like A Pirate Day :P
timotimo we could ship a rakudo with modified error and warning messages 09:36
on that day
nwc10 is that before or after TimToady's birthday?
it would be an interesting day for an Alpha
jnthn Saturday, September 19
nwc10 oh [unpublishable] 09:37
timotimo hm?
nwc10 e,r hangon 09:38
*this*
www.google.at/search?q=larry+wall's+birthday
jnthn Close! 09:39
09:39 felher joined
nwc10 more usefully, before. but not ages before 09:40
er, context. Pirates are before Larry
timotimo do pirates ever say "sorry"? 09:41
otherwise, there'd have to be something like
masak being a pirate means never having to say "sorry"
timotimo ==AVAST!==
It be a good day for plundering the sea to find that missing block! 09:42
btyler haha. I think I like that better than 'SORRY'
timotimo Pass the Grog, matey! The Syntax is OK! 09:43
er, i meant
"Yer Syntax be OK"
nwc10 this program has been scanned and found clear of all errors, because AVAST!
timotimo ouch 09:44
nwc10 I'm not sure if I assocated the term "AVAST" with warm fuzzy confidence 09:46
09:49 rindolf joined, hoelzro joined 09:55 Akagi201_ left 09:56 Akagi201 joined, Akagi201 left
moritz ok, let's use CLAMAV instead :-) 09:56
==KASPERSKY!==
timotimo has to admit not being very educated about piratey speech 09:57
it's a very difficult thing to pull off if you're living around almost exclusively german-speaking people
psch "how appropriate, you fight like a cow!"
moritz timotimo: and surrounded by land on all sides, too :-) 09:59
timotimo yes :( 10:00
cognominal well, the only pirate in the eu parlement is from germany
10:01 RabidGravy left 10:02 skids joined
nwc10 shame. Would have been funnier if the pirate came from a land-locked country. 10:03
although now-re-reading, it *is* funny in the context here 10:04
10:04 dolmen joined 10:12 araujo joined 10:15 araujo left 10:16 araujo joined 10:27 alini left 10:28 alini joined, alini left 10:29 alini joined, Ven joined 10:30 alini left, alini joined 10:45 araujo left 10:48 sqirrel joined, fhelmberger joined 10:50 dolmen left 10:56 [TuxCM] joined
[TuxCM] m: my Int @i; my Int @x = 4; my @z = @i[@x]; @z.perl.say; 10:56
camelia rakudo-moar a7b45b: OUTPUT«[Int]<>␤»
[TuxCM] I did expect an EMPTY list
10:56 konsolebox joined 10:57 Akagi201 joined
psch m: my Int @x; say @x[0] 10:57
camelia rakudo-moar a7b45b: OUTPUT«(Int)␤»
[TuxCM] m: my Int @i; my Int @x = 4..Inf; my @z = @i[@x]; @z.perl.say;
camelia rakudo-moar a7b45b: OUTPUT«[]<>␤»
psch m: my Int @x; say @x[Inf]
camelia rakudo-moar a7b45b: OUTPUT«Cannot coerce Inf or NaN to an Int␤ in sub postcircumfix:<[ ]> at src/gen/m-CORE.setting:3457␤ in block <unit> at /tmp/XD3Q8E7O0Y:1␤␤»
[TuxCM] m: my Int @i; my Int @x = 4..6; my @z = @i[@x]; @z.perl.say;
camelia rakudo-moar a7b45b: OUTPUT«[Int, Int, Int]<>␤»
[TuxCM] I was told that an array in subscript worked lazy
when the array has a range like ..Inf, it works as I expected 10:58
[TuxCM] has internetz just to the border
10:59 konsolebox left
timotimo hum, are we getting the right candidate? 11:00
the problem could be that my Int @i is undefined 11:01
m: my Int @x; say @x.perl; my Int @y .= new; say @y.perl 11:02
camelia rakudo-moar a7b45b: OUTPUT«Array[Int].new()␤Array[Int].new()␤»
timotimo hum
m: my Int @x; say @x.perl; my Int @y = 1, 2, 3; say @y[@x].perl
camelia rakudo-moar a7b45b: OUTPUT«Array[Int].new()␤()␤»
timotimo here i get an empty list
m: my Int @x; say @x.perl; my Int @y = 1, 2, 3; my @z = @y[@x]; say @z.perl 11:03
camelia rakudo-moar a7b45b: OUTPUT«Array[Int].new()␤[]<>␤»
timotimo huh? now what am i doing differently from your thingie?
m: my Int @x; my Int @y = 1; my @z = @y[@x]; say @z.perl
camelia rakudo-moar a7b45b: OUTPUT«[]<>␤»
timotimo oh
i've got @x and @y reversed
m: my Int @x; my Int @y = 1; my @z = @x[@y]; say @z.perl
camelia rakudo-moar a7b45b: OUTPUT«[Int]<>␤»
timotimo well, that does make sense, doesn't it?
if you want it to blow up on positional access outside its "bounds", you may want a shaped array instead? 11:04
(nyi, last i tried it)
jnthn A range 1..* works becuase the * stands in for the number of elements 11:05
[TuxCM] in @y[@x], @x=1..Inf behaves completely different from @x=1..2
jnthn And 1..Inf is taken to do the same
[TuxCM] m: my Int @i = 2..*;
camelia ( no output )
timotimo oh, interesting
so if you (1..Inf).list ...
psch m: my Int @x; say @x[1..Inf] 11:06
camelia rakudo-moar a7b45b: OUTPUT«␤»
[TuxCM] so I can replace Inf with * ?
jnthn m: my @a = 1..10; say @a[3..*];
camelia rakudo-moar a7b45b: OUTPUT«4 5 6 7 8 9 10␤»
[TuxCM] Interesting ...
jnthn m: my @a = 1..10; say @a[3..11];
camelia rakudo-moar a7b45b: OUTPUT«4 5 6 7 8 9 10␤»
timotimo m: say (1..*).WHAT
jnthn Oh, seems it always trims at the end with a range.
camelia rakudo-moar a7b45b: OUTPUT«(Range)␤»
psch got his tr/// mergeable again o/
but it's a bit icky 11:07
jnthn m: my @a = 1..10; say @a[my @ = 3..11];
camelia rakudo-moar a7b45b: OUTPUT«4 5 6 7 8 9 10 (Any) (Any)␤»
jnthn But not for the Array case.
[TuxCM] HELP! :)
psch i have to hand a reference to the outer-most $/ through the trans proto/cands and through nearly all of LSM :/
jnthn But that also likely makes a lot of sense.
11:07 alini_ joined
jnthn Becuase for the Array case you've no promise, like with a rnage, that the elements are contiguous. 11:08
[TuxCM] I have code that checks against the right side of a range
$i < Inf works, $i < * does not (I guess)
11:08 alini left, alini_ is now known as alini
jnthn m: say (1..*).perl 11:09
camelia rakudo-moar a7b45b: OUTPUT«1..Inf␤»
timotimo well, $i < * just gives you a WhateverCode
that's syntactic
jnthn Note that 1..* constructs a range 1..Inf
timotimo and it'll likely eat more code towards the outside of that statement, too
[TuxCM] OK, I have about 4½ hours to play with that
jnthn So if you are doing $i < $range.max then you're good
Though $i ~~ $range "$i falls within the range" is probably better overall
Because it takes care of ..^ and ^.. ranges too :) 11:10
timotimo fixes the Junction optimizer in rakudo 11:12
[TuxCM] germany. network will drop any time now
jnthn Germany, land of no network! /o\ 11:13
skids what, no famous german network engineering?
jnthn
.oO( die Network, die! )
 11:15
11:16 konsolebox joined
psch t/spec/S05-transliteration/trans.rakudo.moar (Wstat: 0 Tests: 62 Failed: 0) TODO passed: 28, 31-34, 36-40, 45-46 11:17
i'll merge that now, if the CALLERS:: stuff i'm doing is too weird, please shout :)
11:19 [TuxCM] left
psch actually, there was a stray commit with the message "fixup" around 11:21
so i cleaned that up first, and'll rerun spectest 11:22
11:23 konsolebox left
Ven wants to talk about @a[x..*] but doesn't know where... 11:25
11:26 rurban_ left
Ven Where is it I've already talked about whatever/closure, about ranges, but still near arrays so that it doesn't appear out of nowhere 11:27
psch oh, in the learnxiny? 11:28
psch was a bit confused at first, thinking "why not talk about it here?"...
grr 11:29
Ven psch: yeah. It needs some solid yak shaving, but I'm not good enough at writing to do it I think
psch git is hard, i apparently undid all my repairing of the tr PR ._. 11:30
11:31 sqirrel left, Akagi201 left 11:32 Akagi201 joined
Ven psch: did you commit or stash at some point? 11:33
psch Ven: no, but i luckily kept vim open and could undo
Ven ah, grats :P
psch is glad he made :e! a habit
Ven m: my @a = ^20; @a[5..*]>>++; say @a 11:34
camelia rakudo-moar a7b45b: OUTPUT«0 1 2 3 4 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20␤»
psch instead of e.g. ZQ↑^M or whatever... 11:35
(or ZQ↑↵)
i like urxvt's KEYCAP INSERT MODE :p 11:36
+PICTURE
there 11:38
dalek kudo/nom: 85cbdc2 | peschwa++ | src/ (3 files):
Work towards implementing :squash for &Str.trans.

Conflicts: src/core/Str.pm 6830781 | peschwa++ | src/core/Str.pm: Remove debug output, masak++.
11:38 dalek left 11:39 dalek joined, ChanServ sets mode: +v dalek
psch github.com/rakudo/rakudo/commit/ec...bce05122f9 11:39
11:40 skids left
psch ^^^ that's the "oh god why" commit in that merge 11:40
seeing as every scope has its own $/ (and gist.github.com/peschwa/4fe16e5c5aa2e5b8370c is what happens) i did it like that
masak um. are you adding debug output statements in a commit on nom? 11:41
psch oh fffs
i'll remove those, sorry
masak :/
maybe consider working in a branch?
that way, you also have bigger freedom to rebase interactive and stuff.
psch it was a branch, i just lost track of the debug prints when merging and rebasing 11:42
which went all over the place a few times
masak maybe a good idea to simply not commit debug prints. you can review a patch before you turn it into a commit, making sure that the code that goes in is intentional.
masak is there any documentation about :squash for .trans? I haven't heard about it before.
psch masak: i've implemented it as per S05-transliteration/trans.t 11:43
masak oh, good.
I'll look there, then.
arnsholt Reviewing commits is a good idea in general. I have "idxd = diff --cached" the [alias] section of my ~/.gitconfig
11:44 iv_ is now known as iv
masak also variations of `git commit -p` tend to work quite well. 11:45
Ven Does the spec request that no more elements are reifed than needed? 11:46
psch -p sounds like something i'll want to use habitually
arnsholt Yeah, add -p as well 11:47
Ven psch: DEBUG_PLS? :P
.oO( RAKUDEBUG )
masak arnsholt: `git add -p`, `git checkout -p`, `git reset -p <committish>`, `git stash -p`... 11:49
Ven removes the part about printing 1..* giving an infinite loop in the REPL 11:50
arnsholt Yeah, I <3 the -p option 11:51
Fits so well with the way I tend to work; start on one thing, shave a yak or two, then get the actual thing I wanted to do
[ptc] has `git commit -p` mapped to <leader>c in vim
jnthn is a heavy -p-er too
11:52 tinyblak_ joined
jnthn Though I tend to ues git reset -p without a committish more often. :) 11:52
masak oh, indeed
11:55 tinyblak left
Ven I removed the GLR part from the learnx because it was confusing, I'll add it back once it comes to be 11:56
dalek c: fb6f481 | paultcochrane++ | lib/Language/control.pod:
Make the repeat/while docs viewable
cognominal GLR? 11:58
timotimo Great List Refactor
m: say &[<].signature.params[0].type
camelia rakudo-moar a7b45b: OUTPUT«(Mu)␤»
timotimo ^- why is this Mu now?
cognominal l'Arlésienne? :)
arnsholt timotimo: Has to be Mu to work with junctions, no? 11:59
jnthn m: say &[<].signature.perl
camelia rakudo-moar a7b45b: OUTPUT«:($?, $?)␤»
timotimo arnsholt: it has to be Mu to also handle junctinos without being folded
jnthn Note that multi-dispatch failure also falls back to Junction dispatch 12:00
And that's a more efficient way to go, iirc 12:01
timotimo jnthn: if we find any parameter that doesn't istype(*, Any), we bail out of the junction unfolding optimization
so the evaluation above is why "so 5 < 1|2|3" doesn't unfold any more 12:02
masak m: say 5 < 1|2|3
camelia rakudo-moar a7b45b: OUTPUT«any(False, False, False)␤»
masak m: say so 5 < 1|2|3
camelia rakudo-moar a7b45b: OUTPUT«False␤»
masak m: say so 5 < 1|9|3
camelia rakudo-moar a7b45b: OUTPUT«True␤»
jnthn timotimo: If it's a multi, the optimizer could ask it if it has candidates that'll bind a Junction arg there 12:04
psch oh, i've noticed an inconsistency :/
jnthn The candidate list is known at compile time.
timotimo OK, so i'll have to look at the whole thing 12:05
the whole candidate list, that is
psch Perl6/Actions.nqp says the short names for the adverbs for tr/// are cm, sq and d, but core/Str.pm has them as c, s, d
jnthn timotimo: But you shouldn't really have to re-implmenet it all
psch i suspect that's me only converting half of them, but i'm not sure which are the right ones
timotimo but this code used to work, i think
jnthn timotimo: Since the multi-dispatcher should be able to tell you.
timotimo so maybe at some point $? parameters went from Any to Mu?
psch fwiw, roast has them as single letters
timotimo right, it should be able to, i'll figure it out
jnthn No, in CORE.setting we explicitly have them as Mu
(just checked) 12:06
timotimo interesting
i don't know why we would, though
does it have to be that to allow for native int arguments and such?
jnthn No 12:08
arnsholt Natives are subtypes of Int/Num, so no need for Mu there
jnthn Right
I think it's to make sure the multi-dispatcher does the auto-threading. 12:09
'cus it knows how to insert a "do the auto-thread" thing into the multi cache
Which works out a lot cheaper than a bind failover.
12:11 aborazmeh joined, aborazmeh left, aborazmeh joined
dalek c: 376253f | paultcochrane++ | lib/Language/control.pod:
Document the C<loop> statement

  ... by shamelessly copying the text from the spec....		 12:13
timotimo ok 12:14
[Coke] ugexe: I just checked, I don't have rights to update the project description, it seems. 12:16
dalek kudo-star-daily: 204c5d4 | coke++ | log/ (2 files):
today (automated commit)
12:17 Akagi201 left 12:20 Akagi201 joined, telex left 12:21 aborazmeh left 12:22 telex joined 12:29 raiph joined 12:32 tinyblak_ left 12:45 yqt joined 12:47 tinyblak joined
timotimo now i'm not totally sure any more how to ask the multi-dispatcher if junction fallbacks "will happen" or not 12:48
12:53 laouji left 12:56 yqt left
nine FROGGS: I've seen something about replacing packlists on the agenda for the PQAH. Maybe we find a solution that works for both Perl 5 and Perl 6? 12:56
Ven timotimo: I must've not read the explanation correctly - what's wrong with look at Mu? 12:59
looking*
FROGGS nine: the first you'd have to do is to explain what a packlist is then :o) 13:00
timotimo Ven: you know what my "junction unfolding" optimization does?
Ven timotimo: transforms $a == 3 | 4 to $a == 3 || $a == 4 ?
timotimo yes 13:02
nine FROGGS: it's just a file containing a list of file names that belong to a distribution
timotimo you know what happens when you pass a junction in a parameter that's typed Mu?
nine nine@sunshine:~/install/roast (master=)> cat /usr/lib/perl5/site_perl/5.20.1/x86_64-linux-thread-multi/auto/Time/Mock/.packlist
/usr/lib/perl5/site_perl/5.20.1/Time/Mock.pm 13:03
/usr/share/man/man3/Time::Mock.3pm
FROGGS nine: ahh, our META6.json (described in S22) 13:06
Ven timotimo: it binds it instead of "junctioning" 13:11
timotimo correct 13:12
so in that case the junction unfolding optimization isn't allowed to be done 13:14
13:27 rurban_ joined
[Coke] S22:1 13:47
synbot6 Link: design.perl6.org/S22.html#line_1
13:57 FROGGS left 13:58 adu joined
Ven \o/ synbot6 14:01
masak my inner nwc10 simulation says "do not stop until everything is *bot6" :) 14:02
Ven nicks venbot6 14:04
nwc10 masak++ # well trained :-)
Ven (though really, I'm 20 already..)
nwc10 but, this is the first live IRC bot made with real dogfood?
moritz no
we've had other p6 bots on and off 14:05
nwc10 OK
are we logging the perl 5 channels with a perl 6 bot yet? :-)
masak oh, that would be swell
14:06 cognominal left
moritz nwc10: no 14:06
I have the code to do it, but it simply isn't reliable enough (doesn't reconnect on errors, for example)
dalek kudo/nom: 239d74f | peschwa++ | src/Perl6/Actions.nqp:
Correct tr/// adverbs to match .trans named parameters.
 14:13
14:14 FROGGS[mobile] joined 14:16 Ven left 14:19 Ven joined 14:20 Ven left 14:29 bjz left
TimToady m: 3 % 0; 14:31
yoleaux 14 Apr 2015 22:35Z <japhb> TimToady: What was the purpose of rakudo commit 872d74cd2c320217894a3b0d5331b4c625b13022 (don't cache Rat literals)?
camelia rakudo-moar 683078: OUTPUT«Earlier failure:␤ Divide by zero using infix:<%>␤ in block <unit> at /tmp/t_DavkAgGK:1␤␤Final error:␤ Type check failed for return value; expected 'Int' but got 'Failure'␤ in any return_error at src/vm/moar/Perl6/Ops.nqp:639␤ in sub infix:…»
TimToady .tell japhb Er, to not cache Rat literals? I don't think there are are that many common ones, unlike Ints. 14:35
yoleaux TimToady: I'll pass your message to japhb.
TimToady Are! There be too many! 14:36
14:37 alini left 14:38 grondilu joined
grondilu m: say first { sqrt([+] @$_ X** 2) ~~ Int }, [(^250).pick(4)] xx *; # rosettacode.org/wiki/Euler%27s_sum_...conjecture 14:38
14:38 Rounin left
camelia rakudo-moar 683078: OUTPUT«(timeout)» 14:38
grondilu yeah that was probably too optimistic 14:39
moritz grondilu: sqrt won't ever return an Int 14:41
m: say sqrt(4).^name
camelia rakudo-moar 683078: OUTPUT«Num␤»
moritz m: say sqrt(4).narrow.^name 14:42
camelia rakudo-moar 683078: OUTPUT«Int␤»
moritz might work with .narrow
grondilu testing for integerness should be a bit easier, imho 14:47
14:48 prammer joined
[Coke] m: say first { sqrt([+] @$_ X** 2).narrow ~~ Int}, [(^250).pick(4)] xx * 14:50
camelia rakudo-moar 239d74: OUTPUT«83 28 158 62␤»
moritz grondilu: you can also test against { .Int == $_ }
[Coke] .narrow seems like it could be less wasteful than a full conversion. (but probably isn't) 14:51
moritz .narrow *does* a conversion 14:52
14:54 estrabd joined 14:56 molaf_ joined 14:58 molaf__ left 15:01 skids joined, lichtkind joined
hoelzro o/ #perl6 15:05
grondilu m: (my %fifth){$_}++ for ^250 X** 5; say first { %fifth{[+] @$_ X** 5} }, combinations(250, 4); 15:10
camelia rakudo-moar 239d74: OUTPUT«(timeout)»
15:10 skids left 15:11 muraiki joined, yvan1 left
moritz m: (my %fifth){$_}++ for ^250 X** 5; say %fifth.keys.sort[^20] 15:12
camelia rakudo-moar 239d74: OUTPUT«0 1 100000 10000000000 101621504799 1024 102400000 1048576 104857600000 10510100501 1073741824 108175616801 11040808032 111577100832 115063617043 115856201 11592740743 1160290625 118636749824 11881376␤»
moritz m: (my %fifth){$_}++ for ^250 X** 5; say %fifth.keys.sort(+*)[^20]
camelia rakudo-moar 239d74: OUTPUT«0 1 32 243 1024 3125 7776 16807 32768 59049 100000 161051 248832 371293 537824 759375 1048576 1419857 1889568 2476099␤»
TimToady m: (my %fifth{Any}){$_}++ for ^250 X** 5; say %fifth.keys.sort[^20] 15:13
camelia rakudo-moar 239d74: OUTPUT«0 1 32 243 1024 3125 7776 16807 32768 59049 100000 161051 248832 371293 537824 759375 1048576 1419857 1889568 2476099␤»
TimToady m: (my %fifth{Int}){$_}++ for ^250 X** 5; say %fifth.keys.sort[^20] 15:14
camelia rakudo-moar 239d74: OUTPUT«0 1 32 243 1024 3125 7776 16807 32768 59049 100000 161051 248832 371293 537824 759375 1048576 1419857 1889568 2476099␤»
grondilu m: (my %fifth){$_}++ for my @fifth = ^250 X** 5; say first { %fifth{[+] @fifth[@$_]} }, combinations(250, 4);
camelia rakudo-moar 239d74: OUTPUT«(timeout)» 15:15
grondilu ^this should be quite fast
15:15 rurban_ left 15:16 cognominal joined
moritz m: my \u = 100; (my %fifth){$_}++ for my @fifth = ^u X** 5; say first %fifth{[+] @fifth[@$_]} }, combinations(u, 4); 15:16
camelia rakudo-moar 239d74: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ZuMGcMaVNn␤Unexpected closing bracket␤at /tmp/ZuMGcMaVNn:1␤------> 3* 5; say first %fifth{[+] @fifth[@$_]} 7⏏5}, combinations(u, 4);␤»
moritz m: my \u = 100; (my %fifth){$_}++ for my @fifth = ^u X** 5; say first { %fifth{[+] @fifth[@$_]} }, combinations(u, 4); 15:17
camelia rakudo-moar 239d74: OUTPUT«(timeout)»
moritz m: say combinations(10, 4).perl
camelia rakudo-moar 239d74: OUTPUT«([0, 1, 2, 3], [0, 1, 2, 4], [0, 1, 2, 5], [0, 1, 2, 6], [0, 1, 2, 7], [0, 1, 2, 8], [0, 1, 2, 9], [0, 1, 3, 4], [0, 1, 3, 5], [0, 1, 3, 6], [0, 1, 3, 7], [0, 1, 3, 8], [0, 1, 3, 9], [0, 1, 4, 5], [0, 1, 4, 6], [0, 1, 4, 7], [0, 1, 4, 8], [0, 1, 4, 9], [0,…»
grondilu m: say [*] (100 ... 0) Z/ 1 .. 4; 15:18
camelia rakudo-moar 239d74: OUTPUT«3921225␤»
moritz m: sub postfix:<!>($x) { [*] 1..$x }; say 250! / (4! * 246!)
grondilu m: say [+] (rand for ^3921225);
camelia rakudo-moar 239d74: OUTPUT«1960141.78214114␤» 15:19
rakudo-moar 239d74: OUTPUT«158882750␤»
grondilu m: say [+] (rand for ^[*] (250 ... 0) Z/ 1 .. 4);
camelia rakudo-moar 239d74: OUTPUT«(timeout)» 15:20
15:20 raiph left
[Coke] Yes, I know it does. but it seems like that's the easiest implementation, and that there might be a faster way to get the narrowed type. 15:25
15:27 diana_olhovik left
grondilu m: say 1.0 ~~ Int # to me that being false is LTA 15:27
camelia rakudo-moar 239d74: OUTPUT«False␤»
grondilu m: say 10/2 ~~ Int # same for this 15:28
camelia rakudo-moar 239d74: OUTPUT«False␤»
grondilu m: say 1 + 0i ~~ Int # and that 15:29
camelia rakudo-moar 239d74: OUTPUT«False␤»
15:29 Possum left
TimToady that behavior would also be suboptimal, just differently :) 15:30
15:31 adu left
grondilu or maybe we could have a different operator. 15:31
15:31 zakharyas left
TimToady that's why we have .narrow 15:31
m: say (1+0i).narrow ~~ Int
camelia rakudo-moar 239d74: OUTPUT«True␤»
grondilu well, yeah. 15:32
TimToady maybe you want something on the Int side that says "can narrow to Int without loss of round-tripping" 15:33
Int() doesn't quite do that
grondilu std: Int(Num)
camelia std 28329a7: OUTPUT«ok 00:00 135m␤»
grondilu m: say 1 + 0i ~~ Int(Num)
camelia rakudo-moar 239d74: OUTPUT«False␤»
grondilu ^ would that make some sense? 15:34
TimToady not before more coffee, and probably not after either :)
m: say 1 + 0i ~~ Int(Real) 15:35
camelia rakudo-moar 239d74: OUTPUT«False␤»
grondilu this is something that bothers me sometimes. In maths, integers form a subsets of the rationals, which form a subset of real numbers etc. But in programming languages, that's not how things work. 15:36
I mean, a Rat should be a Int as long as its denominator is 1.
subset Int of Rat where *.nude[1] == 1; 15:37
subset Real of Complex where *.im == 0; # and so on.
it's like everything is upside down in the core. 15:38
TimToady m: say Int ~~ Rational 15:41
camelia rakudo-moar 239d74: OUTPUT«False␤»
TimToady I thought that used to be true...
star-m: say Int ~~ Rational
camelia star-m 2013.03: OUTPUT«False␤»
moritz never
[Coke] m: say Int ~~ Rat 15:42
camelia rakudo-moar 239d74: OUTPUT«False␤»
moritz m: say Int ~~ Real
camelia rakudo-moar 239d74: OUTPUT«True␤»
TimToady ah, that were it
moritz Rational is the thing with nu and de :-)
TimToady m: say Rational[Int,1]
camelia rakudo-moar 239d74: OUTPUT«(Rational[Int,Int])␤»
TimToady oopsie :) 15:43
I think I was speculating about defined a fixed point type as Rational[Int,100000] or so 15:44
"assumed decimal point"
15:44 Possum joined 15:48 Akagi201 left 15:51 tinyblak left 15:55 RabidGravy joined 16:03 skids joined 16:05 alini joined 16:10 anaeem1 joined, robinsmidsrod left, robinsmidsrod joined
nine S11 talks about different versions of a compunit. How can those exist? As I understand it, I'd say different versions of a module are different compunits 16:12
16:14 spider-mario joined 16:15 cschwenz left 16:20 eli-se joined 16:23 nbdsp joined 16:24 RabidGravy left
nbdsp Greetings! Could someone advise please how to determine the object identity of the caller of a function. Something like CALLER::<WHICH>? 16:26
16:31 anaeem1 left, yvan1 joined, anaeem1 joined 16:32 raiph joined
moritz nbdsp: I don't think you can 16:33
nbdsp: what do you need it for?
nbdsp moritz: I want to emulate protected class members by checking the identity of the caller. 16:34
moritz nbdsp: but for that you'd need the package of the caller, no? 16:35
not the identity
FROGGS[mobile] bartolin: are you comming to the lobby or are you hunting for the Kuchenkaiser then?
nbdsp It depends on the definition of 'protected': class-wise or object-wise. But how to check the caller's package? 16:36
eli-se hi 16:37
moritz nbdsp: either callframe() offers that, or I'm sure you can get to it though the backtrace
16:38 RabidGravy joined
moritz doc.perl6.org/type/Backtrace 16:38
nbdsp I just thinked of a better way: add some checking function to derived class methods and call it from the base class. 16:40
But isn't it indeed impossible to get the WHICH data from the caller? 16:42
b2gills m: sub postfix:<!>( UInt() $x ) is tighter(&[*]) { [*] 2..$x }; #`( I'm not sure its precedence is correct ) say $_! for 0..5,-1 16:45
camelia rakudo-moar 239d74: OUTPUT«1␤1␤2␤6␤24␤120␤Earlier failures:␤ Coercion to UInt out of range. Is: -1, should be in 0..Inf␤ in sub postfix:<!> at /tmp/Xgg4cpAM7S:1␤ in block <unit> at /tmp/Xgg4cpAM7S:1␤␤Final error:␤ Cannot call Real(Failure: ); none of these …»
16:48 abraxxa left 16:49 Akagi201 joined 16:50 SHODAN left 16:51 SHODAN joined
moritz nbdsp: well, in the VM we have the callee's callframe 16:51
nbdsp: but what good would it do you if you had the callframe's .WHICH?
16:51 tinyblak joined
moritz nbdsp: it just tells you if the current call(stack)frame has called you before (if you store the previous ones) 16:52
nbdsp: anyway, if you really do want protected emulation (which is a bad idea enough in itself), then you should do it statically and through the MOP
16:56 gfldex joined 16:57 tinyblak left, alini left
nbdsp moritz: isn't 'self' is defined in the outer callframe scope? Is it possible to reference 'self' in the caller's scope? 16:57
16:57 dakkar left 16:58 Rounin joined
bartolin FROGGS[mobile]: I'll go to the Kuchenkaiser directly, I think. are you already there? 16:59
17:00 mr-foobar joined
FROGGS[mobile] bartolin: we just arrived 17:00
bartolin FROGGS[mobile]: great. I'll need about half an hour .. 17:02
FROGGS[mobile] \o/ 17:03
nbdsp moritz: you are right, anyways: protected members violate encapsulation and it'd better to implement getters/setters in the base class. 17:05
Though getters/setters also violate encapsulation if can be world called. Protected members express the idea of private data common to the base/derived hierarchy. I don't see why such a shared private data is neccesseraly bad idea. 17:15
17:23 KCL joined 17:26 KCL_ left
eli-se lol 17:27
moritz nbdsp: it's bad because it force everybody who wants to access the data to inherit from your class 17:37
nbdsp: and since child classes can access protected stuff, and anybody can be your child class, it's part of the public API anyway, and must be documented and tested just like the rest of the public API 17:38
nbdsp: so, why not make it public right away?
nwc10 moritz: I forgot to ask earlier - is the code for your Perl 6 IRC logger client online anywhere? 17:39
moritz nwc10: I have no idea
nwc10 that wash't an answer I expected.
"Yes" or "no" or possibly "soon"
but you surprised me.
moritz nwc10: if not, it's likely lost, since my laptop hard disk died two days ago
17:39 [TuxCM] joined
nwc10 oh :-( 17:40
poor laptop
sadly that also explains that "no idea", because you can't ask your laptop
moritz correct 17:41
17:44 espadrine left
skids nbdsp: It isn't implemented yet IIRC, but if you really must have protected attributes, you would put them in a role with a "trusts" statement. 17:45
17:46 nbdsp left, [TuxCM] left
skids ndsp: but for the reasons mortiz mentioned, you just end up making a mess because someone will "need" to get at them and abuse a subclass to do it. 17:46
geekosaur .oO {the rigor of mortiz } 17:52
17:53 brrt joined 17:55 brrt left 17:56 raiph left 17:57 brrt joined
skids My typing deteriorates on a full stomach :-) 17:58
18:01 eli-se left 18:06 telex left 18:08 telex joined 18:11 yvan1 left 18:15 raiph joined 18:27 eli-se joined, andreoss joined 18:28 Vlavv_ left 18:32 cognominal left
moritz skids: trusts is implemented 18:39
skids moritz: Oh cool, I'll have to ponder whether I had a use for it. 18:40
18:40 Vlavv_ joined
moritz fwiw rakudo uses trusts in one place in its setting 18:43
Promise trusts Vow (and Vow is a private class within Promise)
18:51 vendethiel- joined 18:52 brrt left, yqt joined 18:53 vendethiel left 18:57 yqt left
timotimo 19:04
sorry, cat was at it again
19:06 kaare__ joined, kaare_ left 19:07 rindolf left 19:11 rindolf joined 19:17 robinsmidsrod left 19:18 alini joined 19:19 robinsmidsrod joined 19:25 pdcawley left 19:30 rindolf left 19:31 andreoss` joined 19:35 andreoss left 19:38 andreoss` left
TimToady psch: I almost never use :e! since I have autowrite set, and typically write it out by suspending 19:42
though I don't suppose that'd work so well for gvim... 19:43
psch TimToady: i used :e! after switching branches
TimToady: and thus could undo beyond the branch switch
or before rather 19:44
TimToady sure, and I can undo after a suspend just the same way
psch oh, right
TimToady I typically leave all my vim sessions in the background while compiling
or when doing anything else
psch i'm mostly running screen
TimToady works fine there too 19:45
grondilu &/quit
19:45 grondilu left
psch well, yeah. screen is why i don't suspend, and i :e after switching branches to look 19:45
there's probably potential to improve my workflow, seeing as i overlooked debug says... :) 19:46
arnsholt Bound to happen eventually, to be fair. That's why reviews are a thing =) 19:47
19:47 Zoffix left
TimToady also relies heavily on vim's undo as a mini-revision-control system 19:48
psch m: $_ = "bookkeeper"; tr:s/a..z/a..z/; .say
19:48 Zoffix joined
camelia rakudo-moar 239d74: OUTPUT«bokeper␤» 19:48
19:48 Zoffix left, Zoffix joined
psch i've read again and again about undotree, but i'm usually too lazy/(scared of data loss) to heavily customize my environment 19:49
the bit of escape codes that puts PWD and $0 (plus filename for vim) into screens ^A" is about as far as i'll go 19:50
otherwise i'll just rename the screens manually and that's something that should be automated
19:54 tinyblak joined 19:58 tinyblak left 20:05 kaare__ left 20:07 yvan1 joined, dalek left 20:08 dalek joined, ChanServ sets mode: +v dalek 20:15 vendethiel- left
dalek c: 6dfe113 | moritz++ | WANTED:
Add Str.substr-eq to WANTED
 20:16
20:20 darutoko left, vendethiel joined
dalek ecs: 724a503 | paultcochrane++ | S04-control.pod:
Fix minor typo
 20:20
20:29 dalek left 20:30 dalek joined, ChanServ sets mode: +v dalek 20:31 mohij joined 20:39 eli-se left
dalek c: 005b7ca | (Steve Mynott)++ | lib/Type/Proc/Async.pod:
typos
 20:41
ast: e1b2360 | peschwa++ | S05-transliteration/trans.t:
Unfudge now passing tests.
 20:43
masak 'night, #perl6 20:51
raydiak good night masak
good afternoon #perl6 20:52
20:56 nbdsp joined
dalek c: 6b814dd | (Steve Mynott)++ | lib/Type/ (2 files):
moar typos
 20:58
20:58 eli-se joined
nbdsp Greetings! Could someone advise please is it possible to reference 'self' variable vesible in a class method from a called function? Something like this: sub foo { say CALLER::<$self.WHAT> }; class C { method meth { foo } } 21:00
PerlJam nbdsp: that method has no "$self variable" 21:05
nbdsp Well, yes, that was a pseudocode. Is it possible to reference from 'foo' a variable which in C::meth is referenced as 'self'? 21:07
21:08 anaeem1 left
moritz you can only access variables through caller if they were declared as dynamic variables 21:10
and self isn't
psch m: class A { method meth($*self:) { f } }; sub f { say $*self }; A.new.meth; # o.o
camelia rakudo-moar 239d74: OUTPUT«A.new␤»
psch i don't think that's a sensible thing to do, fwiw 21:11
vendethiel ...that's amazing. 21:12
paging eli-se for some reason.
eli-se paging me?!
vendethiel pinging
eli-se oh
I'm confused.
jercos help, my pager is beeping and it won't stop 21:13
vendethiel eli-se: "method x($*self:)" dynamic caller.
eli-se terrible 21:14
21:14 bjz joined
eli-se I avoid dynamic variables like the plague until they're fixed (i.e. namespaced). 21:14
nbdsp moritz: not neccesseraly through CALLER. Maybe through OUTER or somehow else? 21:15
japhb .botsnack 21:17
yoleaux :D
synbot6 om nom nom
yoleaux 14:35Z <TimToady> japhb: Er, to not cache Rat literals? I don't think there are are that many common ones, unlike Ints.
b2gills m: sub test { CALLERS::<&?ROUTINE>.package.d }; class C { method d { say 'hi'}; method a { test }}; C.new.a
camelia rakudo-moar 239d74: OUTPUT«hi␤»
japhb Ah, so it was simply a matter of "the hit ratio will probably be low, so why cache these at all?"
PerlJam "CALLERS"? 21:18
PerlJam learns something new
nbdsp b2gills: thanks!
21:20 alini left 21:21 skids left 21:22 RabidGravy left
b2gills Note that it doesn't call the method on an instance of the object 21:23
psch which means it doesn't actually get self but the type object 21:25
21:25 coffee` joined
psch m: class A { has $.foo = "bar"; method m { baz } }; sub baz { CALLERS::<&?ROUTINE>.package.foo.say }; A.new.m 21:26
camelia rakudo-moar 239d74: OUTPUT«Cannot look up attributes in a type object␤ in method foo at src/gen/m-CORE.setting:4009␤ in sub baz at /tmp/kC7eF3Xycb:1␤ in method m at /tmp/kC7eF3Xycb:1␤ in block <unit> at /tmp/kC7eF3Xycb:1␤␤»
21:26 cschwenz joined
cschwenz Hello #perl6! 21:27
psch although iirc the point was implementing «protected», so that doesn't matter
hi cschwenz \o
hoelzro o/ cschwenz 21:28
good to see you finally showing your face here =)
cschwenz I need some help with getting perl6 to run via a #! line.
o/ hoelzro
pastebin.com/1uqxs1kT
what am i doing wrong?
hoelzro cschwenz: what about "type perl6" instead of "which perl6"? 21:29
cschwenz result is: perl6 is hashed (/usr/local/bin/perl6)
hoelzro your shell may be using a hashed path for perl6
hmm
ok
psch cschwenz: it's not really you doing something wrong. perl6 is itself a bash script, but most *nix-y shells don't chain shebangs
err, maybe not "most"
and maybe it's not the shells fault
i do know that linux allows it... :) 21:30
irclog.perlgeek.de/perl6/2015-01-31#i_10038252 21:31
cschwenz ah, that's good to know :-)
PerlJam Does env fix it? 21:32
cschwenz \o/ #!/usr/bin/env perl6 works on OS X
:-D
thanks all! 21:33
and i learned perl6 is a bash script :-P
TimToady shh!
japhb Any sufficiently advanced Bash script is indistinguishable from Perl 6 21:34
TimToady or I gues I should just say sh!
*ss
japhb I guess I should have said 'Any sufficiently advanced Bash script is indistinguishable from `perl6`' 21:35
.ask timotimo What becamse of JSON::Fast? Was the faster parser merged into Rakudo?
yoleaux japhb: I'll pass your message to timotimo.
21:37 eli-se left 21:46 robinsmidsrod left 21:47 robinsmidsrod joined 21:49 pyrimidi_ joined 21:56 lichtkind left 21:57 muraiki left 22:03 lizmat joined 22:05 vendethiel left 22:07 vendethiel joined 22:13 mohij left
cschwenz does unlink do the same thing in perl6 as it does in perl5? (asking because doc.perl6.org/routine/unlink is 404 :-/ ) and if so, would it be called on the file-handle ($fh.unlink) or in the exact same way as perl5 (unlink $path)? 22:22
lizmat .tell nine S22:235 specifies depends as "A list of run-time dependencies, specified as C<use> strings." 22:24
yoleaux lizmat: I'll pass your message to nine.
synbot6 Link: design.perl6.org/S22.html#line_235
lizmat .tell nine so you can depend on a specific version / auth of a module
yoleaux lizmat: I'll pass your message to nine.
lizmat cschwenz: "foo".IO.unlink would unlink the file "foo" 22:25
unlink "foo" would do the same
cschwenz ah, thanks :-)
is the 404ed page intentional (that is, is it in flux)? 22:26
also, fwiw, doc.perl6.org/type/IO has no mention of unlink. 22:29
lizmat well, I'm now at the QA Hackathon in Berlin 22:30
my personal goal is to not have to deal with the newio branch by the end of the Hackathon
cschwenz :-)
lizmat I'm not sure yet what that means :-)
fact is, I lost track of what is already in nom and what is still in newio 22:32
and what is obsolete, and what isn't 22:33
anyway, first getting some sleep now
&
cschwenz good night lizmat! and thanks for the help. 22:34
22:35 gfldex left 22:37 lizmat left 22:43 rararara_ joined 22:44 KimmoNO left 22:46 dolmen joined 22:49 Rounin left 22:55 KimmoNO joined 22:57 dolmen left 22:59 cschwenz left 23:06 avuserow left, avuserow joined 23:07 JimmyZ joined 23:11 skids joined
timotimo japhb: it was not merged yet. it doesn't properly complain about numbers illegaly starting with a 0 23:12
yoleaux 21:35Z <japhb> timotimo: What becamse of JSON::Fast? Was the faster parser merged into Rakudo?
timotimo i'd love for somebody to actually try it out a bit :) 23:13
give me more benchmark data and stuff
and before it gets into rakudo, i may want to fix up the "junction unfolding" optimization and write pieces of JSON::Fast to be less ugly
23:18 andreoss joined 23:20 tinyblak joined, tinyblak left
andreoss m: D: for 1..10 -> \D { for 1..10 -> \E { say D,E and last D if E == 1}} 23:21
camelia rakudo-moar 239d74: OUTPUT«11␤21␤31␤41␤51␤61␤71␤81␤91␤101␤»
andreoss it'd be a nice feature for sigiless variables to be also a loop label 23:22
colomon timotimo: I will try to find time to look at JSON::Fast this week, since faster p6 JSON would be a big win for my $work. :) 23:26
23:29 rararara_ left 23:31 fhelmberger_ joined 23:32 dj_goku_ joined 23:34 gargamel_ joined, mr-fooba_ joined, pmichaud_ joined, perlpilot joined, sivoais_ joined 23:35 huf_ joined, torbjorn_ joined, zoosha_ joined, haroldwu_ joined, atta_ joined 23:38 charsbar__ joined, mr-foobar left, sivoais left, vytas left, smi1er_ left, gargamel left, vytas joined, telex left, BenGoldberg joined 23:39 fhelmberger left, ShimmerFairy left, torbjorn left 23:40 ShimmerFairy joined, telex joined 23:42 smi1er joined 23:46 aborazmeh joined, aborazmeh left, aborazmeh joined 23:49 jobine joined 23:50 jobine left
timotimo i'm glad 23:52
23:53 tinyblak joined 23:56 Akagi201 left, Akagi201 joined, Akagi201 left, Akagi201 joined 23:57 Akagi201 left
timotimo it should be as easy as s/use JSON::Tiny/use JSON::Fast/ 23:59

Powered by Cro and the Raku® Programming Language.

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

Thank you!