»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!
Set by masak on 12 May 2015.
n0tjack g'night all, and thanks again 00:00
00:00 n0tjack left 00:01 llfourn left
softmoth m: my $x = 3; (&sin, &cos X $x :with({&^a($^b)})).perl.say; # I thought there was something like this :with adverb for Z and X? 00:05
camelia rakudo-moar f3b535: OUTPUT«5===SORRY!5=== Error while compiling /tmp/hnbnlZB1PK␤You can't adverb $x␤at /tmp/hnbnlZB1PK:1␤------> 3= 3; (&sin, &cos X $x :with({&^a($^b)})7⏏5).perl.say; # I thought there was someth␤»
00:07 leont left
softmoth m: cross((&sin, &cos), 3, :with({&^a($^b)})) 00:11
camelia rakudo-moar f3b535: OUTPUT«WARNINGS:␤Useless use of "cross((&sin, &cos), 3, :with({&^a($^b)}))" in expression "cross((&sin, &cos), 3, :with({&^a($^b)}))" in sink context (line 1)␤»
softmoth m: cross((&sin, &cos), 3, :with({&^a($^b)})).perl.say 00:12
camelia rakudo-moar f3b535: OUTPUT«(0.141120008059867e0, -0.989992496600445e0).Seq␤»
00:12 NoRefill left 00:14 hohoho joined
[Coke] m: .say for .reals 00:14
camelia rakudo-moar f3b535: OUTPUT«Method 'reals' not found for invocant of class 'Any'␤ in block <unit> at /tmp/IaG_qS1uZH:1␤␤»
[Coke] m: given 42 { .say for .reals } 00:15
camelia rakudo-moar f3b535: OUTPUT«Method 'reals' not found for invocant of class 'Int'␤ in block <unit> at /tmp/kwojiobzrB:1␤␤»
00:16 laouji joined
softmoth sub infix:<-$->(&a, $b) { &a($b) }; (&sin, &cos X[-$-] 3).say.perl 00:17
m: sub infix:<-$->(&a, $b) { &a($b) }; (&sin, &cos X[-$-] 3).say
camelia rakudo-moar f3b535: OUTPUT«(0.141120008059867 -0.989992496600445)␤»
00:33 spider-mario left
kid51 This bug ticket, rt.perl.org/Ticket/Display.html?id=126349, lacks a meaningful Subject. It just says '[BUG]'. Can someone suggest a better Subject for it? If so, I or someone with rt privileges can add that Subject. 00:35
00:36 tokuhirom joined, telex left 00:38 telex joined 00:39 laouji left, CurtisOvidPoe left 00:40 sufrostico left, tokuhirom left 00:41 softmoth left 00:43 laouji joined 00:45 CurtisOvidPoe joined 00:47 Kogurr joined
BenGoldberg m: sub infix:<of> { &^x($^y) }; say &cos of 3; say &sin, &cos, &tan Xof 3; 00:48
camelia rakudo-moar f3b535: OUTPUT«-0.989992496600445␤(0.141120008059867 -0.989992496600445 -0.142546543074278)␤»
00:52 hummeleBop left 00:54 n0tjack joined
n0tjack last annoying newb question of the day: is there a pointfree way of expressing map {'0123456789'.index($_)}, "string".comb? ? 00:55
it just feels like overkill to introduce a new block context just to invoke a single function with a single unbound param
00:55 ofzyh_ joined 00:56 sQuEE left
[Coke] kid51: perhaps you can ask the person who opened the ticket. 00:57
00:58 sQuEE joined, adhoc joined
gfldex kid51: "substitution of Array of Str fails to provide backreferences" and i can understand why he left the subject out :) 00:58
m: my @strs = ("11 22 33"); for @strs { $_ ~~ s/(\S+)\s(\S+)\s(\S+)/$0-$1-$2/; }; say @strs;
camelia rakudo-moar f3b535: OUTPUT«Use of Nil in string context in code at /tmp/pq2S60RUgD:1␤Use of Nil in string context in code at /tmp/pq2S60RUgD:1␤Use of Nil in string context in code at /tmp/pq2S60RUgD:1␤[--]␤»
[Coke] updating..
00:58 ofzyh left
gfldex m: my @strs = ("11 22 33"); for @strs { $_ ~~ s/\s(\S+)\s(\S+)/$0-$1-$2/; }; say @strs; 00:59
camelia rakudo-moar f3b535: OUTPUT«Use of Nil in string context in code at /tmp/NG8WX_iNEp:1␤Use of Nil in string context in code at /tmp/NG8WX_iNEp:1␤Use of Nil in string context in code at /tmp/NG8WX_iNEp:1␤[11--]␤»
gfldex m: my @strs = ("11 22 33"); for @strs { say $_ ~~ s/\s(\S+)\s(\S+)/$0-$1-$2/; }; say @strs;
camelia rakudo-moar f3b535: OUTPUT«Use of Nil in string context in code at /tmp/MoBfYeEcUI:1␤Use of Nil in string context in code at /tmp/MoBfYeEcUI:1␤Use of Nil in string context in code at /tmp/MoBfYeEcUI:1␤｢ 22 33｣␤ 0 => ｢22｣␤ 1 => ｢33｣␤[11--]␤»
BenGoldberg j: my @strs = ("11 22 33"); for @strs { $_ ~~ s/\s(\S+)\s(\S+)/$0-$1-$2/; }; say @strs;
ShimmerFairy gfldex: if that's what it can be called, I could swear there's an existing ticket about s/// in a loop :)
camelia rakudo-jvm f3b535: OUTPUT«Use of Nil in string context in code at /tmp/UtguVw6Bg0:1␤Use of Nil in string context in code at /tmp/UtguVw6Bg0:1␤Use of Nil in string context in code at /tmp/UtguVw6Bg0:1␤[11--]␤»
BenGoldberg n: my @strs = ("11 22 33"); for @strs { $_ ~~ s/\s(\S+)\s(\S+)/$0-$1-$2/; }; say @strs;
camelia niecza v24-109-g48a8de3: OUTPUT«sh: mono: command not found␤»
[Coke] updated. 01:00
timotimo n0tjack: time to build something like python's "itemgetter" or "methodcaller"? :P 01:02
gfldex #118705 is dupe of #123005 is dupe of #123005
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=118705
gfldex it's double duped twice! That's to much for me. Good night.
n0tjack timotimo: I feel there's a whatever-star solution lurking, that I just can't see 01:03
maybe with .assuming?
01:04 sufrostico joined
timotimo putting a whatever star into the method list will not cause whatever currying 01:04
01:04 sufrostico left
timotimo m: (ord *)('a') 01:04
camelia rakudo-moar f3b535: OUTPUT«5===SORRY!5=== Error while compiling /tmp/T6RywmNXwk␤Calling ord(Whatever) will never work with declared signature (Cool $s)␤at /tmp/T6RywmNXwk:1␤------> 3(7⏏5ord *)('a')␤»
timotimo right
not in that case, either 01:05
BenGoldberg m {ord $^x}('a') 01:09
m: {ord $^x}('a').say
camelia rakudo-moar f3b535: OUTPUT«97␤»
n0tjack but it can't be order
I mean ord
it's gotta be a => 0, b => 1, etc
timotimo well, you can subtract ord('0') from the values 01:10
n0tjack and ord('a') - ord('0') won't work, because in the general case it won't be ASCII
ha, beat me by a millisecond
timotimo you'll still have to check for out-of-bounds
n0tjack yeah, but I have plans for that 01:11
BenGoldberg m: my %lookup = ('0'..'9') X (0..9); my @whatever = %lookup{ "1234".comb }; @whatever.say; 01:12
camelia rakudo-moar f3b535: OUTPUT«[(Any) (Any) (Any) (Any)]␤»
BenGoldberg m: my %lookup = ('0'..'9') X (0..9); %lookup.say 01:13
camelia rakudo-moar f3b535: OUTPUT«0 0 => (0 1), 0 2 => (0 3), 0 4 => (0 5), 0 6 => (0 7), 0 8 => (0 9), 1 0 => (1 1), 1 2 => (1 3), 1 4 => (1 5), 1 6 => (1 7), 1 8 => (1 9), 2 0 => (2 1), 2 2 => (2 3), 2 4 => (2 5), 2 6 => (2 7), 2 8 => (2 9), 3 0 => (3 1), 3 2 => (3 3), 3 4 => (3 5), 3 6 …»
timotimo it's not getting any better :D
BenGoldberg headscratches
timotimo X is a *bit* much :D
n0tjack X is cartesian, no?
timotimo yeah
BenGoldberg m: my %lookup; %lookup{'0'..'9'} = 0..9; %lookup.say
camelia rakudo-moar f3b535: OUTPUT«0 => 0, 1 => 1, 2 => 2, 3 => 3, 4 => 4, 5 => 5, 6 => 6, 7 => 7, 8 => 8, 9 => 9␤»
BenGoldberg m: my %lookup; %lookup{'0'..'9'} = 0..9; my @whatever = %lookup{ "1234".comb }; @whatever.say; 01:14
camelia rakudo-moar f3b535: OUTPUT«[1 2 3 4]␤»
n0tjack BenGoldberg: Yeah, I did that too, but it forces me to create intermediate vars.
I'm trying to stay as functional, pointfree, and clean as I can
I mean, I can achieve my ends, but I'm trying to make it hard for others to maintain my code ;) 01:15
If you want to see the kind of monsters I'm inflicting on the world: gist.github.com/anonymous/483ece62552981cb47dc 01:16
BenGoldberg m: my %lookup = map { $_ => $_ } 0..9; my @whatever = %lookup{ "1234".comb }; @whatever.say;
camelia rakudo-moar f3b535: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ok5PJnAIZC␤Missing comma after block argument to map␤at /tmp/ok5PJnAIZC:1␤------> 3my %lookup = map { $_ => $_ }7⏏5 0..9; my @whatever = %lookup{ "1234".c␤ expecting any of:␤ infix␤…»
BenGoldberg m: my %lookup = map { $_ => $_ }, 0..9; my @whatever = %lookup{ "1234".comb }; @whatever.say;
camelia rakudo-moar f3b535: OUTPUT«[1 2 3 4]␤»
BenGoldberg m: my @whatever = ((map { $_ => $_ }, 0..9).hash){ "1234".comb }; @whatever.say; 01:17
camelia rakudo-moar f3b535: OUTPUT«[1 2 3 4]␤»
BenGoldberg n0tjack, What about doing: $/.make( :10[$/.comb] ); 01:21
On the one hand, it's not "doing it the hard way", but on the other, it restricts to base ten integers, just like your present grammar. 01:22
n0tjack BenGoldberg: That does work, but I need a more general solution for other situations (also, for this particular grammar, I'm trying to "do the work myself", build up the values from fundamental principles).
check out line 48, for example 01:24
01:34 vendethiel joined 01:36 n0tjack left 01:37 softmoth joined 01:42 tokuhirom joined, softmoth left 01:44 rhr joined 01:46 tokuhirom left 01:49 adu joined 01:59 llfourn joined 02:05 uruwi joined
uruwi Ho! 02:05
awwaiid How do I get a module of mine listed in panda? Do I need to have the master branch be the stable branch? 02:06
uruwi I installed Rakudo* 2015.09, but can't find Panda, which is claimed to be packaged with the distribution, anywhere. 02:07
garu ilmari: follow up on the HTTP::UserAgent issue. According to tokuhirom, current version expects a \r\n, not just \n. Since RFC2616 also says so, I wonder if it should be implemented at all
uruwi Is there any kind of problem?
02:12 dayangkun joined 02:16 vendethiel left 02:17 kid51 left
uruwi Never mind, found it! 02:18
02:19 noganex_ joined
awwaiid Yeah. me too. 02:19
02:22 noganex left 02:24 colomon left 02:25 colomon joined
dalek osystem: 7f59a4f | (Brock Wilcox)++ | META.list:
Add LREP project
 02:26
osystem: a775816 | (Zoffix Znet)++ | META.list:
Merge pull request #77 from awwaiid/patch-1

Add LREP project github.com/awwaiid/p6-lrep/
awwaiid In case anyone else doesn't know and wants to join the fun: to add a module to the ecosystem (ie panda) do a PR against github.com/perl6/ecosystem editing META.list
dalek osystem: 889c8e5 | (Zoffix Znet)++ | README.md:
Suggest adding viewable URLs when adding modules to eco

For ease of viewing during announcements. Suggested by Issue #76		 02:35
Zoffix awwaiid, you'll also need to run `panda update` for it to refresh its modules list, if you want your module to be installable 02:37
(on your box) 02:38
02:40 colomon left
dalek href="https://modules.perl6.org:">modules.perl6.org: e5ad02f | (Zoffix Znet)++ | web/build-project-list.pl:
Ensure sortable table "arrows" pics are copied to output dir
 02:41
02:43 colomon joined
dalek ecs: 8fb3cd9 | (Nova Patch)++ | / (13 files):
various minor casing/spacing/grammar/syntax fixes

  ◦ unicode → Unicode [except unicode pragma]
  ◦ Perl6 → Perl 6
  ◦ Perl5 → Perl 5 [except Perl5ish]
  ◦ a → an [before vowel sounds]
  ◦ c<…> → C<…>
  ◦ $*ARGS → @*ARGS
  ◦ :p5 → :P5
  ◦ Unicode database → Unicode Character Database [unicode.org/glossary/#unicode_chara..._database]
  ◦ unicode normalisation form → Unicode normalization form [unicode.org/glossary/#normalization_form]
also minor punctuation changes on some of the lines modified		 02:45
timotimo .u 02:46
yoleaux Search for a Unicode character by codepoint, name, or raw character
timotimo .u l 6
yoleaux U+1974 TAI LE LETTER TONE-6 [Lo] (ᥴ)
U+16F97 MIAO LETTER TONE-6 [Lm] (𖾗)
timotimo damn, not what i meant
02:49 aborazmeh joined, aborazmeh left, aborazmeh joined
ShimmerFairy timotimo: what did you mean, ooc? 02:50
m: say "◦".uniname
camelia rakudo-moar f3b535: OUTPUT«WHITE BULLET␤»
timotimo m: say "l 6".uniname 02:51
camelia rakudo-moar f3b535: OUTPUT«LATIN SMALL LETTER L␤»
timotimo m: say " ".uniname
camelia rakudo-moar f3b535: OUTPUT«SPACE␤»
timotimo that's not the right space to go between Perl and 6! 02:52
m: say " ".uniname
camelia rakudo-moar f3b535: OUTPUT«NARROW NO-BREAK SPACE␤»
timotimo ^- has to be this!
uruwi Just a question...
ShimmerFairy timotimo: that's the space I use, not NBSP :)
timotimo it's a NNBSP, though
uruwi how do I make perl6 recognize an internally used module?
02:52 adu left
timotimo uruwi: can you be a little more specific? 02:53
ShimmerFairy timotimo: I always saw NBSP in documents (a purple-ish underline on the space in emacs, which AFAIK is just for NBSP. maybe NNBSP is counted too)
uruwi Is it possible to use a module internally?
(within a project?)
Zoffix uruwi, sure 02:54
ShimmerFairy uruwi: yes it is. I've got quite a few projects where I may have a test file at the top level, module files in lib/, and I run the script with perl6 -Ilib test.p6
timotimo you may have to "use lib" for things
or -I, yeah
Zoffix uruwi, you should be able to just `use` it if the paths are right. If they aren't. Just `use lib 'foobar';` or add them with -I
ShimmerFairy uruwi: you could also put use lib "lib/"; in such a test script, as has been mentioned a couple times already :)
Zoffix Answer all the things \o/
Zoffix goes to bed & 02:55
dalek c: a8437e9 | (Tokuhiro Matsuno)++ | doc/Language/5to6-perlvar.pod:
$*PROGRAM_NAME was renamed to $*PROGRAM-NAME.
 02:56
c: d4c4bfc | (Zoffix Znet)++ | doc/Language/5to6-perlvar.pod:
Merge pull request #161 from tokuhirom/program_name

  $*PROGRAM_NAME was renamed to $*PROGRAM-NAME.
uruwi So if I have a module with the source at lib/Terminal/Basic.pm6 then what would I type? 02:57
Zoffix uruwi, use lib 'lib'; or perl6 -Ilib
ShimmerFairy you'd use that as use Terminal::Basic , so long as you've told Perl 6 to look in that lib/ directory. 02:58
uruwi It doesn't work in the REPL, does it? 02:59
ShimmerFairy uruwi: I'm not sure, but you should be able to type 'use lib "lib/"' (replacing that string with the correct relative path to lib/), and then use Terminal::Basic on the next line 03:01
uruwi: you should also be able to start the REPL as perl6 -Ilib instead of the 'use lib' statement
uruwi Thanks ShimmerFairy 03:02
03:04 mr_ron left
uruwi Any different for source files? 03:12
03:18 atweiden joined 03:25 softmoth joined 03:30 softmoth left 03:43 laouji left, tokuhirom joined 03:46 BenGoldberg left
atweiden i'm testing this github.com/atweiden/nightscape/blo...ader.t#L99 03:47
it passes in isolation here: github.com/atweiden/nightscape/blo...der.t#L118
and here: ix.io/lpw
but the original test fails
is this a rakudobug?
03:48 BenGoldberg joined, tokuhirom left
atweiden (running latest nom) 03:51
uruwi Is there any way to get a keypress without waiting for the user to press [Enter]? 04:01
Ideally I'd use the NCurses module but at this time it doesn't work with Windows.
04:17 WizJin joined
atweiden oops, figured it out 04:23
dalek osystem: b42c954 | lestrrat++ | META.list:
Add Apache::LogFormat
osystem: 2c8368d | (Tokuhiro Matsuno)++ | META.list:
Merge pull request #78 from lestrrat/patch-1

Add Apache::LogFormat
04:23 WizJin left 04:28 atweiden left 04:34 WizJin joined 04:39 laouji joined
uruwi another question: where does perl6 search for dlls on Windows? 04:57
moritz in PATH I believe
05:02 FROGGS[mobile] joined
FROGGS[mobile] dlls are searched in CWD too I believe 05:03
uruwi I'm trying to get the NCurses module to work on Windows. 05:07
FROGGS[mobile] cool
uruwi What does "No appropriate parametric role variant available for 'Native'" mean?
I tried to search for it, but I got confused
FROGGS[mobile] hmmm 05:09
uruwi Does that mean that there's no function found with the appropriate signature in the dll? 05:10
FROGGS[mobile] m: use NativeCall; sub foo is native(<foo bar>) { *}
camelia rakudo-moar f3b535: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Ji4wHTAxFh␤No appropriate parametric role variant available for 'Native'␤at /tmp/Ji4wHTAxFh:1␤»
FROGGS[mobile] ha!
05:10 khw left
FROGGS[mobile] uruwi: pass a string to the is native trait, not a list or something else 05:11
uruwi Ah, it was because $NCURSES_LOC didn't exist. 05:12
05:14 softmoth joined
uruwi Perl still can't locate the dll though 05:14
FROGGS[mobile] ohh
uruwi even if I add the directory to the WINDOWS %Path% variable
FROGGS[mobile] yeah, traits run at compile time, which can be confusing
do you add it to PATH before starting your script? 05:16
uruwi Yes 05:17
Not every time, though
05:18 softmoth left
FROGGS[mobile] setting it in the script it too late though 05:18
but you can also provide an absolut path to is native, without setting Path at all I think 05:20
05:22 sQuEE is now known as sQuEE`
uruwi Still no luck. 05:22
Does it matter if I use a version from Cygwin? 05:24
05:25 Jaglor left
uruwi Tried to use a Windows binary and still doesn't work. 05:30
oh, wrong file name 05:31
unfortunately Murphy's law strikes back "Error opening terminal: xterm." 05:32
This much work just to be able to read keypresses without pressing Enter. 05:35
And even less luck on cmd 05:36
On a break from all this madness, I've abandoned this project long long ago github.com/bluebear94/sun1p1 05:38
05:45 tokuhirom joined, bjz joined 05:47 jkva joined 05:49 tokuhirom left, BenGoldberg left
uruwi It's a problem with initscr() 05:50
05:52 rindolf joined
uruwi It seems I'll have to link a Cygwin-aware DLL, but I can't find libncursesw6.dll anywhere 05:53
05:54 bjz left 05:55 skids left 06:01 kjones_ joined 06:02 kjones_ left 06:08 bjz joined, kmel joined
kmel m: say 'Hello' 06:10
camelia rakudo-moar f3b535: OUTPUT«Hello␤»
uruwi say 99 ** 99 06:12
m: say 99 ** 88 06:14
camelia rakudo-moar f3b535: OUTPUT«41294967113388846862799854538691617968752357209343777782657514166985700076424987472478122770530747238801087245466136620792550517990555186110294214867126284223076111559302271201␤»
uruwi m: say 99 ** 99
camelia rakudo-moar f3b535: OUTPUT«369729637649726772657187905628805440595668764281741102430259972423552570455277523421410650010128232727940978889548326540119429996769494359451621570193644014418071060667659301384999779999159200499899␤»
TEttinger m: say 99 ** 99 ** 99 06:15
camelia rakudo-moar f3b535: OUTPUT«0␤»
TEttinger m: say 99 ** 99 * 99 06:16
camelia rakudo-moar f3b535: OUTPUT«36603234127322950493061602657251738618971207663892369140595737269931704475072474818719654351002695040066156910065284327471823569680179941585710535449170757427389035006098270837114978219916760849490001␤»
TEttinger m: say 99 * 99 ** 99
camelia rakudo-moar f3b535: OUTPUT«36603234127322950493061602657251738618971207663892369140595737269931704475072474818719654351002695040066156910065284327471823569680179941585710535449170757427389035006098270837114978219916760849490001␤»
TEttinger interesting
so ** really does have the right priority for the exponentiation operator
06:18 FROGGS[mobile]2 joined
ShimmerFairy .tell jnthn just taking a quick glance at NQP's regex stuff, I notice case-insensitive stuff uses uc and lc variants. Now that we have nqp::fc, would it be beneficial to have regexes use fc for :i and :ii ? Or do those modifiers explicitly require just uc/lc checking in spec somewhere (nothing suggesting such in S05)? 06:18
yoleaux ShimmerFairy: I'll pass your message to jnthn.
06:20 bjz_ joined, [Sno] left 06:21 bjz left, FROGGS[mobile] left 06:24 kmel left
Hotkeys I like how sequence generation looks when multiplying 06:25
m: my @test = 1, 2, 3, * * * * * ... *; say @test[^10];
camelia rakudo-moar f3b535: OUTPUT«(1 2 3 6 36 648 139968 3265173504 296148833645101056 135345882205792807436868315512832)␤»
uruwi So that was the product of the 3 previous terms right? 06:26
Hotkeys yeah
uruwi m: my @test = 1, 1, 2, * * * * * ... *; say @test[^10];
camelia rakudo-moar f3b535: OUTPUT«(1 1 2 2 4 16 128 8192 16777216 17592186044416)␤»
Hotkeys just a line of asterisks
ShimmerFairy Yeah, I think it's delightfully cryptic. I can't help but think it could also serve as a good tool for teaching the difference between "term" and "operator", at least for parser-writing things :) 06:27
Hotkeys beauty
uruwi m: my @test = 1, 2, 3 * ** * ** * ... *; say @test[^10];
camelia ( no output )
06:27 firstdayonthejob joined
uruwi m: my @test = 1, 2, 3 * ** * ** * ... *; say @test[^5]; 06:27
camelia ( no output )
uruwi m: my @test = 1, 2, 3, * ** * ** * ... *; say @test[^10];
camelia rakudo-moar f3b535: OUTPUT«(1 2 3 1 8 3 1 512 3 1)␤»
Hotkeys that's neat
ShimmerFairy m: my @test = 1, 1, 2, {$^a * $^b * $^c} ... *; say @test[^10]; # the fun-ruining clearer version of that line of asterisks, fwiw 06:28
camelia rakudo-moar f3b535: OUTPUT«(1 1 2 2 4 16 128 8192 16777216 17592186044416)␤»
Hotkeys m: my @test = 1, 2, 3, * ** * ** * ... *; say @test[^20];
camelia rakudo-moar f3b535: OUTPUT«(1 2 3 1 8 3 1 512 3 1 134217728 3 1 2417851639229258349412352 3 1 14134776518227074636666380005943348126619871175004951664972849610340958208 3 1 28240139587082174969491088422046278633513539118515775246834019308626938303611984999058739209952299969708978654…»
Hotkeys interesting that it goes to 3 1 every time
uruwi m: my @test = 2, 2, 2, * ** * ** * ... *; say @test[^10];
camelia rakudo-moar f3b535: OUTPUT«(2 2 2 16 200352993040684646497907235156025575044782547556975141926501697371089405955631145308950613088093334810103823434290726318182294938211881266886950636476154702916504187191635158796634721944293092798208430910485599057015931895963952486337236720300291…»
uruwi m: my @test = 2, 3, 4, * ** * ** * ... *; say @test[^10];
camelia rakudo-moar f3b535: OUTPUT«(2 3 4 2417851639229258349412352 1 0 2417851639229258349412352 1 0 2417851639229258349412352)␤»
Hotkeys I never knew perl 5, only started perl 6 a few days ago
this is fun
uruwi I broke Camelia 06:29
Hotkeys m: 3 ** 4 ** 99999999999999999999999999
camelia rakudo-moar f3b535: OUTPUT«WARNINGS:␤Useless use of "**" in expression "3 ** 4 **" in sink context (line 1)␤»
uruwi Still trying to get NCurses module to work with Windows
06:30 kmel joined, n0xff left
uruwi m: my @test = 2, 2, 2, * ** * ** * ... *; say @test[^10].map(&log); 06:30
camelia rakudo-moar f3b535: OUTPUT«Attempt to divide Inf by zero using /␤ in block <unit> at /tmp/v2TkLbcMfT:1␤␤Actually thrown at:␤ in block <unit> at /tmp/v2TkLbcMfT:1␤␤»
uruwi m: (1, 2).map(&log) 06:31
camelia ( no output )
uruwi m: say (1, 2).map(&log)
camelia rakudo-moar f3b535: OUTPUT«(0)␤»
uruwi m: say (1, 2, 3).map(&log)
camelia rakudo-moar f3b535: OUTPUT«(0 1.09861228866811)␤»
uruwi m: say (1, 2, 3, 4, 5).map(&log)
camelia rakudo-moar f3b535: OUTPUT«(0 0.792481250360578 1.6094379124341)␤»
uruwi say (1).map(&log) 06:32
m: say (1).map(&log)
camelia rakudo-moar f3b535: OUTPUT«(0)␤»
uruwi m: my @test = 2, 2, 2, * ** * ** * ... *; say @test[^10].map(&length);
camelia rakudo-moar f3b535: OUTPUT«5===SORRY!5=== Error while compiling /tmp/XBTugJvACT␤Undeclared routine:␤ &length used at line 1. Did you mean '&elems', '&chars', '&graphs', '&codes'?␤␤»
uruwi m: my @test = 2, 2, 2, * ** * ** * ... *; say @test[^10].map(&chars);
camelia rakudo-moar f3b535: OUTPUT«(1 1 1 2 19729 1 1 19729 1 1)␤»
uruwi m: my @test = 1, 3, 5, * * * * * ... *; say @test[^10].map(&chars); 06:33
camelia rakudo-moar f3b535: OUTPUT«(1 1 1 2 3 5 8 15 27 49)␤»
Hotkeys also lazy geometric sequences are nice too
m: my @geo = 1, 2, 4 ...^ * > 10000; say @geo
camelia rakudo-moar f3b535: OUTPUT«[1 2 4 8 16 32 64 128 256 512 1024 2048 4096 8192]␤»
uruwi m: my @test = 1, 3, 5, * * * * * ... *; say @test[^80].map(&chars);
camelia rakudo-moar f3b535: OUTPUT«(timeout)»
uruwi m: my @test = 1, 3, 5, * * * * * ... *; say @test[^40].map(&chars); 06:34
camelia rakudo-moar f3b535: OUTPUT«(timeout)»
TimToady m: my @geo = 1, 2, 4 ... 10000; say @geo
camelia rakudo-moar f3b535: OUTPUT«[1 2 4 8 16 32 64 128 256 512 1024 2048 4096 8192]␤»
uruwi m: my @test = 1, 3, 5, * * * * * ... *; say @test[^20].map(&chars);
camelia rakudo-moar f3b535: OUTPUT«(1 1 1 2 3 5 8 15 27 49 90 164 302 555 1020 1875 3448 6342 11664 21453)␤»
Hotkeys TimToady, oh
TimToady intuited sequences don't have to match exactly any more
Hotkeys neat
TimToady m: my @geo = 1, -2, 4 ... 10000; say @geo 06:35
camelia rakudo-moar f3b535: OUTPUT«[1 -2 4 -8 16 -32 64 -128 256 -512 1024 -2048 4096 -8192]␤»
TimToady even if they change sign
Hotkeys cool
m: my @geo = -1, 2, -4 ... 10000; say @geo
camelia rakudo-moar f3b535: OUTPUT«[-1 2 -4 8 -16 32 -64 128 -256 512 -1024 2048 -4096 8192]␤»
uruwi m: my @what = 1, 6, 8, ... 10000; say @what
camelia rakudo-moar f3b535: OUTPUT«5===SORRY!5=== Error while compiling /tmp/CRSEYRiGhm␤Comma found before apparent series operator; please remove comma (or put parens␤ around the ... listop, or use 'fail' instead of ...)␤at /tmp/CRSEYRiGhm:1␤------> 3my @what = 1, 6, 8,7…»
uruwi m: my @what = 1, 6, 8 ... 10000; say @what 06:36
camelia rakudo-moar f3b535: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,6,8 (or did you really mean '..'?)␤ in block <unit> at /tmp/rF_mweXWcH:1␤␤»
uruwi m: my @what = 1, 4, 9 ... 10000; say @what
camelia rakudo-moar f3b535: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,4,9 (or did you really mean '..'?)␤ in block <unit> at /tmp/7sXfX4HhmV:1␤␤»
Hotkeys so it ignores sign for the upper limit?
uruwi m: my @what = 1, 4, 16 ... 10000; say @what
camelia rakudo-moar f3b535: OUTPUT«[1 4 16 64 256 1024 4096]␤»
TimToady yes, it takes absolute value for that
ShimmerFairy That error message suddenly makes me think of what use something like 1, 2, 3, fail 10000 would have :P 06:37
TimToady that's more to tell people what the ... was interpreted as than a serious suggestion
06:37 FROGGS joined
ShimmerFairy TimToady: I figured it's one of the less common intentions in that context, but it's still a funny expression to consider :) 06:38
06:38 FROGGS[mobile]2 left
Hotkeys so if you wanted to take into consideration the sign, 06:38
m: my @geo = -1, 2, -4, ...^ * > 10000; say @geo
camelia rakudo-moar f3b535: OUTPUT«5===SORRY!5=== Error while compiling /tmp/eLM7NZ5w_8␤Comma found before apparent series operator; please remove comma (or put parens␤ around the ... listop, or use 'fail' instead of ...)␤at /tmp/eLM7NZ5w_8:1␤------> 3my @geo = -1, 2, -4,7…»
Hotkeys m: my @geo = -1, 2, -4 ...^ * > 10000; say @geo
camelia rakudo-moar f3b535: OUTPUT«[-1 2 -4 8 -16 32 -64 128 -256 512 -1024 2048 -4096 8192 -16384]␤»
uruwi Anyone managed to build and run the NCurses module on Windows? 06:39
FROGGS uruwi: I've never tried
uruwi wb
FROGGS uruwi: is there a Perl 5 version that works on windows?
maybe we can look what they do
uruwi Well, not ncurses per se, but it has what I want metacpan.org/pod/Games::Roguelike::Console 06:42
Hotkeys m: my @alphabet = 'a', 'b', ... 'z'; say @alphabet 06:43
camelia rakudo-moar f3b535: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tG32An3I0Q␤Comma found before apparent series operator; please remove comma (or put parens␤ around the ... listop, or use 'fail' instead of ...)␤at /tmp/tG32An3I0Q:1␤------> 3my @alphabet = 'a', 'b'…»
Hotkeys m: my @alphabet = 'a', 'b' ... 'z'; say @alphabet
camelia rakudo-moar f3b535: OUTPUT«[a b c d e f g h i j k l m n o p q r s t u v w x y z]␤» 06:44
Hotkeys neat
uruwi m: my @hiragana = 'あ', 'い' ... 'ん'; say @hiragana
FROGGS uruwi: so, port that instead? 06:45
camelia rakudo-moar f3b535: OUTPUT«(timeout)»
Hotkeys lol
uruwi You know what, the framebuffer functionality would be useful too
ShimmerFairy m: say "あ" ~~ /<rangechar>/ 06:46
camelia rakudo-moar f3b535: OUTPUT«Method 'rangechar' not found for invocant of class 'Cursor'␤ in block <unit> at /tmp/1R_5FaG3d_:1␤␤»
FROGGS grr, I can't even copy+paste these properly in my chat client
06:46 tokuhirom joined
ShimmerFairy I could swear that was a rule name... 06:46
uruwi m: my @alphabet = 'α', 'β' ... 'ω'; say @alphabet
camelia rakudo-moar f3b535: OUTPUT«[α β γ δ ε ζ η θ ι κ λ μ ν ξ ο π ρ σ τ υ φ χ ψ ω]␤»
06:47 aborazmeh left
TimToady m: say "あ"..."ん" 06:48
camelia rakudo-moar f3b535: OUTPUT«(あ ぃ い ぅ う ぇ え ぉ お か が き ぎ く ぐ け げ こ ご さ ざ し じ す ず せ ぜ そ ぞ た だ ち ぢ っ つ づ て で と ど な に ぬ ね の は ば ぱ ひ び ぴ ふ ぶ ぷ へ べ ぺ ほ ぼ ぽ ま み む め…»
06:49 [Sno] joined 06:51 tokuhirom left 06:52 ecocode joined
uruwi FROGGS I probably should have used a frame buffer from the start 06:52
06:53 bjz_ left
uruwi Anyway, it's way too late at night for me to port this today. 06:54
I'll probably begin tomorrow.
FROGGS ++uruwi :o)
06:55 bjz joined
uruwi Yikes, what's been done for Win32... 06:55
kmel m: say 'ا'...'ي' 06:57
camelia rakudo-moar f3b535: OUTPUT«(ا ب ة ت ث ج ح خ د ذ ر ز س ش ص ض ط ظ ع غ ػ ؼ ؽ ؾ ؿ ـ ف ق ك ل م ن ه و ى ي)␤»
uruwi m: say [+] 'ا'...'ي' 06:58
camelia rakudo-moar f3b535: OUTPUT«Cannot convert string to number: base-10 number must begin with valid digits or '.' in '7⏏5ا' (indicated by 7⏏)␤ in block <unit> at /tmp/Ue8ms31u19:1␤␤Actually thrown at:␤ in block <unit> at /tmp/Ue8ms31u19:1␤␤»
uruwi m: say [~] 'ا'...'ي'
camelia rakudo-moar f3b535: OUTPUT«ابةتثجحخدذرزسشصضطظعغػؼؽؾؿـفقكلمنهوىي␤»
kmel i see additional letters is this normal? 06:59
uruwi say chr(0) .. chr(65535)
m: say chr(0) .. chr(65535)
camelia rakudo-moar f3b535: OUTPUT«"\x[0]".."\x{FFFF}"␤»
uruwi m: say chr(0) .. chr(255)
camelia rakudo-moar f3b535: OUTPUT«"\x[0]".."ÿ"␤»
uruwi say chr(7)
m: say chr(7) 07:00
camelia rakudo-moar f3b535: OUTPUT«␤»
uruwi m: for 0 .. ^100 {say chr(7);}
camelia rakudo-moar f3b535: OUTPUT«Range objects are not valid endpoints for Ranges␤ in block <unit> at /tmp/AQvN6cHeEK:1␤␤»
uruwi m: for 0 .. 100 {say chr(7);}
camelia rakudo-moar f3b535: OUTPUT«␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤…»
uruwi m: for 0 .. 100 {say "\a";}
camelia rakudo-moar f3b535: OUTPUT«␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤…»
Hotkeys oh
I wondered where the bels were coming from
moritz .u ␤ 07:01
yoleaux U+2424 SYMBOL FOR NEWLINE [So] (␤)
moritz not a BEL
Hotkeys oh 07:02
07:02 softmoth joined
Hotkeys well it's making me bel 07:02
07:06 softmoth left 07:10 CIAvash joined 07:19 laouji left
ShimmerFairy Hotkeys: one reason I disabled my system beep :) (the main reason being that, using headphones as my speakers, there's never a time where I could hear system noises that won't also interrupt me listening to something) 07:20
moritz: I'm sure it's the 0x07 being printed between those NL symbols, not the symbols themselves
m: say "\a".uniname 07:21
camelia rakudo-moar f3b535: OUTPUT«BELL␤»
moritz ShimmerFairy: ah
07:22 laouji joined, abraxxa joined 07:24 RabidGravy joined, firstdayonthejob left
Hotkeys night all 07:31
07:31 Hotkeys left 07:34 leont joined 07:35 FROGGS left
masak good morning, #perl6 07:35
07:38 ely-se joined 07:41 adhoc left, cognominal left 07:43 adhoc joined, leont left 07:46 RabidGravy left 07:48 RabidGravy joined
[ptc] o/ masak 07:50
07:50 eliasr joined 07:54 WizJin left 07:55 xinming joined 07:57 ][Sno][ joined 08:01 [Sno] left, xinming left, zakharyas joined 08:02 FROGGS joined 08:08 WizJin joined 08:12 darutoko joined
kmel m: 'good morning'.say 08:13
camelia rakudo-moar f3b535: OUTPUT«good morning␤»
08:15 labster joined 08:21 Kogurr left
nine FROGGS: Perl 5 for Windows is strawberryperl.com/ 08:25
FROGGS nine: hi, I know strawberry... why are you telling about this? 08:26
08:26 jkva_ joined
FROGGS I'm a windows + strawberry + activeperl user after all :o) 08:26
Timbus he probably misinterpreted this: <FROGGS> uruwi: is there a Perl 5 version that works on windows? 08:27
08:28 ely-se left
FROGGS ohh 08:28
nine: this was about an ncurses lib for windows
nine FROGGS: < FROGGS> uruwi: is there a Perl 5 version that works on windows?
ok :)
FROGGS :o)
08:28 jkva left 08:29 Hor|zon joined, virtualsue joined
nine Probably been skimming the backlog too fast. It's reeeeeally hard to keep up when one is afk for an evening. 08:29
FROGGS aye 08:31
08:32 ely-se joined 08:34 adhoc left, dakkar joined 08:36 sivoais joined 08:42 adhoc joined 08:45 sivoais left 08:46 sivoais joined 08:47 tokuhirom joined 08:50 luiz_lha joined, softmoth joined, luiz_lha is now known as Guest13051 08:51 davido_ joined, Guest13051 is now known as luizl_lha, luizl_lha is now known as luiz_lha 08:52 tokuhirom left 08:54 sivoais left 08:55 softmoth left 08:57 sivoais joined
Ulti how do you set the max threads for the thread pool? 08:59
(and does that include GC threads or are they lower level) 09:00
09:00 FireFly left
moritz I don't think the GC runs in a separate thread 09:00
Ulti oh really, I thought it runs in parallel itself when it gets called 09:01
moritz maybe I'm wrong though 09:02
09:02 davido_ left
moritz m: say $*SCHEDULER 09:02
camelia rakudo-moar f3b535: OUTPUT«ThreadPoolScheduler.new(initial_threads => 0, max_threads => 16, uncaught_handler => Callable)␤»
moritz Ulti: you can assign a new ThreadPoolScheduler to $*SCHEDULER, and give it a max_threads 09:03
Ulti okedoke thanks
though sort of wonder what happens if there are already threads when you do that o___O 09:04
09:04 jsnathan joined, FireFly joined, sivoais left 09:06 virtualsue left, rurban joined
kmel what is the correct way of referencing perl 6: Perl 6 or Perl6 or perl 6 or perl6 ? 09:06
09:06 rarara joined
moritz kmel: Perl 6, preferably with a non-breaking space between Perl and 6 :-) 09:06
09:07 sivoais joined
kmel thanks moritz 09:07
09:07 ][Sno][ is now known as [Sno]
Ulti perl6 would be the actual compiler you call specifically rather than the language 09:08
09:10 ely-se left 09:12 chenryn_ joined
kmel So Perl 6 is the language specification 09:12
and perl6 is the compiler
09:12 diana_olhovik_ joined
ShimmerFairy I personally don't think it matters how you write it ("the actual compiler" depends on the implementation, happens to be "perl6" for rakudo). I also don't believe NBSP is anything work worrying over :) 09:12
masak kmel: yes. well, `perl6` is the name of the executable. 09:13
kmel: people refer to the compiler+runtime as "Rakudo"
09:13 brrt joined
chenryn_ how to use varibles in regex/token/rule? 09:13
masak chenryn_: just use them
m: my $s = "oo"; say so "foo" ~~ / f $s / 09:14
camelia rakudo-moar f3b535: OUTPUT«True␤»
kmel thanks
09:14 ely-se joined
masak m: my $s = /oo/; say so "foo" ~~ / f $s / 09:14
camelia rakudo-moar f3b535: OUTPUT«True␤»
masak m: my $s = /o+/; say so "foo" ~~ / f $s /
camelia rakudo-moar f3b535: OUTPUT«True␤»
kmel m: say 'thanks everyone :)'
camelia rakudo-moar f3b535: OUTPUT«thanks everyone :)␤»
09:15 sivoais left
RabidGravy Ulti: also for convenience most of the methods that start new threads such as Promise.start take an optional scheduler argument 09:16
09:16 espadrine_ joined 09:17 sivoais joined, TEttinger left
kmel say 'hello world'; and 'hello world'.say yield the same result. Is it a valid example of Tim Toady? If I was to explain what Tim Toady is. 09:18
Ulti RabidGravy does .race() ?
RabidGravy I somewhat doubt it 09:19
baest kmel: I would say so 09:20
brrt good * #perl6, too
masak kmel: sure
kmel thanks baest and masak 09:21
masak kmel: another example: `for @a -> $elem { ... }` vs `@a.map: -> $elem { ... }`
09:21 virtualsue joined
moritz constructing a hash from keys values: my %h; %h{@keys} = @values; or my %h = @keys Z=> @values; 09:22
masak kmel: and you probably want to write it as TIMTOWTDI (There Is More Than One Way To Do It)
llfourn m: say &say; say Str.^find_method('say').package;
camelia rakudo-moar f3b535: OUTPUT«sub say (| is raw) { #`(Sub|60192304) ... }␤(Mu)␤»
llfourn though they are not exactly the same thing under the hood :) 09:23
kmel masak: ok
rarara time perl6 -e 'say "Ciao"' : user 0m0.156s ; time perl6 <<< 'say "Ciao"' : user 0m2.070s. Wow
masak kmel: though, of course, there's more than one way to write it :P
chenryn_ p6: say "key=value" ~~ /$<key>=\N+ '=' $<value>=\N+/; 09:24
camelia rakudo-moar f3b535: OUTPUT«｢key=value｣␤ key => ｢key｣␤ value => ｢value｣␤»
chenryn_ p6: my rule field { $<key>=\N+ '=' $<value>=\N+ }; say "key=value" ~~ &field;
camelia rakudo-moar f3b535: OUTPUT«Nil␤»
chenryn_ can't write named capture directly in rule?
kmel masak :)
09:25 diana_olhovik_ left, sivoais left
DrForr_ rarara: Oddly I get the exact opposite. sys 0m0.053s without <<<, 0m0.008s with. 09:25
But I might be more up-to-date, maybe rebuild your rakudo?
rarara I will
09:27 sivoais joined 09:30 mattprelude joined
ShimmerFairy m: my regex field { $<key>=\N+ '=' $<value>=\N+ }; say "key=value" ~~ &field; 09:30
camelia rakudo-moar f3b535: OUTPUT«｢key=value｣␤ key => ｢key｣␤ value => ｢value｣␤»
RabidGravy It would be amusing to find the thread from clpm from about 1998 wherein someone asked "how do I find the length of a string" and it went on for a week with people posting ever more ludicruous ways of doing it
ShimmerFairy chenryn_: 'rule' considers whitespace significant; 'regex' is equivalent to the ~~ / ... / example you used
RabidGravy and then redo it as Perl 6
ShimmerFairy (there's also 'token' , which is like 'regex' but disables backtracking behavior)
09:31 tadzik joined
llfourn just learnt something new about token 09:31
09:32 M-tadzik1 left
chenryn_ so, can't ~~ a rule but only regex? 09:33
ShimmerFairy chenryn_: that's not it, the problem is that you had whitespace in that rule, but there's no whitespace in the text you were matching 09:34
09:35 virtualsue left, sivoais left 09:37 tokuhirom joined 09:38 sivoais joined
ShimmerFairy chenryn_: oh, forgot to mention that 'rule' also doesn't do any backtracking (just like 'token'), so the first \N+ would've captured that entire string and left nothing for the rest of it to match. 09:38
09:38 jsnathan left 09:39 ely-se left 09:40 vytas left, margeas joined 09:41 tokuhirom left
chenryn_ p6: my $field_split = ' ';my $value_split = '=';my regex field { $<key>=\N+ $value_split $<value>=\N+ };my regex kv { <field>+ % $field_split };say 'key=value k2=value2' ~~ &kv; 09:43
camelia rakudo-moar f3b535: OUTPUT«｢key=value k2=value2｣␤ field => ｢key=value k2=value2｣␤ key => ｢key=value k2｣␤ value => ｢value2｣␤»
chenryn_ oh, yes...
09:47 sivoais left 09:48 sivoais joined, sivoais left, sivoais joined, spider-mario joined 09:52 jsnathan joined
rarara Is it possible to tell rakudobrew to download all the packages and then to rakudobrew to build withoud downloading? 09:54
github.com[0: 192.30.252.128]: errno=Network is unreachable
this is what i get on the cluster
09:57 sivoais left 09:58 sivoais joined
rarara because, btw, a simple rsync -av .rakudobrew me-there@where does also not work 09:58
carlin colomon: any idea why Crypt::Bcrypt isn't listed on smoke.perl6.org ? 10:01
jnthn afternoon, #perl6 10:02
yoleaux 06:18Z <ShimmerFairy> jnthn: just taking a quick glance at NQP's regex stuff, I notice case-insensitive stuff uses uc and lc variants. Now that we have nqp::fc, would it be beneficial to have regexes use fc for :i and :ii ? Or do those modifiers explicitly require just uc/lc checking in spec somewhere (nothing suggesting such in S05)?
jnthn .tell ShimmerFairy The regex stuff doesn't use fc 'cus we only just implemented fc. :) We should use it instead of lc. 10:05
yoleaux jnthn: I'll pass your message to ShimmerFairy.
ShimmerFairy ding! 10:06
yoleaux 10:05Z <jnthn> ShimmerFairy: The regex stuff doesn't use fc 'cus we only just implemented fc. :) We should use it instead of lc.
ShimmerFairy jnthn: I of course figured it's because fc is new, I was just wondering if that was something to be done, or if there was perhaps a specific reason to not change it over :)
10:07 sivoais left
jnthn m: grammar G { token TOP { <digit> <alpha> } }; class A { method TOP((:$digit, :$alpha)) { say $digit; say $alpha; } }; G.parse('1a', actions => A) 10:07
camelia rakudo-moar f3b535: OUTPUT«｢1｣␤｢a｣␤»
jnthn .tell n0tjack about your unpacking $/ question in the backlog that I don't think anyone answered: grammar G { token TOP { <digit> <alpha> } }; class A { method TOP((:$digit, :$alpha)) { say $digit; say $alpha; } }; G.parse('1a', actions => A) 10:08
yoleaux jnthn: I'll pass your message to n0tjack.
10:08 sivoais joined 10:09 vytas joined 10:11 risou_awy joined, risou_awy is now known as risou 10:14 hohoho left, hohoho joined 10:18 sivoais left 10:19 sivoais joined 10:23 ely-se joined
brrt afternoon jnthn 10:23
jnthn o/ brrt 10:25
10:28 g5 joined, sivoais left, g5 left 10:29 cygx joined
cygx o/ 10:29
10:29 sivoais joined
cygx any thoughts on adding a :$match flag to Str.split and have it retunr strings by default in case of :all? 10:30
I found it rather surprising to get back a list of mixed strings and Match objects... 10:31
10:31 vytas left 10:36 vytas joined
lizmat good *, #perl6! 10:37
dalek kudo/nom: 5271fc7 | lizmat++ | src/core/ (4 files):
Use private methods for better optimizations

Well, at some point in the future.
10:38 sivoais left, ely-se left, softmoth joined
jnthn lizmat: I think we already do something with 'em :) 10:39
lizmat well, I tested it yesterday, didn't see a difference in performance though :-(
10:39 sivoais joined
lizmat class A { method a() {}; method b() { self.a } }; for ^1000000 { A.b } # versus: 10:40
class A { method !a() {}; method b() { self!a } }; for ^1000000 { A.b }
jnthn It'll be hard to see 10:41
I suspect in those cases everything gets inlined and it's a wash
And non-privates get dynamic optimization
lizmat ok, in any case, I thought it would be better that way :-) 10:42
10:42 tokuhirom joined
lizmat starts backlogging 10:42
jnthn So while the static code generation is perhaps a bit better for privates (a lookup by name less), once spesh gets its hands on the code it'll all be the same. :)
dalek kudo/nom: 9db48e9 | cygx++ | src/core/Str.pm:
Add missing Slip.new(...) to Str.split(:all)
 10:43
kudo/nom: b16110d | lizmat++ | src/core/Str.pm:
Merge pull request #551 from cygx/patch-3

Add missing Slip.new(...) to Str.split(:all)
10:43 softmoth left
cygx lizmat++ 10:44
10:47 tokuhirom left 10:48 sivoais left
|Tux| test 50000 36.887 36.778 10:49
test-t 50000 37.560 37.452
10:49 sivoais joined 10:52 brrt left
nine Regardless of performance considerations, making these methods private sounds like a good idea. 10:56
jnthn Yes, agree
10:57 diana_olhovik joined 10:58 sivoais left 10:59 sivoais joined 11:01 diana_olhovik left 11:03 ely-se joined 11:04 cvxcv joined, cvxcv left
|Tux| lizmat, your CSV6 commit causes a syntax error (you do not even need to run the test suite) 11:11
===SORRY!=== Error while compiling /pro/3gl/CPAN/Text-CSV6/lib/Text/CSV.pm
Missing block
at /pro/3gl/CPAN/Text-CSV6/lib/Text/CSV.pm:1680
lizmat grrr :-)
11:15 hohoho left
dalek osystem: c7f50ff | azawawi++ | META.list:
Add Selenium::WebDriver
 11:15
11:16 larion joined
lizmat m: class A { method a(@a) { @a.grep(self) } }; say A.a(A xx 5) # [Tux] not sure why that fails in CSV 11:16
camelia rakudo-moar 5271fc: OUTPUT«((A) (A) (A) (A) (A))␤»
lizmat [Tux]: better ? 11:18
|Tux| yep 11:19
lizmat I do so many recompiles, and needing to reinstall everything... I hope we get a better fix for that soon 11:20
jnthn too :) 11:21
Hairy problem is hairy
11:25 azawawi joined
ely-se Hairy Potter 11:27
azawawi hi
ely-se hi azawawi!
azawawi is there an http client library in Perl 6 that .post actually works? 11:28
RabidGravy HTTP::UserAgent will happily do a POST 11:30
though this may depend on what you want to actually post, lots of work going in at the moment 11:31
if it doesn't work file an issue ;-)
anyway, getting cabin fever - off out to the pub for a bit ;-) 11:32
11:32 diana_olhovik_ joined, ggoebel2 joined 11:33 hohoho joined, mayuresh joined
mayuresh p6: say 3; 11:33
camelia rakudo-moar b16110: OUTPUT«3␤»
mayuresh p6: say "Hello, world!"; 11:34
camelia rakudo-moar b16110: OUTPUT«Hello, world!␤»
mayuresh ah, the camelia bot is rather cool ...
:)
11:35 mayuresh left 11:36 araujo joined, RabidGravy left 11:37 kmel left, araujo left, eliasr left, araujo joined 11:38 pmurias joined
lizmat m: Str.new(value => q/Hello, world!/ ~ "\n").print # more complicated way 11:39
camelia rakudo-moar b16110: OUTPUT«Hello, world!␤»
|Tux| test 50000 36.255 36.146
test-t 50000 37.664 37.555
lizmat .5 sec less ?
azawawi github.com/sergot/http-useragent/b...rAgent.pm6 # doesnt contain even a post method :) 11:40
pmurias hi 11:42
11:42 diana_olhovik_ left
dalek p: 1764d85 | (Pawel Murias)++ | src/vm/js/nqp-runtime/io.js:
[js] Fix bug in nqp::getenvhash.
 11:43
p: 0945366 | (Pawel Murias)++ | src/vm/js/ (3 files):
[js] Implement nqp::spawn using the nqp-js-io module (For now lives on github as pmurias/nqp-js-io).
p: 7330628 | (Pawel Murias)++ | / (7 files):
[js] Load the setting in the bootstrapped nqp-js.

node nqp-bootstrapped.js t/nqp/01-literals.t passes.
It uses a bunch of hacks that need to be cleaned up.
11:43 dayangkun left
jnthn pmurias++ 11:43
11:44 doublec_ joined
CIAvash azawawi: github.com/sergot/http-useragent/b...equest.pm6 11:44
11:44 kid51 joined 11:45 travis-ci joined
travis-ci NQP build failed. Pawel Murias '[js] Load the setting in the bootstrapped nqp-js. 11:45
travis-ci.org/perl6/nqp/builds/85310473 github.com/perl6/nqp/compare/6ac0c...30628d94a6
11:45 travis-ci left 11:47 doublec_ is now known as doublec
pmurias hmm, nqp-js-io fails to build on travis for some reason :( 11:49
11:51 loren joined
iv looks like it's using a very old version of node 11:52
loren good evening, perl6 11:53
dalek p: 16cd47d | (Pawel Murias)++ | src/vm/js/ (2 files):
[js] Implement nqp::getmessage.
 11:54
p: de79fcf | (Pawel Murias)++ | t/nqp/44-try-catch.t:
Test nqp::getmessage.
11:56 kid51 left 11:57 travis-ci joined
travis-ci NQP build failed. Pawel Murias 'Test nqp::getmessage.' 11:57
travis-ci.org/perl6/nqp/builds/85312361 github.com/perl6/nqp/compare/73306...79fcf514f8
11:57 travis-ci left
nine pmurias++ # still my lonesome hero :) 11:59
12:05 uruwi left, pierrot_ left 12:08 ely-se left 12:11 telex left 12:12 ely-se joined, telex joined, laouji left 12:16 pierrot joined, hohoho left 12:18 ZoffixW left, hohoho joined, espadrine_ is now known as espadrine 12:19 Jaglor joined, Jaglor left 12:21 cognominal joined 12:22 reneeb joined 12:27 softmoth joined, RaTTuS|BIG joined, jkva joined 12:28 margeas left 12:29 CIAvash left 12:30 jkva_ left 12:31 softmoth left
azawawi sergot: ping 12:39
12:39 sufrostico joined, sufrostico left 12:42 kmel joined 12:44 tokuhirom joined
moritz ok, heretic question: python has inverted stack traces, and the error message at the bottom 12:44
what do you think of that? something to adopt? or rather avoid?
arnsholt Confuses me endlessly, when it's been a while since I programmed in Python =) 12:45
But potentially a good idea, yeah
[Coke] -1 from me; I like having the error be right next to the last output
12:46 llfourn left
arnsholt It's actually very ergonomic when the stack traces start getting long 12:46
lizmat simple to add with an env variable. fwiw
well, the non --ll-exception case
12:48 tokuhirom left 12:50 ZoffixW joined
azawawi sigh, hit another http redirection bug with HTTP::UserAgent :( 12:50
so far POST + HTTP redirection is not working. Nearly tried everything, LWP::Simple, HTTP::Client (hangs), HTTP::UserAgent(hangs) 12:51
ZoffixW Hey. Does a "clock sequence" have meaning in Perl 6? What is it and how do I obtain it? It's mentioned in 4.1.5 of RFC 4122, but I'm not 100% sure what it is (www.rfc-base.org/txt/rfc-4122.txt)
azawawi, how did HTTP::Tinyish fair?
12:51 ely-se left
azawawi let me check :) 12:52
12:54 rurban left
[Coke] .see curtisovidpoe 12:55
.seen curtisovidpoe
yoleaux I saw CurtisOvidPoe 13 Oct 2015 13:44Z in #perl6: <CurtisOvidPoe> Heh.
CurtisOvidPoe looks for CurtisOvidPoe
[Coke] FOUND HIM!
dalek p: cc2b977 | (Pawel Murias)++ | / (2 files):
[js] Implement goal and dba rxtypes.

Pass rx_goal.		 12:56
p: e5c012a | (Pawel Murias)++ | / (3 files):
[js] Add a make js-test-bootstrapped target.
12:59 eone joined, travis-ci joined
travis-ci NQP build failed. Pawel Murias '[js] Add a make js-test-bootstrapped target.' 12:59
travis-ci.org/perl6/nqp/builds/85321737 github.com/perl6/nqp/compare/de79f...c012aeb1d4
12:59 travis-ci left 13:01 ely-se joined
[Coke] shakes his fist at travis. 13:03
azawawi ZoffixW: i tried it with HTTP_TINYISH_DEBUG=1, it uses curl and gets stuck
ZoffixW: This is working perfectly github.com/azawawi/perl6-selenium-...antomjs.pl
ZoffixW: this gets stuck github.com/azawawi/perl6-selenium-...antomjs.p6
hmmm even curl get stuck 13:05
ZoffixW azawawi, what HTTP method does it redirect to?
IIRC a POST should be only redirected to a GET
13:05 brrt joined
azawawi ZoffixW: gist.github.com/azawawi/e30986625e31cae4f2e9 13:06
ZoffixW: it is basically against a `phantomjs --webdriver=5555`
ZoffixW: it is installed via 'sudo apt-get install phantomjs' # 1.9 13:07
LWP::UserAgent from Perl 5 has no problem with it. i think it is doing a workaround. 13:08
ZoffixW I don't have that package. But I'm guessing "* Disables POST, goes with GET" and "* Empty reply from server" are clues. It's expecting the redirect to follow with a POST, not GET. 13:09
Oh, there's actually "> POST /wd/hub/session/185fea20-"... then maybe the other way around, it's expecting a GET, but receiving a POST :P
In other words... ¯\_(ツ)_/¯ 13:10
azawawi well that's true because i added the following github.com/azawawi/perl6-selenium-...mjs.pl#L47 13:12
colomon carlin: Crypt::Bcrypt got on my “skip” list because it was hanging too often. I’ll try reinstating it. 13:13
13:14 llfourn joined
ZoffixW azawawi, aha! :) 13:14
sergot azawawi: pong 13:15
it does not, we need to add it :)
you can build your post request using HTTP::Request
azawawi sergot: i was trying to use HTTP::UserAgent and somehow hit a redirection bug 13:16
sergot: already did :)
sergot and then $useragen->request($your_request)
azawawi++
azawawi the goal here is to run automate testing for websites over phantomjs as phase #1 and then go with firefox, ... 13:17
s/automate/automated/
im basically implementing code.google.com/p/selenium/wiki/Js...reProtocol 13:18
kmel is there a function I can call on a scalar that returns its type (Int, Str ect.)?
cygx m: 42.WHAT
camelia ( no output )
ely-se m: my $x = 42; say $x.WHAT
camelia rakudo-moar b16110: OUTPUT«(Int)␤»
psch m: say Int.^can('WHAT') # it's not quite a function though 13:19
camelia rakudo-moar b16110: OUTPUT«()␤»
sergot azawawi: what is the problem now?
could you show me the code?
psch but i imagine that probably doesn't matter... :)
kmel thanks ely-se cygx psch 13:20
azawawi sergot: LWP::UserAgent is doing some magic with POST redirection. We're not. Same with curl. 13:21
sergot: the Perl 5 version that is
13:21 pmurias left
azawawi sergot: basically phantomjs 1.9 which is now in debian stable responds with an HTTP status 303 to redirect 13:23
13:23 rindolf left
azawawi sergot: 2.0 fixed that 13:23
sergot: let me update the script to use HTTP::UserAgent
13:23 pmurias joined
dalek href="https://modules.perl6.org:">modules.perl6.org: 05c783d | (Zoffix Znet)++ | web/js/main.js:
Coerce sorter plugin to treat all module names as text
 13:27
azawawi sergot: github.com/azawawi/perl6-selenium-...mjs.p6#L50 13:29
dalek href="https://modules.perl6.org:">modules.perl6.org: 31e9990 | (Zoffix Znet)++ | web/ (2 files):
Disable sort on columns that do not make sense

Use sorter plugin's options instead of half-fixing this with CSS		 13:30
13:31 eone left
azawawi starts debugging HTTP::UserAgent 13:33
sergot azawawi: hnn
hmm
What magic actually do you mean? :) 13:35
literal spacebat: there's also perl6-mode, but it's not quite there yet in terms of highlighting
ZoffixW sergot, I'm guessing this: github.com/azawawi/perl6-selenium-...mjs.pl#L47 13:36
sergot oh, ok. 13:40
13:41 skids joined 13:43 sufrostico joined, sufrostico left, bjz left
azawawi sergot: aha 13:46
sergot: headers are not preserved on redirection in HTTP::UserAgent?
13:46 bjz joined
sergot azawawi: let me check 13:46
azawawi sergot: they're not 13:47
sergot azawawi: right 13:48
13:49 bjz_ joined
sergot azawawi: please create an issue for that 13:49
azawawi sergot: im trying to fix it :)
sergot azawawi++ that's nice! thank! :) 13:50
13:50 bjz left
dalek osystem: f30d76f | moznion++ | META.list:
Add Stream::Buffered
 13:50
13:54 rurban joined 13:55 margeas joined 13:57 softmoth joined 14:01 kmel left 14:04 khw joined 14:05 g4 left, ely-se left 14:06 ZoffixW left
[Coke] .tell nine - please log out and log back in to RT. 14:06
yoleaux [Coke]: I'll pass your message to nine.
14:06 hohoho left
[Coke] Also - if folks need help with RT that isn't "make me a perl 6 bug admin", there's an email addy at the bottom of each page, the rt bugadmins seem very helpful. 14:06
dalek kudo/nom: e3ab6cc | lizmat++ | src/core/Rakudo/Internals.pm:
Add WeightedRoll class

It might be of general usage, and was the only way to not have the code live twice: once for Mix and once for MixHash		 14:09
kudo/nom: 254c1a8 | lizmat++ | src/core/ (6 files):
Some more Bag/BagHash/Mix/MixHash reorganization

Basically, backport my experiment of using 2 hashes from last Monday by taking only the good bits I did.
14:09 rindolf joined, cygx_ joined
dalek ast: 08e8c9a | lizmat++ | S32-basics/xxKEY.t:
Fix test because of simplified code

Previously, we had an ASSIGN-KEY for Bag/Mix, now we just let the fact that there is no container, take care of throwing the error.		 14:11
lizmat cycling&
azawawi sergot: going to debug it @home. nearly there :)
sergot: thanks for your help
Zoffix: thanks :)
14:12 cygx left, azawawi left
sergot azawawi: thank you too 14:14
[Coke] RT: 1029: nom: 8; glr: 4; weird: 11; LTA: 87; testneeded: 9; xmas: 75
14:17 reneeb left 14:20 ely-se joined
tadzik oh gods that Ovid's blog post is hilarous 14:23
and a bit sad as well
14:25 sufrostico joined
[Coke] I hire consultants here. That rings pretty true, and I get the ones that have -already gone through- a consultanting company screener before they get to me! 14:25
*consulting
14:27 loren left 14:33 CJKinni joined 14:34 weihan joined 14:35 cognominal left 14:42 muraiki joined
zostay is react/whenever/supply documented somewhere? i'm not even finding them described in S17 14:42
14:43 RabidGravy joined 14:45 edehont joined, tokuhirom joined, ShimmerFairy joined 14:48 cygx_ left, llfourn left
[Coke] jnthn++'s presentations are the best place to find that at the moment. 14:49
14:50 tokuhirom left
[Coke] try this: www.jnthn.net/papers/2015-spw-concurrency.pdf 14:50
jnthn Also there's a gist somewhere that describes it in ratehr more detail that is to be incorporated into S17 at some point.
14:50 ecocode left
[Coke] and a todo for docs to get it in there. 14:50
14:51 edehont left 14:53 brrt left 14:54 llfourn joined 15:01 CJKinni_ joined 15:03 sufrostico left 15:04 CJKinni left 15:05 blackcat_ joined, blackcat_ is now known as loren 15:06 ely-se left 15:07 hohoho joined, zakharyas left 15:12 hohoho left 15:14 spider-mario left
daxim who is responsible for <p6weekly.wordpress.com/2015/10/12/...ling/>? "# even though there are officially no words starting with ß" is a daxim bait if I've ever seen one 15:14
that only needs one counter example to disprove 15:15
"ßo"
(that's sorbian)
jnthn m: "ßo".fc 15:16
camelia ( no output )
jnthn m: say "ßo".fc
camelia rakudo-moar 254c1a: OUTPUT«sso␤»
15:16 ely-se joined
jnthn m: say "ßo".uc 15:16
camelia rakudo-moar 254c1a: OUTPUT«SSO␤»
jnthn m: say "ßo".tc
camelia rakudo-moar 254c1a: OUTPUT«Sso␤»
jnthn :)
jnthn had to look up what sorbian is :) 15:17
daxim no smiley, because that particular german orthographic rule does not apply
gfldex Sso is most likely wrong in that case because ß is a legature of s and z (the so called round z)
daxim it's a letter, not a ligature for at least 120 years 15:18
jnthn Doesn't matter. Perl 6 implements the Unicode SpecialCasing/CaseFolding rules.
If you want anything other than those, that's for module space.
daxim and it combined ſ and s, not s and z
yeah, I' know, it's a bug in unicode 15:19
they cannot fix it because of reasons. it's complicated
jnthn :)
ShimmerFairy daxim: I concur with you, it really should be ß <-> ẞ :)
daxim I have a p5 module that attempts to fix it
jnthn Yeah, I started reading a little on it and was like...OK, I don't even have time for this, I'll just get on with doing what Unicode says.
15:20 s34n left
jnthn They seem to take compatability quite seriously... That's probably a good thing overall. 15:20
ShimmerFairy daxim: AFAIU, the best/only solution would be if they just added two new symbols that implemented the proper casing, and deprecated the existing ones. (Though just changing the properties would be much simpler)
15:20 jkva left
jnthn Well, stability, not compatibility. 15:20
loren my $l = IO::Socket::INET.new(:listen, :localport(xxxx)); my $c = $l.accept; while my $buf = $c.recv(:bin) { $buf.decode('utf-8').say; }; $c.close; 15:21
The simple service will terminate when i press CTRL+C in the client side.
zostay thx [Coke] 15:22
15:22 weihan left
ShimmerFairy jnthn: btw, I do think we should be doing what Unicode says in core (as you've done), even if they let an erroneous application of NFKC slip through :P 15:22
daxim at least now I have a strong motivation to write my first p6 module 15:23
15:23 weihan joined 15:24 uruwi joined
uruwi Back 15:24
15:25 weihan left, weihan joined
uruwi Still going to port Games::Rougelike 15:26
15:27 liztormato joined, sufrostico joined 15:28 liztormato left, liztormato joined, ely-se left
loren '$buf.decode('utf-8')' will throw a exception when $buf is not a string. the document not mention that. 15:29
jnthn loren: It's still appropriate. You can't write robust code like that. What if you get a multi-byte UTF-8 sequence spread over a couple of packets? 15:31
15:31 s_kilk joined
jnthn The documentation for .decode should mention what happens if it can't decode though, so if it doesn't that deserves an doc issue 15:32
loren ok, thks very much, jnthn 15:33
15:33 RabidGravy left, jonathan__ joined
jnthn (At some point we should also support things like inserting a replacement char too) 15:33
loren i'll take more test
15:33 jonathan__ is now known as RabidGravy
mrf jnthn: is there currently any way to decode streamed in bytes without waiting for the entire stream to complete? 15:34
jnthn mrf: Use the string-level rather than byte-level API on sockets.
mrf: I plan to expose a streaming API for it too
15:35 jkva joined
jnthn (Hopefully pre-xmas) 15:35
RabidGravy while we're on the subject of sockets, does the NQP socket api expose timeout for connect?
jnthn RabidGravy: Not afaik 15:36
jnthn will take a good pass over I/O things in November
dalek p: 932548b | (Pawel Murias)++ | src/vm/js/ (2 files):
[js] avoid emitting a needless warning
 15:37
p: b3a4cbf | (Pawel Murias)++ | src/vm/js/nqp-runtime/core.js:
[js] Stringify the key value things we get while iterating on hash to the key.
p: 25008c3 | (Pawel Murias)++ | src/vm/js/nqp-runtime/core.js:
[js] Make nqp::istype work on the result of nqp::hash.
p: 7b61ed4 | (Pawel Murias)++ | t/nqp/89-istype.t:
Test calling nqp::istype on nqp::list() and nqp::hash().
p: a133692 | (Pawel Murias)++ | t/nqp/68-associative-for.t:
Test stringifying the keyvalue pair to the key.
jnthn mrf: In more detail: MoarVM has a nice implementation of these things that can handle not only codepoints hanging over multi-byte sequences, but also graphemes hanging over multi-codepoint sequences, to get NFG stuff right. So I want to provide a Perl 6 level API for that stuff. 15:38
pmurias jnthn: it's funny that I have to port over code from MoarVM to support doing some IO things synchronously on node.js
15:39 FROGGS left
mrf jnthn: that would be nice for Web Streaming stuff 15:40
15:40 n0tjack joined, travis-ci joined
travis-ci NQP build failed. Pawel Murias 'Test stringifying the keyvalue pair to the key.' 15:40
travis-ci.org/perl6/nqp/builds/85354276 github.com/perl6/nqp/compare/e5c01...3369226b28
15:40 travis-ci left
jnthn mrf: Indeed. I'll try to make it happen pre-xmas so everyone'll have an easier time implementing such things. 15:40
mrf jnthn: is the idea that the methods would just return Str rather than Buf? 15:42
15:42 jax joined 15:43 virtualsue joined
dalek p: a187d4c | (Pawel Murias)++ | .travis.yml:
Disable testing the js backend on travis as the node.js version is too old.
 15:43
15:43 liztormato left
15:43 liztormato_ joined
jnthn mrf: There's a char-based API there today; the use case here is more like when you need to start out working at byte level because you didn't work out what encoding the rest of the message has to be decoded in 15:44
mrf: But the API will basically be something you can throw Buf/Blob into, and then ask for a Str with as many chars as it can completely decode/normalize at that point
timotimo oh, neat 15:46
15:46 tokuhirom joined 15:47 travis-ci joined
travis-ci NQP build passed. Pawel Murias 'Disable testing the js backend on travis as the node.js version is too old.' 15:47
travis-ci.org/perl6/nqp/builds/85355579 github.com/perl6/nqp/compare/a1336...87d4c93f69
15:47 travis-ci left
gtodd1 hmm the "naive parsing benchmark" made a great leap forward in speed with GLR ... but seems to have slowed down again 15:48
www.perlmonks.org/?node_id=1098390
or perhaps I am imagining it all
15:48 gtodd1 is now known as gtodd, pierrot left
n0tjack lear 15:49
yoleaux 10:08Z <jnthn> n0tjack: about your unpacking $/ question in the backlog that I don't think anyone answered: grammar G { token TOP { <digit> <alpha> } }; class A { method TOP((:$digit, :$alpha)) { say $digit; say $alpha; } }; G.parse('1a', actions => A)
n0tjack oh, neat
15:49 jkva left
n0tjack goes off to refactor his grammar 15:49
15:51 tokuhirom left, Ven joined 15:57 n0xff joined 15:58 larion left 16:00 virtualsue left, yakudza left 16:07 s_kilk left, Ven left 16:08 hohoho joined 16:10 virtualsue joined, liztormato_ left
timotimo gtodd: what parts have you tried? 16:11
16:11 liztormato joined, liztormato left 16:12 Ven joined, hohoho left, diana_olhovik_ joined
timotimo like, which of these alternatives? 16:13
loren good night, perl6 16:14
also timotimo
timotimo good night loren 16:15
16:15 loren left, virtualsue left
gtodd timotimo: well I am not doing anything very formal .... I have a bunch of different command lines and some time a few weeks back compiled a post GLR perl6 and was shocked at how much it had sped up 16:16
timotimo how did you reach the conclusion that it has slowed down again?
16:16 yakudza joined
timotimo just by compiling a newer rakudo and re-running? 16:16
gtodd timotimo: yeah ... something like perl6 -n -e 'say $/ if m/<<\w**5>>/' logs.txt 16:17
had dropped from 40 to < 10 seconds ... unless I made a mistake
timotimo can i has your logs.txt? 16:18
gtodd recompiling from yestyerday it was back to something in the order of 30 seconds ...
timotimo wow
gtodd timotimo: ok just a sec trying with a freshly built rakudo again
timotimo that'd be a big loss indeed
16:20 Ven left 16:22 ^elyse^ joined
TimToady you might try it with 'say $/ if .simplematch(/<<\w**5>>/)' to see if it's the gather/take overhead or not 16:22
gtodd ok
timotimo: easiest explanation I must have made a mistake seeing the earlier speed up
timotimo oh 16:23
that's actually a bit sad :(
gtodd but
one thing I notice is: 16:24
time perl6 -e 'for lines(:eager) { say ~$/ if m/(<<\w**5>>)/}' 1k-lines.txt 0.67s
TimToady is the -n loop being lazy or batching?
16:24 diana_olhovik_ left, virtualsue joined
timotimo :eager doesn't exist any more, though 16:25
gtodd ah ok
my script needs tweaking then :-)
jnthn gtodd: What about just for lines() { ... } on a modern Rakudo?
tony-o_ m: say $?ENC
camelia rakudo-moar 254c1a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/VjdRC4pxWx␤Variable '$?ENC' is not declared. Did you mean '@?INC'?␤at /tmp/VjdRC4pxWx:1␤------> 3say 7⏏5$?ENC␤»
jnthn TimToady: I'm not sure -n has been looked at since GLR...
tony-o_ r: say $?ENC; 16:26
camelia rakudo-{moar,jvm} 254c1a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Variable '$?ENC' is not declared. Did you mean '@?INC'?␤at /tmp/tmpfile:1␤------> 3say 7⏏5$?ENC;␤»
tony-o_ hm
gtodd timotimo: trying with a file that's bigger doesn't errm "scale" ... perl6 -e 'for lines(:eager) { say ~$/ if m/(<<\w**5>>)/}' 100k-lines.txt 30.90s
timotimo what happens if you drop the :eager? 16:27
gtodd ie. it's less than 100 times the time for 100 times as many lines ... which I assume is good
timotimo japhb: i'm getting some trouble with perl6-bench; after a ./bench setup i get Don't know how to process component '' 16:28
TimToady how does it compare with slurp.lines?
timotimo after "Setup complete."
tony-o_ timotimo: ugexe fixed a bug in it last night, if you haven't reinstalled since then
gtodd same ... I now grok why :eager would not be needed post GLR
tony-o_ it being perl6-bench
16:28 edehont joined
timotimo i just git pull'd it i think? 16:29
let me see.
"already up to date"
16:29 weihan left
timotimo but the latest commit i see is 3 weeks old 16:30
16:30 matt_ joined
gtodd slurp.lines is ~ 2 seconds faster :-) 28.75 vs 31.25 16:30
:-D
timotimo and there's no pull requests or anything
tony-o_: where did that commit disappear to? 16:31
ugexe are you sure its not assuming its up to date based on the version in the meta.info file?
tony-o_ github.com/tony-o/perl6-bench/pull/8
timotimo: i just merged in the other two PRs from ugexe 16:33
16:33 Ven joined
tony-o_ so a git pull should pull down some killer stuff 16:33
timotimo oh, ok 16:34
oh, you have your own fork
gtodd and the winner is Larry Wall (AKA TimToady) ....
timotimo that explains it i guess
16:35 captain-adequate joined
gtodd time perl6 -n -e ' say ~$/ if .simplematch(/<<\w**5>>/)' 100k-lines.txt 7.49s user 0.08s system 98% cpu 7.681 total 16:35
timotimo ... wait
oh, it's about setting the outer's $/? 16:36
no, hold on, that's in simplematch, too
TimToady no, simplematch does that
gtodd so 7.49 seconds versus 30 seonds ... cause of .simplematch .... maybe I will check to see if the output is the same :-P
TimToady it's just all the rigamarole around handling multiple matches
16:36 cygx_ joined
TimToady was using it the other day to do my timings on the tweet database 16:36
gtodd will m// "know" when it is supposed to be .simplematch in some closer to Xmas variant of perl6 :-) 16:37
timotimo merge conflict in readme.md
TimToady where about ⅓ of the overhead was gather/take, and about ⅔ was :i
gtodd wow perl5 --> perl -wnl -E 'say $1 if /\b(\w{5})\b/' 100k-lines.txt 0.52seconds ... so it's only ~ 15 times faster 16:38
cygx_ Now that people are around, I'll ask again: Any thoughts on letting split(..., :all) return strings and only intermix Match objects in case of split(..., :all, :match)?
gtodd wheee!! (it used to be 75-80 times faster)
cygx_ We're no supposed to be doing high-energy physics, where pulling things apart can spawn all kinds of crazily complicated things ;)
*not
16:38 pRiVi_ joined 16:39 Ven left, Shozan joined
TimToady HLL::Compiler is very confused about what is the language name/version vs compiler name/version, and this shows up in both -v and -V 16:39
16:40 Upasaka_ joined, intrigue_ joined, intrigue_ left, intrigue_ joined, iv_ joined, btyler_ joined
TimToady "This is perl6 version 2015.09-373-g254c1a8" 16:40
16:41 ZoffixW joined
ZoffixW Shouldn't this flatten? 16:41
m: my @a = <foo bar baz>; my @b; @b.append: @a, 'mer'; say @b.perl
camelia rakudo-moar 254c1a: OUTPUT«[["foo", "bar", "baz"], "mer"]␤»
timotimo tony-o_: dooooood
tony-o_: that's a totally different perl6-bench you have there
[Coke] stares at htmlify
TimToady should be more like "This is rakudo version 2015.09-373-g254c1a8 implementing Perl v6b" or so 16:42
jnthn TimToady: I susepct the code you're looking at predates us having much clue how such things would pan out :)
16:44 apotheon_ joined
TimToady RT'd it, since it's fruit that's hanging somehwere 16:46
timotimo bonks tony-o_ for deceiving him so bad
gtodd reads about simplematch() and m/ and m:i etc. ... argh :-)
ZoffixW doesn't follow the subleties between .push and .append: 16:47
TimToady ZoffixW: no, that's two arguments under the one-arg rule
ZoffixW m: my @a = <foo bar baz>; my @b; @b.push(@a, 'mer'); my @c; @c.append(@a, 'mer'); say [ @b, @c ].perl
camelia rakudo-moar 254c1a: OUTPUT«[[["foo", "bar", "baz"], "mer"], [["foo", "bar", "baz"], "mer"]]␤»
TimToady .push only pushes arguments, .append only pushes a list
16:47 Ven joined
ZoffixW Alright. I'll read up on the difference between the two. Thanks. 16:48
TimToady because .push only pushes arguments, it doesn't care if any of those arguments are lists
16:48 sftp_ joined
TimToady because .append only pushes a list, it doesn't care whether that list was produced by argument comma or something else 16:48
timotimo m: my $spec = "/home/oldhome/timo/perl6/fresh_bench/components/arane"; my $COMPONENTS_DIR = "/home/oldhome/timo/perl6/fresh_bench/components"; $spec .= subst(/^ $COMPONENTS_DIR '/' /, ''); say $spec.perl
camelia rakudo-moar 254c1a: OUTPUT«"arane"␤»
timotimo for some reason that doesn't remove the text locally 16:49
16:49 Kogurr joined, espadrine left, jkva joined 16:50 softmoth left, margeas left, skids left, colomon left, apotheon left, Alina-malina left, PerlJam left, Upasaka left, Humbedooh left, go|dfish left, SHODAN left, jantore left, oha left, freeze left, f3ew left, spacebat left, danstoner left, pRiVi left, sjohnson left, mst left, btyler left, lucs left, jdv79 left, confuseAcat left, bjz_ left, adhoc left, Woodi left, jevin_ left, |Tux| left, woodruffw left, Util left, bbarker`` left, tony-o_ left, ingy left
timotimo m: my @specs = "/home/oldhome/timo/perl6/fresh_bench/components/arane"; my $COMPONENTS_DIR = "/home/oldhome/timo/perl6/fresh_bench/components"; for @specs -> $spec is copy { $spec .= subst(/^ $COMPONENTS_DIR '/' /, ''); say $spec.perl } 16:50
ZoffixW TimToady++ # That makes sense now. Thanks.
Weeeeeeeeeeeeeeeeee \o/
camelia rakudo-moar 254c1a: OUTPUT«"arane"␤»
16:50 sftp_ is now known as sftp
TimToady m: say 'foo' ~~ S/foo/bar/ 16:51
camelia rakudo-moar 254c1a: OUTPUT«Potential difficulties:␤ Smartmatch with S/// can never succeed because the subsequent string match will fail␤ at /tmp/TCGa4pn2jf:1␤ ------> 3say 'foo' ~~ S/foo/bar/7⏏5<EOL>␤False␤»
TimToady which is not strictly true, but close enough
m: say 'foo' ~~ S/foo/foo/
camelia rakudo-moar 254c1a: OUTPUT«Potential difficulties:␤ Smartmatch with S/// can never succeed because the subsequent string match will fail␤ at /tmp/Ffuqn5y8DN:1␤ ------> 3say 'foo' ~~ S/foo/foo/7⏏5<EOL>␤True␤»
timotimo damn 16:52
we have a strange bug going on here
16:52 jkva_ joined
timotimo $COMPONENTS_DIR gets imported from a module "Bench::Globals" 16:52
by puting a "my $COMPONENTS_DIR = OUTER::('$COMPONENTS_DIR')" at the beginning of this sub, i can get the regex to work 16:53
16:54 jkva left
timotimo wizards, figure this out. i'm too tired and would prefer to not move a single inch today (but still have to do grocery shopping) 16:54
16:54 k2490klm joined 16:55 k2490klm left
TimToady does it also fail with $spec.subst-mutate instead of .=subst? 16:56
timotimo i believe this is about the regex not matching because the $COMPONENTS_DIR doesn't get interpolated
TimToady I'm just wondering about different code paths in the compiler 16:57
I doubt the actual problem would be in .subst
16:58 patrickz joined, abraxxa left, Ven left
TimToady and does it fail with $spec ~~ s///? 16:59
16:59 bjz_ joined, adhoc joined, Woodi joined, jevin_ joined, |Tux| joined, woodruffw joined, Util joined, bbarker`` joined, tony-o_ joined, leedo joined, agentzh joined, mprelude joined, breinbaas joined, ssflying joined, revdiablo joined, sergot joined, atta joined, ingy joined
TimToady looks around for the backlog he dropped a few minutes ago... 17:00
timotimo hold on
i'm running ./bench build at the moment, so can't check immediately 17:01
TimToady np
17:02 firstdayonthejob joined 17:06 softmoth joined 17:07 bjz_ left, adhoc left, Woodi left, jevin_ left, |Tux| left, woodruffw left, Util left, bbarker`` left, tony-o_ left, leedo left, agentzh left, mprelude left, breinbaas left, ssflying left, revdiablo left, sergot left, atta left, ingy left, boggard joined, mrf joined, freeze joined, Grrrr joined, danstoner joined, lucs joined, yoleaux joined, mst joined, PerlJam joined, f3ew joined, oha joined, jantore joined, colomon joined, margeas joined, krakan joined, spacebat joined 17:08 yeltzooo joined, burnersk joined, sjohnson joined, go|dfish joined, skids joined 17:09 Humbedooh joined
timotimo i'll be doing a few nqp benchmarks now 17:09
moar vs jvm
17:10 Humbedooh left, Humbedooh joined
timotimo nine: need a bit more help with building the p6forloop desugar? 17:10
17:10 jdv79 joined, bjz_ joined, adhoc joined, Woodi joined, jevin_ joined, |Tux| joined, woodruffw joined, Util joined, bbarker`` joined, tony-o_ joined, leedo joined, agentzh joined, mprelude joined, breinbaas joined, ssflying joined, revdiablo joined, sergot joined, atta joined, ingy joined
timotimo as soon as that happens, i'll also do rakudo benchmarks 17:11
17:11 virtualsue left 17:12 llfourn left, wtw joined, confuseAcat joined, avar joined, avar left, avar joined, literal joined, crux joined
timotimo TimToady: with only ~~ it says Nil 17:13
hold on, i mistaked 17:14
17:14 FROGGS joined
timotimo i get a match IFF i put in the OUTER:: thingie 17:14
17:15 dakkar left
FROGGS tried yesterday to unbust s/// in for loops 17:16
FROGGS did this after banging the head against jvm for hours
17:17 dha joined 17:18 Alina-malina joined
dalek c/makefile: 6478396 | coke++ | t/pod-convenience.t:
fix test that was passing for the wrong reason
 17:21
17:23 uruwi left 17:24 jkva_ left 17:25 Peter_R joined
TimToady oh yeah, that's the old subst in loop bug, dur 17:31
17:32 n0tjack left, sufrostico left 17:33 apotheon_ is now known as apotheon 17:34 apotheon left, apotheon joined
FROGGS m: my @a = <a b c>; for @a { s[(.)] = "$0" }; say @a # it is as easy as this 17:34
camelia rakudo-moar 254c1a: OUTPUT«Use of Nil in string context in code at /tmp/PD5OPjMW0F:1␤[ a b]␤»
FROGGS m: my @a = <a b c>; for @a { my $/; s[(.)] = "$0" }; say @a
camelia rakudo-moar 254c1a: OUTPUT«[a b c]␤»
17:35 ^elyse^ left
nine [Coke]: wooohooo...many thanks :) 17:37
yoleaux 14:06Z <[Coke]> nine: - please log out and log back in to RT.
ab5tract_ .tell uruwi regarding Roguelike: I can recommend my own github.com/ab5tract/Terminal-Print
yoleaux ab5tract_: I'll pass your message to uruwi.
dalek c: 1a0ecc6 | (David H. Adler)++ | doc/Language/5to6-perlvar.pod:
Changed C<$*EXECUTABLE_NAME> to C<$*EXECUTABLE-NAME> in 5to6-perlvar.pod
 17:38
ab5tract_ .tell uruwi it doesn't yet do the 'capture character without printing it to screen' via any API, but you have access to all of the tput commands that you could need, and a peg to hang more from if desired: github.com/ab5tract/Terminal-Print...mmands.pm6 17:40
yoleaux ab5tract_: I'll pass your message to uruwi.
ab5tract_ .tell uruwi keep in mind though, that I'm currently working on a rewrite of the internals to use a multidimensional array and a react/whenever loop so that we can have multiple threads modifying the screen independently 17:41
yoleaux ab5tract_: I'll pass your message to uruwi.
dalek c: 9413f6c | (Zoffix Znet)++ | doc/Language/5to6-perlfunc.pod:
Mention different flattening behaviour
 17:42
ab5tract_ speaking of -- can anyone clarify the syntax for a multi-dimensional ASSIGN-POS ? I tried emulating core, but it didn't work. Unless it needs to be declared as a multi? But I only want to support a three arg form (x,y,v). 17:43
17:43 kmel joined 17:44 kmel left, kmel joined
nine timotimo: a look at Actions.nqp has made quite clear what you want :) What I have not yet figured out is what exactly the $qast argument of the block contains. 17:45
timotimo: currently trying: method statement_control:sym<for>($/) { my $xblock := $<xblock>.ast; make QAST::Op.new( :op<p6for>, :node($/), $xblock,); } 17:46
FROGGS we've got a p6for op?
dha Zoffix - re that commit: how recent is that change? I may need to rebuild rakudo, but I'm not seeing that behavior.
ZoffixW dha, pretty recent. 17:47
dha Ah, I'll rebuild and see what happens.
ZoffixW m: my @a = <foo bar>; my @b.push: @a; my @c.append: @a; say [ @b, @c ].perl 17:48
camelia rakudo-moar 254c1a: OUTPUT«[[["foo", "bar"],], ["foo", "bar"]]␤»
17:48 tokuhirom joined
nine FROGGS: I'm creating a p6for op_desugar 17:49
FROGGS aha
17:51 ^elyse^ joined
timotimo nine: the thing you get in $qast is the qast node that you've put into the tree with the :op('p6forloop') 17:51
dha Would @a be the first element of @b if @b already had elements
? 17:52
FROGGS no, push goes at the end
ZoffixW Nope
m: my @a = <foo bar>; my @b = <moo mar mer>; @b.push: @a; say @b.perl
camelia rakudo-moar 254c1a: OUTPUT«["moo", "mar", "mer", ["foo", "bar"]]␤»
17:52 tokuhirom left
dha In that case, I find the wording there confusing. 17:52
17:53 xfix joined, jkva joined
dalek c: 3aa60c6 | (Zoffix Znet)++ | doc/Language/5to6-perlfunc.pod:
Use clearer wording for flattening behaviour
 17:53
timotimo nine: i can't help but notice the p6for seems to be lacking the thing to iterate about 17:54
El_Che is it me or is the modules.perl6.org growing fast? 17:55
ZoffixW El_Che, I notice that as well
TimToady maybe you're both growing fast
timotimo it's exploding
ab5tract_ m: class A { has @!a; method ASSIGN-POS(**@b) { my $v = @b.pop; @!a[ @b[0],@b[1] ] = $v }; method AT-POS(**@b) { @!a[ @b[0], @b[1] ] } }; my $t = A.new; $t[6;7] = "fooo"; dd $t[6;7] 17:56
camelia rakudo-moar 254c1a: OUTPUT«("fooo", Any)␤»
ZoffixW 15 new modules in the last 4 days
ab5tract_ TimToady: what's going on with that Any? am I wielding this right? 17:57
FROGGS m: "abc".match(/./); say $/
camelia rakudo-moar 254c1a: OUTPUT«｢a｣␤»
ab5tract_ also, happy belated 6.Birthday! :)
FROGGS I'm considering breaking that...
nine timotimo: isn't that $xblock? 17:58
timotimo: oh
timotimo :) 17:59
tony-o_ timotimo: did you get perl6-bench working?
timotimo tony-o_: i wasn't interested in *that* perl6-bench, i was interested in japhb's :) 18:00
nine timotimo: in the original code $xblock[0] is the thing iterated over and $xblock[1] is the loop body
timotimo but yeah, i was able to get it working
nine timotimo: so my assumption was that having $xblock as p6for's child will give me both
lizmat FROGGS: not returning match objects by default?
timotimo nine: ah, right.
in the desugar, you'll ge texactly that QAST::Op node back 18:01
and you can work from that
FROGGS lizmat: not setting $/ for method calls
the method returns the match anyway
18:01 kmel left
FROGGS and methods should not have side effects usually 18:01
lizmat ah, ok
true
18:02 kmel joined
FROGGS so, if the refactor I am attempting solves the bugs... 18:02
we'll see
timotimo oof
nine timotimo: ok, then it will be interesting to find out why exactly it's failing 18:03
timotimo put more debug statements in :D 18:04
would you like to push it to a branch so i can have a look? 18:05
though, i'll be fooding soon
nine What's the NQP equivalent of $foo.WHAT.say?
arnsholt nqp::say(nqp::what($foo))
FROGGS nqp::say($foo.HOW.name($foo))
nine I definitely need to beef up my nqp debugging abilities. I want to contribute more in this area :) 18:06
18:06 margeas left, azawawi joined
arnsholt Yeah, what FROGGS said is probably what you actually want 18:06
FROGGS nqp-m: my $foo := nqp::list(); nqp::say(nqp::what($foo)); nqp::say($foo.HOW.name($foo))
camelia nqp-moarvm: OUTPUT«␤NQPArray␤»
timotimo debugging nqp stuff is really hard, aye
FROGGS you get used to it :o) 18:07
dha speak for yourself.
azawawi hi 18:08
sergot: ping
18:10 luiz_lha left, luiz_lha joined 18:11 luiz_lha is now known as Guest33524
azawawi .tell sergot just submitted a PR for the 303 redirection logic + preserve headers for HTTP::UserAgent. Thanks for your time. 18:12
yoleaux azawawi: I'll pass your message to sergot.
18:13 uruwi joined
uruwi Ho!!! 18:13
yoleaux 17:37Z <ab5tract_> uruwi: regarding Roguelike: I can recommend my own github.com/ab5tract/Terminal-Print
17:40Z <ab5tract_> uruwi: it doesn't yet do the 'capture character without printing it to screen' via any API, but you have access to all of the tput commands that you could need, and a peg to hang more from if desired: github.com/ab5tract/Terminal-Print...mmands.pm6
17:41Z <ab5tract_> uruwi: keep in mind though, that I'm currently working on a rewrite of the internals to use a multidimensional array and a react/whenever loop so that we can have multiple threads modifying the screen independently
uruwi ありがとう 18:14
見ますよ
(Thanks, I'll be watching it)
azawawi .tell Zoffix it was the HTTP code 303 after all that caused the hanging. Now it is working :) 18:15
yoleaux azawawi: I'll pass your message to Zoffix.
uruwi Wait, I don't know you, do I?
18:15 azawawi left
ZoffixW botsnack 18:16
synbot6 om nom nom
uruwi I can definitely recall several eaux people on a community I used to frequent long long ago
nine How can a QAST::Op object have no op method?
18:17 n0tjack joined
FROGGS nine: what are you doing? :o) 18:17
nine: can you paste it?
18:18 vendethiel joined
nine FROGGS: oh, debug error. It's really an NQPMu 18:18
FROGGS ahh
18:20 leont joined, s_kilk joined
s_kilk hi, I'm new to perl, just started looking at perl6. One thing I can't figure out: what's the difference between an Array and a List? some of the documentation seems to use the terms interchangably 18:21
[Coke] docs.perl6.org/type/Array : An Array is a List which forces all its elements to be scalar containers, which means you can assign to array elements. 18:22
FROGGS and you can push/pop/unshift/shift an Array, but not a List
[Coke] hurm. list says "Indexes into lists and arrays start at 0 by default." ... there is no way to override that, is there? 18:23
s_kilk ah, so a List is immutable? in that the elements of the list are not containers you can assign to?
18:23 punter joined
FROGGS I think there was meant to be a way to customize the starting index... 18:23
punter Is Ubuntu 16.04 going to include the Christmas version of Rakudo? 18:24
sergot azawawi: looks awesome but also makes some tests failing
yoleaux 18:12Z <azawawi> sergot: just submitted a PR for the 303 redirection logic + preserve headers for HTTP::UserAgent. Thanks for your time.
FROGGS s_kilk: if you put a literal value into a list, you cant assign to it.... if you put a (scalar) variable into a list, you probably can assign to that variable
sergot .seen azawawi
yoleaux I saw azawawi 18:15Z in #perl6: <azawawi> .tell Zoffix it was the HTTP code 303 after all that caused the hanging. Now it is working :)
[Coke] punter: doubtful.
FROGGS punter: we hope so... looks good so far 18:25
lizmat s_kilk: same for Map vs Hash
18:25 azawawi joined
[Coke] But, that's really up to ubuntu. 18:25
azawawi sergot: hi :)
s_kilk thanks lizmat, not sure I understand containers yet, but that does help
[Coke] defers to FROGGS.
FROGGS [Coke]: hmmm, well, let's get it into debian first and then I will hunt some ubuntu packagers down :o) 18:26
sergot azawawi: hi :))
FROGGS (with cake or beer or something)
lizmat m: my $p = a => 42; $p.value = 666 # 42 is not a container
camelia rakudo-moar 254c1a: OUTPUT«Cannot modify an immutable Int␤ in block <unit> at /tmp/NdR_kY50e2:1␤␤»
lizmat m: my $a = 42; my $p = a => $a; $p.value = 666 # $a *is* a container 18:27
camelia ( no output )
vendethiel ab5tract_: omg, your Terminal::Print looks amazing!
I was looking for something like this at the SPW
sergot azawawi: looks like your PR makes some tests failing
azawawi: could you please check that?
azawawi sergot: yeah im looking at them 18:28
sergot azawawi++ 18:29
azawawi sergot: the funny thing is that they are random
sergot azawawi: hmm, that's weird 18:30
azawawi SSL?
sergot I've just rerun travis build to see if it happens again 18:32
and it does :(
nine It works!
lizmat nine++
[Coke] tries to run the terminal print examples and ends up in panda hell.
vendethiel nine++ lizmat++
sergot ua-simple.t and ua.t are failing, so this is not SSL problem I think
brb azawawi
kmel It Works! => localhost :) 18:33
;)
azawawi sergot: sure i will take a look at it
kmel wasn't this the apache welcome page 18:34
18:34 diana_olhovik_ joined
azawawi sergot: found it... cloning headers is bad in redirection... my bad... fixing 18:35
dha Ugh. installing Task::Star got shot down by File::Temp failure. Is that something I should be reporting somewhere? 18:38
dalek kudo/nom: 59d4c40 | ugexe++ | src/core/CompUnitRepo/Local/Installation.pm:
Supress CURLI warnings

If the manifest is missing precomp info for a module it shouldn't warn each time one is loaded/used in a script, or complain about trying to use a value from an uninitialized hash key.
This carries over the mistaken `:has_precomp` argument		 18:39
moritz dha: in the File::Temp repo's issue track
dalek kudo/nom: 3f3e032 | lizmat++ | src/core/CompUnitRepo/Local/Installation.pm:
Merge pull request #552 from ugexe/patch-4

Supress CURLI warnings
[Coke] panda is broken.
------> my $panda = Panda.new(:ecosystem(⏏make-default-ecosystem($prefix)));
Calling make-default-ecosystem(Str) will never work with declared signature ()
^^ few dozen of those.
uruwi Can't install any modules with Panda on cygwin. pastebin.com/XiNY0yqs 18:40
dha Oh. so is this Panda's fault or File::Temp's fault?
uruwi I have no idea. 18:41
[ptc] has seen that in association with File::Temp more than once 18:42
sergot azawawi: why is it bad?
18:42 ^elyse^ left
dha Huh. also, the list of modules in Task::Star's META.info file on modules.perl6.org does not include File::Temp... 18:43
uruwi ptc elaboration?
dha Ah. It's installing it in order to install p6doc.
18:44 telex left
[ptc] uruwi: I wish I had the tracebacks, but unfortunately I don't. Upon running e.g. ./rebootstrap.pl or simply installing the modules for perl6-examples, I'd noticed File::Temp failing 18:44
azawawi sergot: it needs to exclude some headers as it broke some tests. I will revisit it.
sergot: the main problem was the 303 redirection
uruwi Is it the same error message?
[ptc] uruwi: then when I tried again later, it seemed to work (usually I'd also updated the full rakudo stack, panda etc., thus it's very hard to nail down where the problem was
uruwi: sorry mate, I can't remember 18:45
uruwi What version of Rakudo are you running?
2015.09?
ugexe 303 redirection used to work correctly
[ptc] uruwi: This is perl6 version 2015.09-363-g70eea19 built on MoarVM version 2015.09-81-g41002db
18:46 telex joined
uruwi Strangely enough, mine doesn't provide that much detail 18:46
same problem in cmd.exe 18:47
sergot azawawi: well, travis shows it still fails :(
uruwi So it happens with Windows, period.
[ptc] one thing I have had reasonably consistently is "Could not find file 'Build.pm' for module Build.pm" when installing Linenoise
this is also something which is a bit flappy 18:48
uruwi Thinking of giving up on Cygwin and using full-on Linux instead... 18:51
How does no one else have this error, by the way? 18:52
Some person *has* to be running Perl6 on Windows 18:53
azawawi sergot: it sometimes hangs on 08-ua.t . Is that normal?
kmel m: say 'What is windows?'
camelia rakudo-moar 254c1a: OUTPUT«What is windows?␤»
El_Che uruwi: you got rakudo built on cygwin? 18:54
uruwi No, I'm using the regular windows build
Panda doesn't work from cmd either so this probably doesn't have anything to do with the shell
El_Che ah ok (I tried and failed :) )
18:55 s_kilk left
[ptc] uruwi: afaik FROGGS and jnthn both use Windows 18:55
18:55 jkva left
ugexe i run perl6 on windows 18:56
FROGGS and panda does work with latest star release (2015.09)
azawawi sergot: i think it is now a problem with github.com/perlpilot/p6-File-Temp/
uruwi Hmm
FROGGS I've not checked a newer version of rakudo
El_Che uruwi: starting a vm with windows 7, cygwin and rakudo 18:57
uruwi Oh
FROGGS why cygwin?
uruwi I use Win10 :P
ugexe im pretty sure its file::temp or file::directory::tree, and star is bundled what a specific version that still works
azawawi sergot: File::Temp is actually failing its tests with the latest batch of commits 18:58
18:58 jkva joined
azawawi wishes for panda module version dependency 18:58
18:59 rindolf left
El_Che uruwi: same result in the Win VM 19:01
uruwi: same error message
uruwi: latest rakudo installed
uruwi Wait the modules in Rakudo* don't auto update do they. 19:02
I infer not.
El_Che however panda install Task::Star completed without errors (although it probably did nothing) 19:03
19:03 jkva left
uruwi Succeeds on my machine too 19:04
sergot azawawi: looks so.. hmm. 19:06
El_Che uruwi: skipping tests (--notests) builds and installs the modules succesfully
19:06 [Sno] left, jkva joined
ugexe zef works fwiw 19:06
task::star probably isnt trying to install a new File::Temp if you already have panda installed
dalek kudo/nom: 17987c6 | lizmat++ | src/core/Exception.pm:
Add 'type' to X::Adverb::Slice
 19:07
kudo/nom: 3cdaf9a | lizmat++ | src/core/Any.pm:
Handle unhandled named parameters to zen slice
uruwi Panda still tries to test even with --notests
azawawi sergot: so basically my headers fix was ok :)
El_Che panda --notests install ...
the params matching is a little basic it seems 19:08
dalek ast: 29f5477 | lizmat++ | S32- (2 files):
Adapt test for zen slice with unknown adverb
ugexe the thing is, we were cloning http::response like a week ago and it was removed
uruwi oh --notests comes before install
ugexe and its starting to get ugly with http::request generating other requests and modifying attributes instead of the transport/ua handling that
whereas with the cloned history the transport/ua can handle that stuff 19:09
19:09 punter left, rotcpy joined 19:10 llfourn joined, jkva left
sergot ugexe: we're talking about travis, it should install the latest File::Temp, right? 19:10
19:11 jkva joined
ugexe yes. but its not testing it so you cant see that its broke from your travis output 19:13
azawawi m: $*SPEC.catpath('A', 'B'); 19:14
camelia rakudo-moar 254c1a: OUTPUT«Too few positionals passed; expected 4 arguments but got 3␤ in block <unit> at /tmp/Q5pTBq7vrM:1␤␤»
timotimo a few push replaced with append and bench is a bit happier
azawawi m: $*SPEC.catpath('', 'A', 'B'); 19:15
camelia ( no output )
19:15 n0tjack left
nine timotimo: the desugar op breaks loop control statements and for's return value. Will need some restructuring to get that working 19:15
19:15 llfourn left
azawawi this commit broke File::Temp, github.com/perlpilot/p6-File-Temp/...6ed6056429 19:15
sergot ugexe: oh, ok, right 19:16
uruwi Different error message though 19:18
timotimo nine: damn :(
jnthn: do you have a hunch why a desugar for for loops in rakudo would break things like nine just described?
19:18 ZoffixW left
nine timotimo: I guess it breaks the return value because the QAST::Want must be the outermost op 19:19
No idea where the "control exception without handler" comes from though 19:20
timotimo so are you going to push it up so i can have a quick look? 19:21
nine timotimo: current work in progress of trying to push the QAST::Want to the top: gist.github.com/niner/c08db40160de90a7c46b 19:23
19:23 rindolf joined
timotimo something's wrong with the perl5 portion of perl6-bench, it seems like o_O 19:25
azawawi sergot: anyway, github.com/azawawi/perl6-selenium-...antomjs.p6 is now working with the fix and it is controlling phantomjs :) 19:26
sergot: thanks 19:27
timotimo oh! 19:28
could have something to do with the timings files for nqp-jvm being 260 megabytes big?
huh. the ones in nqp-moar are just half that size 19:29
japhb: anything we can do to make this less of a problem? 19:30
[Coke] ooh, p6 selenium
azawawi [Coke]: yup :) 19:31
sergot azawawi++ : awesome! thank you! :) 19:34
azawawi &
19:34 johnjohn101 joined
johnjohn101 hi perl six peeps 19:34
sergot johnjohn101 hi \o
PerlJam johnjohn101: greetings 19:35
johnjohn101 what ever happened to the perl 6 on .net project? seems like it hasn't had an update in a while.
PerlJam johnjohn101: champions are hard to come by. 19:36
johnjohn101 i understand
moritz in the past, I've wondered if writing a .NET backend for NQP was a good way to learn C# 19:37
19:37 azawawi left
moritz I came to the conclusion that it likely isn't :-) 19:37
dalek kudo/nom: ee96fc1 | lizmat++ | src/core/Any-iterable-methods.pm:
Make List.first/first-index 2x as fast
19:38 pmurias_ joined
[Coke] pokes the one guy he knows that works at microsoft about it. 19:40
19:41 pmurias left
masak m: class A { method m { "A" } }; class B is A { method m { "B"; } }; class C is A { method m { "C" } }; class D is B is C {}; say D.new.m 19:42
camelia rakudo-moar 3cdaf9: OUTPUT«B␤»
19:44 kmel left 19:46 jkva left
masak spec.perl6.org/ doesn't take me to spec.perl6.org/specification/ :( 19:49
it takes me to the main page
19:49 tokuhirom joined
timotimo t.h8.lv/p6bench/2015-10-14-nqp_shootout.html - this was also supposed to have nqp-jvm in it, but it already took super long to do only the nqp-moar bits 19:50
i'm going to re-run the analyze command with also jvm parts
19:54 tokuhirom left
dalek kudo/nom: dd15d70 | lizmat++ | src/core/Supply.pm:
Give Supply.squish same semantics as List.squish
 19:59
PerlJam isn't spec.perl6.org/specification/ kind of a misnomer anyway? 20:00
dalek ast: 7ed4852 | lizmat++ | S17-supply/squish.t:
Add tests for 1 elem semantics of Supply.squish
20:03 larion joined
timotimo well, it does link to the test suite 20:03
20:04 ^elyse^ joined 20:05 cygx_ left
timotimo t.h8.lv/p6bench/2015-10-14-nqp_shoo...h_jvm.html 20:07
20:07 s_kilk joined 20:08 hohoho joined
uruwi Ho! 20:11
timotimo Hey!
lizmat timotimo: that's pretty bad, no ? 20:12
[Coke] have a 750 line or so perl6 script. add a multi MAIN("TEST") near a sub I'm trying to debug so I can quickly invoke a test without changing the rest of the code.
timotimo what is?
lizmat that shootout
the while empty ones
20:13 hohoho left
timotimo why do you think so? 20:13
PerlJam lizmat: bad for moar compared to jvm?
lizmat well, the nqp::jvm does 4x more ?
FROGGS head seems identical to 2015.07 and 2015.09
timotimo when it warms up, yeah
FROGGS lizmat: they payed more ppl to optimize that vm
paid*
lizmat hehe: I bet there's some dieselgate in there as well :-) 20:14
FROGGS one could say that the JVM is optimized to do nothing :o)
hehe
gugod or collecting garbage
dalek ast: 364b771 | usev6++ | S (6 files):
Unfudge some tests for rakudo.jvm
 20:15
timotimo i'm not sure those benchmarks are worth anything, tbh :)
in a few hours i'll have some rakudo benchmarks
lizmat looking forward to that :-)
20:16 colomon left, TEttinger joined
jnthn Such benchmarks are mostly a good test of how good your optimizer is at getting rid of "junk" and spitting out concise machine code. So, some value. :) 20:18
[Coke] if I have a parallel pipeline (e.g. something with .race.map.***) - what can I use to stop the pipeline? jnthn's example uses map...
timotimo right
jnthn 'night, #perl6
timotimo gnite jnthn!
lizmat gnight jnthn 20:19
20:23 darutoko left
[Coke] er, uses 'max' 20:24
dalek ast: 4a8cdeb | usev6++ | S02-literals/numeric.t:
Add another test for RT #124559
 20:25
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=124559
dha List.eager doc says "Evaluates all elements in the list eagerly, and returns them as a list." What does that mean? 20:33
specifically, what does it mean to evaluate something "eagerly"? 20:34
20:34 colomon joined
PerlJam some elements may be a "thunk" that generates list elements rather than actually be elements. Eager evaluation will cause it to generate all of the values it can. 20:37
pink_mist that may lead to out of memory problems with infinite lists =) 20:38
dha ok. The docs should probably explain that. I'm not convinced I'm the person to do that, however.
psch "Eager evaluation" is a wikipedia article 20:39
that's not to say we don't need it in the docs, though
20:42 DrForr_ left
dha Ok, I'm confused. What is going on here: say (1, 2, 3).reduce: * - *; 20:46
This results in -4, but I'm not clear why.
psch m: say 1 - 2 - 3
camelia rakudo-moar dd15d7: OUTPUT«-4␤»
psch m: [-] 1, 2, 3
camelia ( no output )
psch m: say [-] 1, 2, 3
camelia rakudo-moar dd15d7: OUTPUT«-4␤»
dha ok. I get that. What are the stars doing? 20:47
psch just a different form of specifiying the operator or function
'* - *' is a two-ary function that substracts the second argument from the first
dha and where is that documented?
Oh, placeholders. damn that looks weird. 20:48
psch just like &infix:<-> is a two-ary function that substracts the second argument from the first
dha C<*> means too many things, dammit.
PerlJam aye, there is a little of that.
psch m: say *.WHAT 20:49
camelia rakudo-moar dd15d7: OUTPUT«(Whatever)␤»
PerlJam same with { }
psch Whatever should be somewhat documented, iirc
yeah, pretty thoroughly even, it looks like
RabidGravy doc.perl6.org/type/Whatever
20:52 [Sno] joined
dha Having to read a section of an entirely different doc page seems problematic for an example. :-/ 20:53
skids
.oO(maybe .last-index should be renamed to .hind-index so we can have .hind which would not conflict with "last" and kinda rhymes with "find" anyway.)
psch dha: you're doc'ing .reduce? 20:54
PerlJam would rather all of the *-index routines go away first
20:54 FROGGS left
dha I'm looking through List.pod for where I can put in examples, and reading the ones that have examples already as I go. 20:54
That one caught my eye.
psch ah, i see 20:55
lizmat PerlJam: how would they be called then ?
PerlJam lizmat: something like grep :index, { ... }, @stuff; 20:57
skids .first(:k)?
masak .oO( .grep :kv, .grep :v ) 20:58
PerlJam in any case, some modifier that say we want the ordinals rather than the values.
lizmat well, I'm fine with that: that's how I originally implemented them, I think 20:59
masak PerlJam: also, some other modifier that says we want the mutable container rather than the value in the container
lizmat but then the whole thing about named parameters are a smell came about
PerlJam masak: that would be nice too
lizmat masak: you get the container already if they are containers 21:00
21:00 kmel joined, niceperl joined
kmel i am reading doc.perl6.org/language/control 21:01
can someone explain what does 'with' actually do?
PerlJam kmel: with $thing { ... } is like if defined($thing) { ... } 21:02
masak m: my @a = 1, 2, 3; @a.grep(2)[0] = 7; say @a
camelia rakudo-moar dd15d7: OUTPUT«Cannot modify an immutable Int␤ in block <unit> at /tmp/lTm_2l8Fws:1␤␤»
masak lizmat: don't look like it 21:03
lizmat masak PerlJam : I will let TimToady rule on that, and make changes accordingly
kmel oh ok. thanks PerlJam
skids kmel: like if defined($thing) { ... }($thing) actually.
or well if defined($thing) {{ ... }($thing)} 21:04
PerlJam lizmat: yeah, he's the best person to ferret out any underlying commonality and unify it into a syntax that makes sense
lizmat masak: oops, looks like they're some "is raw"'s missing
kmel thanks skids
masak lizmat: tbh, I think "just the values" is an excellent default for .grep 21:05
lizmat: 99% of the time I use .grep, that's what I expect
PerlJam lizmat: and I wouldn't hold up 6.christmas for it in any case. :)
masak probably great for optimization, too
lizmat masak: what is ?
21:05 uruwi left
masak returning just the values, not their containers 21:05
it's one less level of indirection to worry about 21:06
PerlJam since you already have the containers, just returning them seems already "optimized"
lizmat m: my @a = 1, 2, 3; $_ = 7 for @a.grep(2); say @a
camelia rakudo-moar dd15d7: OUTPUT«[1 7 3]␤»
skids Depends entirely on use case I would say. 21:07
lizmat masak: I think this is really a problem of the iterator interface
your example did a push-all, which pushes values into a buffer
the for loop does pull-one, which gets the container 21:08
not sure how to solve that
dalek c: 90ebc3c | (David H. Adler)++ | doc/Type/List.pod:
Added examples for grep-index, first-index, and last-index
 21:09
c: 1c87ed7 | RabidGravy++ | doc/Type/List.pod:
Merge pull request #163 from dha/master

Added examples for grep-index, first-index, and last-index
dha I assume that there is a difference between Range.list and Range.flat, but they are documented identically. Or are they in fact just synonyms?
skids m: my @a = 1, 2, 2, 3; @a.grep(2)[0..1] = 7,8; say @a;
camelia rakudo-moar dd15d7: OUTPUT«Unexpected named parameter 'eagerize' passed␤ in block <unit> at /tmp/fGpowvDynL:1␤␤»
21:10 diana_olhovik___ joined
skids decommute 21:10
21:10 llfourn joined, diana_olhovik_ left
PerlJam dha: I think .list gives you an actual List, but .flat gives you a Seq 21:10
dha According to the docs, they both return a List. 21:11
and they both just say "Generates the list of elements that the range represents."
lizmat .tell jnthn couldn't we just not put a $no-sink in IterationBuffer.push ? 21:12
yoleaux lizmat: I'll pass your message to jnthn.
masak lizmat: I see. hm.
21:13 xfix left 21:14 colomon left, skids left
RabidGravy dha, it's entirely possible the docs weren't updated post-GLR 21:15
dha That's what it looks like, yes.
21:16 llfourn left
dha So that should be changed to "returns Seq:D"? And the doc perhaps changed to "Generates a sequence of elements..."? 21:17
lizmat well, not sure we want to expose Seq so clearly ? 21:18
PerlJam m: (^10).flat.WHAT.say 21:19
camelia rakudo-moar dd15d7: OUTPUT«(Seq)␤»
PerlJam It's already exposed to some extent
dha It certainly seems odd to have two methods that do the same thing. Also, since the function documentation has signatures, they should probably be correct... 21:20
21:21 margeas joined 21:22 diana_olhovik___ left
dha I suppose you could just omit the return type from .flat, but that strikes me as inconsistent and weird. But maybe that's just me. 21:22
s_kilk The SCGI module appears to be broken currently, because its dependency Netstring doesn't build: pastebin.com/KaKrK7ku
Which means that Web::App::Dispatch is broken too, I think, or at least the examples are. will dig in a bit further 21:23
PerlJam s_kilk: ping supernovus harder ... there's already a PR to fix it from back in May 21:24
dalek kudo/nom: d578a56 | lizmat++ | src/core/Supply.pm:
Introduce Supply.throttle

So you can throttle your supplies by number of elems per time-unit
  (in seconds, or parts thereof)
PerlJam ptc++ (I'm pretty sure he did bunches of PRs to add unit declarations to most modules in the ecosystem)
s_kilk Thanks PerlJam, I'll open an issue on the SCGI repo. I'm new to perl6 (new to perl actually), and have been digging around for a place to start contributing, looks like I found it :) 21:25
PerlJam s_kilk: alternatively, if no one can contact supernovus, someone can fork his repo, add the fix, then replace his repo in the ecosystem with their own. 21:26
21:26 Peter_R left
PerlJam .seen supernovus 21:27
yoleaux I saw supernovus 30 Sep 2015 21:08Z in #perl6: <supernovus> I changed my build script so instead of using bootstrap.pl it uses rebootstrap.pl and it appears to work properly now.
PerlJam He was here a couple of weeks ago apparently. 21:28
dha Can I assume the examples for .floor, .ceiling, etc. from Cool.pod should be appropriate for, for instance, Real.pod?
s_kilk I'm tempted to do that actually, what's the current state of HTTP/Web stuff in Perl6? As a newcomer it seems that supernovus is responsible for most of the web projects in the ecosystem, is he pulling the cart alone? Is there a group of contributors I could speak to about me pitching in? I'd be keen on even just doing doc work
dha (Cool has lots of examples, Real, not so much) 21:29
RabidGravy s_kilk, there's a few people sparked up recently - I think that stuff is going to grow a bit 21:30
s_kilk Cool, I bet a handful of people could make some real progress with the right co-ordination
RabidGravy Crust came in just recently and those guys have been patching a load of other stuff which is cool 21:31
s_kilk The Web::App::Dispatch framework looks really promising, but it's currently broke.
Ah, thanks RabidGravy, I hadn't seen Crust before. 21:32
RabidGravy I think in the fullness of time something really good will pop up, someone clever will have the urge to write the best web framework ever and actually have the time to try 21:36
21:37 colomon joined, ZoffixMobile joined 21:38 virtualsue joined
ZoffixMobile I really wish sri of the Perl 5's Mojolicious fame would finally decide to do a Perl 6 port. So far, they're unsure/need funding. 21:38
21:39 AlexDaniel joined
RabidGravy I don't, I want someone new to write something new 21:39
ZoffixMobile Even aside from the web framework, I miss the Mojo Toolkit in P6 (Mojo::DOM in particular)
RabidGravy, how come? sri knows the specs like the back of their hand and has creating two very successful web frameworks on their CV. 21:40
masak ZoffixMobile: in the meantime, I'm pretty sure you can use Mojo from Inline::Perl5 21:41
ZoffixMobile masak, yeah, you can (I tried).
21:42 sufrostico joined, sufrosti1o joined
RabidGravy ZoffixMobile, that doesn't mean someone else can't make something better and more Perl 6 like :) It would be really nice if someone who wasn't an existing Perl 5 author did that :) 21:43
ZoffixMobile But I wouldn't make a P6 module that has Inline::Perl5 and a Perl 5 module as a prereq; which is my more accurate feeling about missing Mojo :p 21:44
21:45 pierrot joined 21:46 muraiki left
ZoffixMobile RabidGravy, that's definitely true. :) 21:47
tony-o_ ZoffixMobile: i'd be interested in feedback on github.com/tony-o/perl6-hiker
21:47 ggoebel joined, sufrostico left 21:48 sufrosti1o left
s_kilk so, what would it take to replace the Netstring module in the ecosystem with a forked version? 21:48
ZoffixMobile .tell Zoffix check out github.com/tony-o/perl6-hiker when you're sober
yoleaux ZoffixMobile: I'll pass your message to Zoffix.
ZoffixMobile :p
21:48 ggoebel2 left
tony-o_ lol ZoffixMobile++ 21:48
leont integrating perl6's async power into a framework is where it starts being truly interesting to me, copying existing things is so-so
RabidGravy I'm with leont :) 21:49
tony-o_ hiker uses the async stuff
ZoffixMobile s_kilk, I suggest editing META.list in the github.com/perl6/ecosystem, while doing an @ mention of the original author of the intent to overtake. Wait 24 hours. Merge your overtake 21:50
21:51 tokuhirom joined
s_kilk Thanks ZoffixMobile, I've forked the project and patched it here: github.com/ShaneKilkelly/perl6-netstring . it seems to work, i think. 21:51
RabidGravy I'm taking a more relaxed approach to Cache::Memcached - I sent a PR to make it actually work in May, emailed him yesterday and will wait a week or two 21:52
ZoffixMobile Oh sure 21:53
Sorry, I assumed s_kilk attempted to contact the original author with no avail.
21:54 hohoho joined, patrickz left
s_kilk yeah, I think Netstring may be more broke than I realized, it installs now but crashes the server with "Invalid netstring stream data.". Not sure which side of the Web::App / SCGI / Netstring boundary is at fault there. 21:55
21:55 tokuhirom left
s_kilk sorry, I thought if it was an easy fix that it might be good to take some pressure off of supernovus. I've just wandered in off the street really, I'm quite taken with perl6, and was starting to play around with some of the web modules. 21:56
RabidGravy well, sounds like Netstring may or may not be properly tested 21:57
I'll put this on my TODO list for tomorrow 21:58
(it's actually more of a stack than a list nowadays but hey)
21:59 hohoho left, virtualsue left 22:00 kid51 joined
dalek ast: ef6d4ee | lizmat++ | S17-supply/throttle.t:
Add basic Supply.throttle tests
 22:00
RabidGravy right bed time 22:01
lizmat gnight RabidGravy
ZoffixMobile s_kilk, yeah, Perl 6 is pretty cool :) If you haven't already, submit a PR to supernovus that fixes the issues. If they don't merge it soon enough, try to contact them about the merge, and if that fail, do the ecosystem editing stuff I suggested :) Sorry, my fault for not being clearer earlier :)
gnight
s_kilk Thanks ZoffixMobile, it looks like someone has had the same PR open for the last few months. I've added a comment to the PR too. 22:03
dha sees the Document All Modules message and looks for modules in need of documentation. 22:04
22:06 RabidGravy left
dalek kudo/nom: 271e84c | lizmat++ | t/spectest.data:
Run the Supply.throttle tests
 22:06
ZoffixMobile Thinking more of this, we probably need a more official guideline for module overtakes. I can imagine people getting touchy about someone just taking over their work willy-nilly
22:06 bjz_ left
ZoffixMobile dha, make a list! :) 22:07
dha On what basis?
That is, one what basis would I add something to this hypothetical list? :-)
22:07 bjz joined
_sri ZoffixMobile: thanks for the vote of confidence :) 22:08
ZoffixMobile _sri, :) Join us! We have cookies! :p 22:09
_sri RabidGravy: it's not like perl6 is full of new programming paradigms that could result in radically new ways of developing web apps 22:10
(and that doesn't mean it's bad at all!)
dha We do? cookies++
ZoffixMobile dha, does the documentation included with the module tell everything about how to use that module? The "this is a Port of Perl 5 blah" doesn't cut it. :)
dha *nod* ok.
Ooh. But now I have to get going. Something for tomorrow, I guess. 22:11
later, all.
22:11 dha left
ZoffixMobile \o 22:11
dalek c: 913b7b8 | lizmat++ | doc/Type/Supply.pod:
Document Supply.throttle
 22:12
rl6-most-wanted: 8849c04 | (Nova Patch)++ | most-wanted/ (2 files):
add wanted i18n/NLP
lizmat good night, #perl6!
ZoffixMobile night, lizmat 22:13
masak 'night, lizmat 22:15
lizmat .tell daxim if you can give me a paragraph about why my statement re ß in the P6W was wrong, I will add it as an erratum next week 22:16
yoleaux lizmat: I'll pass your message to daxim.
lizmat really afk&
22:17 ^elyse^ left
dalek rl6-most-wanted: e1c1feb | (Nova Patch)++ | most-wanted/ (2 files):
i18n and NLP headers
 22:17
22:18 s_kilk left
leont _sri: I've seen few frameworks/languages handle concurrency and async well at the same time 22:19
few meaning "I haven't seen it, though it might exist, it must have not broken through"
[ptc] PerlJam: thanks :-) Yes, I did. I also wanted to fix kebab-case issues as well, but time sort of ran out on me 22:20
_sri leont: phoenix is pretty phenomenal
i don't think anything in perl6 will ever be able to compete with that honestly 22:21
(which again, is not bad! ;p)
due to the erlang vm and otp
play framework should be closer to what a perl6 framework will ultimately look like i guess, it's pretty good at concurrency/parallelism 22:23
leont I should look into that 22:24
_sri but yea, if you wanted to make concurrency/parallelism your big selling point you'll have to compete with phoenix/elixir, which would be an uphill battle 22:26
22:26 edehont left, niceperl left
_sri it's not even such a radically innovative framework, it just takes advantage of all the erlang infrastructure 22:26
leont Only worked with erlang once, but yeah that was impressive 22:27
22:28 vendethiel left
johnjohn101 erlang is ok but seems to me specialized usage 22:28
_sri would focus more on expressiveness, like in perl5 22:29
22:31 johnjohn101 left 22:32 firstdayonthejob left
_sri totally take advantage of the awesome concurrency features (love await and whenever), but have perl-ish stuff like one-liner web apps 22:32
leont Fair point 22:33
ugexe m: use trace; my $x = { :ok(1), :nok(0) }; my @a; @a.append($x); say @a.perl 22:35
camelia rakudo-moar 271e84: OUTPUT«2 (/tmp/WXFbbkbxjB:1)␤my $x = { :ok(1), :nok(0) }␤9 (/tmp/WXFbbkbxjB:1)␤my @a␤10 (/tmp/WXFbbkbxjB:1)␤@a.append($x)␤11 (/tmp/WXFbbkbxjB:1)␤say @a.perl␤[-> ;; $_? is raw { #`(Block|63458872) ... }]␤»
ugexe m: my $x = { :ok(1), :nok(0) }; my @a; @a.append($x); say @a.perl
camelia rakudo-moar 271e84: OUTPUT«[{:nok(0), :ok(1)},]␤»
22:37 uruwi joined
uruwi stupid mouse broke :( 22:38
22:39 ZoffixMobile left 22:40 kmel left 22:45 mattprelude left 22:48 intrigue_ is now known as intrigue 22:49 bjz left
AlexDaniel say '㎉' 22:52
m: say '㎉'
camelia rakudo-moar 271e84: OUTPUT«㎉␤»
22:52 tokuhirom joined
AlexDaniel does anybody know any symbol with more than 4 letters? 22:52
㌲ I'm not sure if japanese counts
but if yes, then I'm searching for a thing with more than 5… ㌮ 22:54
㌖ or 6… 22:56
22:56 tokuhirom left
masak pretty sure those Japanese things you pasted don't contain any letters... :) 22:58
AlexDaniel masak: well, the last one is “kiromeetoru” written with katakana, if I'm not mistaken 23:00
TimToady funny they have that as well as ㎞ 23:01
23:01 skids joined, pmurias_ left
timotimo t.h8.lv/p6bench/2015-10-15-rakudo_m...lease.html - i don't really trust these measurements for some reason ... 23:05
TimToady by some reckoning there's nine letters in 𒐎
23:05 cognominal joined
AlexDaniel TimToady: I wonder if there is any font that supports it 23:07
23:08 rurban left, kmel joined, kmel left 23:09 kmel joined
masak 'night, #perl6 23:10
AlexDaniel split_string_constant, split_string_regex – ouch!
timotimo tries to get some rest and get well 23:12
TimToady AlexDaniel: shows up on my screen just fine :)
timotimo++ 23:13
23:13 llfourn joined
TimToady AlexDaniel: but maybe that's because: 23:14
> ls .fonts/
Aegean.ttf Aegyptus_R.ttf Analecta.ttf Gardiner.ttf HAN NOM B.ttf Musica.ttf Unidings.ttf
Aegyptus_B.ttf Akkadian.ttf Anatolian.ttf HAN NOM A.ttf Maya.ttf Symbola.ttf
23:19 kmel left, llfourn left 23:23 EricMobiel joined
ShimmerFairy m: say "𒐎".uniname 23:30
camelia rakudo-moar 271e84: OUTPUT«CUNEIFORM NUMERIC SIGN NINE DISH␤»
AlexDaniel m: say 'ﯹ'.uniname 23:32
camelia rakudo-moar 271e84: OUTPUT«ARABIC LIGATURE UIGHUR KIRGHIZ YEH WITH HAMZA ABOVE WITH ALEF MAKSURA ISOLATED FORM␤»
AlexDaniel OK, it seems like this one is 18 characters or so 23:34
m: say 'ﷺ'.uniname
camelia rakudo-moar 271e84: OUTPUT«ARABIC LIGATURE SALLALLAHOU ALAYHE WASALLAM␤»
23:46 tokuhirom joined 23:50 tokuhirom left
gfldex m: my constant @a := map { .chr.chars }, 1..33500; say [max] @a; 23:51
camelia rakudo-moar 271e84: OUTPUT«(signal SEGV)»
gfldex i think that's my first segfault rakudobug :)
sadly i cant think of a version that doesnt use map 23:53
r: my constant @a := map { .chr.chars }, 1..33500; say [max] @a; 23:54
camelia rakudo-moar 271e84: OUTPUT«(signal SEGV)»
..rakudo-jvm 271e84: OUTPUT«1␤»
ugexe m: my constant @a := (1..33500)>>.chr>>.chars; say [max] @a; 23:56
camelia rakudo-moar 271e84: OUTPUT«(signal SEGV)»
23:57 bigrattus joined 23:58 luiz_lha joined, luiz_lha is now known as Guest93853, [particle]1 joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!