»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
timotimo you just need to be the user that builds the documentation 00:00
i don't think it has a password set
Zoffix su: Authentication failure ¯\_(ツ)_/¯ tomorrow it is :) 00:01
00:02 Xliff_zzz left 00:03 Xliff_zzz joined 00:08 killbill joined 00:14 zacts left 00:34 jack_rabbit left 00:40 killbill left 00:47 Actualeyes joined 00:49 pdcawley left, pdcawley_ joined 00:57 flaviusb joined 00:58 kurahaupo left 01:02 jack_rabbit joined 01:07 kalkin-_ joined, kalkin- left 01:10 pdcawley_ left 01:19 z3ndrag0n left 01:21 z3ndrag0n joined 01:38 AlexDaniel left
BenGoldberg This is kindof an idle question, but how are character sets, such as /[abc]/ implemented? As bitsets, or a hash, or ... ? 01:39
01:39 km3 joined
BenGoldberg wonders if rakudo uses universal hashes anywhere, and if it would be a space or speed improvement to do so. 01:40
01:44 AlexDaniel joined 01:45 molaf left 01:57 molaf joined 01:59 km3 left, km3 joined 02:00 km3 left 02:07 z3ndrag0n left 02:11 ssotka left
TEttinger BenGoldberg, I'd be curious too. I've seen character classes implemented with bitsets before 02:14
02:25 fhorck left 02:27 girafe left
ShimmerFairy BenGoldberg: not sure, since that's not typically called a character set :P 02:36
geekosaur it is in Icon :p
anyway I think bitsets went out when multibyte characters showed up 02:38
02:39 grassass left 02:42 pmichaud joined
pmichaud m: sub term:foo { }; my $x = foo // 'meow'; 02:42
camelia rakudo-moar 29a110: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Null regex not allowed␤at <tmp>:1␤------> 3sub term:foo { }; my $x = foo //7⏏5 'meow';␤»
pmichaud m: sub term:foo() { }; my $x = foo // 'meow'; 02:43
camelia rakudo-moar 29a110: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Null operator is not allowed␤at <tmp>:1␤------> 3sub term:foo()7⏏5 { }; my $x = foo // 'meow';␤»
BenGoldberg I meant character class, of course ... I was thinking (mathematical) set, because a test for whether a character is in a character class can be thought of as a set-membership test.
pmichaud m: sub term:sym<foo> { }; my $x = foo // 'meow';
camelia ( no output )
Zoffix waves to pmichaud 02:44
pmichaud waves to Zoffix
02:49 skids joined
BenGoldberg Actually, if our character classes are implemented with the Set class, then speeding up Set will of course speed up regex matching... 02:50
02:52 grassass joined 02:59 noganex_ joined 03:02 noganex left 03:13 Actualeyes left 03:23 Calibell_ joined 03:29 AlexDaniel left 03:31 ssotka joined 03:34 kid51 left 03:36 Actualeyes joined
Zoffix .ask masak can this ticket be closed, after adding the tests for current behaviour? It was suggested I ask you: rt.perl.org/Ticket/Display.html?id=70297 03:37
yoleaux Zoffix: I'll pass your message to masak.
03:54 canopus left 04:00 canopus joined 04:12 notostraca joined 04:13 BenGoldberg left 04:14 TEttinger left, Calibell_ left 04:16 notostraca is now known as TEttinger 04:36 pdcawley joined 04:49 Cabanossi left 04:50 Cabanossi joined, khw left 05:07 skids left 05:10 buharin joined 05:11 Actualeyes left 05:17 ssotka left 05:22 djbkd joined 05:23 pdcawley left 05:24 pdcawley joined
sortiz m: sub foo { fail }; foo // say 'meow'; # Ambiguous //, bare foo expects TERM, not OP 05:26
camelia rakudo-moar 29a110: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Null regex not allowed␤at <tmp>:1␤------> 3sub foo { fail }; foo //7⏏5 say 'meow'; # Ambiguous //, bare foo ex␤»
sortiz m: sub foo { fail }; (foo) // say 'meow'; # Works 05:27
camelia rakudo-moar 29a110: OUTPUT«meow␤»
sortiz m: sub foo { fail }; foo() // say 'meow'; # Works
camelia rakudo-moar 29a110: OUTPUT«meow␤»
05:31 addison left 05:38 Sgeo left
buharin hi 05:40
;-)
sortiz \o buharin
buharin however I used Perl
but I can't install any cpan ui framework 05:41
it is failing both os x and debian
I tried Gtk3
Wx
Tkx
Tk
all failing
konobi perl5? 05:42
buharin yep 05:43
sammers good afternoon from Japan 05:44
sortiz \o sammers
sammers hi sortiz
05:45 molaf left
sammers ok, so I am messing around with promises and I am getting this weird intermitent error (Use of Nil in string context). here is a gist gitlab.com/snippets/22152 05:47
sometimes it works perfectly, sometimes the error is throw. I have to use a sleep inside the promise to ensure that the error never appears. 05:48
konobi buharin: try #perl on irc.perl.org 05:50
buharin konobi, I tried
sammers so I guess my question is how do I track down why 'Use of Nil in string context' is thrown?
konobi buharin: you'll just need to wait about
sortiz sammers, Can you try creating $http inside the start block? (I'm only guessing) 05:51
sammers yeah, I had that but it seems to work the same. 05:52
Woodi hi #perl6 :) 05:54
sammers is my assumption correct that this *should* work as intended? where the http request is queued and then the await *should* wait for it to complete before moving on?? 05:55
hi Woodi
I am fine with bugs, I am just trying to validate my assumptions 05:56
Woodi jnthn++ for kicking obvious abuser. obviously Allah approved :) however: a) it's possible Allah is almighty; b) for sure, jnthn++ isn't... beware over-proudness... 05:57
05:58 Sgeo joined 06:05 rgrinberg left 06:08 domidumont joined
sammers sortiz, I have updated the gist to run the promise a few times in a row to highlight the issue a little better. sometimes it works, sometimes it throws 'Use of Nil in string context'. 06:08
sortiz sammers, I'm seeing the HTTP::Tinysh docs and seems that it has specific async support, have you tried that? 06:10
sammers no, I haven't I will take a look.
06:11 domidumont left 06:12 domidumont joined
sammers sortiz, but shouldn't we be able to use promises with blocking calls? in this example I am just using HTTP::Tinyish for experimentation purposes. 06:12
so this might as well be a DB call or disk IO 06:13
sortiz Yep, but In particular that support uses an adhoc FileTempFactory, so seems related to your specific problem with File::Temp. 06:14
sammers hmm 06:15
ok
06:18 CIAvash joined, jjido joined 06:19 z3ndrag0n joined
sammers sortiz, I am unfamiliar with adhoc FileTempFactory. is there some issues wilt File::Temp that are causing this? 06:21
06:23 jjido left, z3ndrag0n left
sortiz Dunno, but the "async" support the author added seems to be mainly about tempfiles, so my best guess is that he found a problem with that. 06:23
06:26 girafe joined 06:27 z3ndrag0n joined 06:29 buharin left 06:31 firstdayonthejob joined 06:32 Actualeyes joined, buharin joined, buharin left 06:41 RabidGravy joined 06:46 firstdayonthejob left 06:50 jjido joined 06:52 sno left, Actualeyes left 06:54 girafe left 07:00 jjido left 07:07 Actualeyes joined 07:13 darutoko joined 07:14 wamba joined 07:15 z3ndrag0n left 07:17 domidumont left 07:27 domidumont joined 07:34 bartolin left 07:35 zakharyas joined 07:40 kent\n is now known as kentnl 07:43 kentnl is now known as kent\n, bartolin joined 07:59 ilogger2 joined, ChanServ sets mode: +v ilogger2 08:08 abraxxa joined 08:17 alnewkirk joined, buharin joined 08:22 z3ndrag0n joined 08:36 holyghost joined
holyghost Hello, it seems that 'panda install SDL' does not build. 08:38
timotimo maybe you can try SDL2::Raw instead
holyghost Thanks that builds, I hun on SDL 1.2 as I do not know SDL 2 08:40
s/hun/run
08:43 rindolf joined 08:47 eugen joined, pdcawley_ joined 08:52 pdcawley_ left
timotimo OK 08:57
09:15 zacts joined 09:26 iH2O joined 09:44 yqt joined 09:54 labster joined 09:55 maybekoo2 joined 09:57 zacts left 09:59 labster left
gfldex timotimo: did you update Pod::To::HTML on docs.perl6.org yet? 10:09
RabidGravy If anyone fancies an amusing little project it's write an emulator for this stuff github.com/chrislgarry/Apollo-11/ 10:10
timotimo did you bump the version of Pod::To::HTML? 10:11
apparently not
well, there's still --force
updated 10:12
gfldex thanks
lizmat RabidGravy: would that involve moon dust ?
10:13 cognominal joined
lizmat touched a piece of moon rock the other day 10:13
nine woah!
lizmat all part of a visit to the Kennedy Space Center :-) 10:14
dalek c: 33b2f80 | (Wenzel P. P. Peppmeyer)++ | / (2 files):
Yellow and black stripes are used by mother nature to indicate stingy

light gray zebra table should play better with other Pod::FormattingCodes when we start to support them in tables.
c: b908f29 | (Wenzel P. P. Peppmeyer)++ | / (2 files):
Table headers where missing padding, making them clash with table data.
c: 8701cfa | (Wenzel P. P. Peppmeyer)++ | / (2 files):
Headings are links to move to the top, giving them the wrong color.
c: 286320e | (Wenzel P. P. Peppmeyer)++ | / (2 files):
give h1 a slight top padding to separate it more from the previous
10:14 zacts joined
RabidGravy lizmat I'm sure it could ;-) I reckon a monstrous Perl 6 version of "Lunar Lander" with the actuall AGC code in it would be a killer application ;-) 10:14
gfldex to get the CSS changes on docs may need a forced reload 10:26
something went wrong with table background, checking 10:28
solved with force reload on all pages :-/ 10:29
timotimo bluh :( 10:33
10:34 yqt left 10:35 clkao joined
gfldex all cache problems will solve themself eventually, right? 10:35
iH2O "Lunar Lander" hmm... No. 10:36
Zoffix Yeah. You can also add, say, ?v=1 query param to the CSS file here to speed things up: github.com/perl6/doc/blob/master/t...ad.html#L5 10:37
gfldex it's purely changed colours, so there is no real need to force it 10:38
do we want a printable version (without redundat auto-gen stuff, etc.) of the docs? 10:42
dalek c: 572dcb4 | Altai-man++ | doc/Language/unicode_texas.pod6:
"Since" column was filled
 10:43
10:50 holyghost left 10:51 holyghost joined 10:55 yqt joined
Zoffix Yeah, since it looks weird on paper. I also add this boilerplate to print sheets for sites at $work though it might be too distracting on the docs site: gist.github.com/zoffixznet/31a2679...9f2e4e6be9 10:55
(adds the URLs for all links into document) 10:56
timotimo oooh
gfldex isn't there a @media type that is respected by browsers on paper?
timotimo also add a piece at the bottom like "please think of the environment before printing this document"
yes, there is
10:56 buharin left
timotimo this css file would be imported only for the paper meida type of course 10:56
10:56 kaare_ joined
gfldex i should be able to use Pod::To::BigPage to make something workable 10:57
Zoffix Yeah, @media print { ... } or you can use media="print" attribute on the <link> element that brings in the print styles 10:59
11:00 huggable joined
gfldex but then the ToC would not have page numbers and the ToC over all .pods is kinda long 11:00
i will see if I can botch a Pod::To:LaTeX together over the weekend
if it's ugly enough some LaTeX wizzard will come along and write a proper style :) 11:01
11:05 gregf_ joined
iH2O well said, gflatex :) 11:07
holyghost So I posted the SDL2 example to CUFP, I will make a game system for it tomorrow 11:09
Most people comprehend rendering but it need e.g. png image loading for games
CUFP == Cool Uses for Perl, perlmonks.org 11:10
11:10 Actualeyes joined
iH2O hmm.. whats the difference between holy ghost and holy spirit 11:12
11:13 kurahaupo joined
gfldex m: enum Foo ( 'Bar' => 1 ); say "{'Bar'}"; 11:16
camelia rakudo-moar ea8c42: OUTPUT«Bar␤»
gfldex :->
11:17 kurahaupo_ joined
gfldex m: enum Foo ( 'Bar' => 1 ); say "{Bar}"; 11:18
camelia rakudo-moar ea8c42: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared name:␤ Bar used at line 1. Did you mean 'Bag'?␤␤»
gfldex that should warn on «'Bar' =>» in my eyes 11:19
11:20 kurahaupo left 11:24 kurahaupo joined 11:26 kurahaupo_ left 11:29 kurahaupo_ joined, kurahaupo left
tbrowder mornin' #perl6 11:33
ref table pod: I'm trying to incrementally add some severe changes as gradually as I can to eliminate problems, so I've added my proposed DebugPod.nqp as a single PR. The current spec table test has been run against it successfully. Please consider PR github.com/rakudo/rakudo/pull/816
11:34 kurahaupo_ left 11:36 freezerburnv joined 11:37 kurahaupo joined
iH2O z.... 11:43
11:46 maybekoo2 left
holyghost I hacked for 8 months on GPL code (off irc), is there a channel on freenode you recommnd a bit for talking ? 11:56
jast about what? :)
holyghost no idea, code or people :-) 11:57
jast well, I *think* there is no super strict off topic policy here
holyghost ok
I got into drinking some pints after code writing 11:58
just because of stress
iH2O join the AAs
holyghost :-)
iH2O :)
jast that seems like, well, not the best possible method to deal with stress... :)
holyghost I really don't want live a bit longerthan just my liver going off course
jast: right 11:59
I don't drink stronger things though
iH2O have inositol, that protects your liver they say
holyghost sure
jast is it the code writing that you get stressed out from? or other stuff?
holyghost mostly not fitting the code where I wanted it 12:00
jast I mainly tend to get angry when writing code ;)
holyghost *lol*
jast usually at the people who invented computers
sometimes at the people who invented the stupid framework/library I have to use
holyghost sure frameworks and so on can be stupid sometimes 12:01
I had this idea not to rely on other code
I am mainly a OSS game hacker, I do sell some stuff too
I am thinking of making "nights into dreams" in Perl 6
jast yeah, sometimes writing it yourself means less overall work 12:02
holyghost that's a sega saturn game
jast depends very much on who wrote the thing you might need, though
12:02 kid51 joined
jast right... I'm not terribly familiar with most console gaming stuff 12:02
holyghost well you get a good emulator and build GNU gcc tools for it
I have a PSVita compiler for it 12:03
jast I don't play a lot of games anyway
holyghost on linux I mean
me neither
jast sure... I just don't care that much though
holyghost I only write them :-)
jast you typically don't install emulators for games you've never really cared about
I did use ScummVM a lot for a while 12:04
holyghost jast : same here
The lucasfilm OSS engine
they made quite some good simmon the sorcerer & monkey island based games
I remember day of the tentacle
dosbox is allright for that too 12:05
Not that I am into infringing games 12:06
90ies games without the internet were ok though
e.g. Westwoord studios, lands of lore, legend of kyrandia series
"apple core dumped" when you dropped an apple 12:07
I really loved the king's quest series starting from number 5
I had this 386 to play it on
jast never played most of those 12:08
holyghost I forgot what scummvm did a bit 12:09
game hacking is cool as you like graphics, once you get into 3D things change though 12:10
lots of math
in 2D you develop much faster, which is logical
jast yeah, I used to try and get into that but that was back when I didn't know nearly enough about math
holyghost sure, to be fair I do not like GL as I do like SGI though 12:11
I gave my O2 to my father
BrokenRobot iH2O: holy spirit is what the holy ghost drinks after a tough day at work ;)
holyghost I need to claim a new one
*lol*
I drink Jupiler, I am in Belgium, so enough choice of beers 12:12
As I said I don't drink high alcohol bevarges
s/bevarges/beverages/
jast I don't like alcoholic anything
holyghost I really like the fact I can talk a bit here 12:13
iH2O Belgium's beers are the best in the world, even better than Germany's
jast Belgium? I'm in Germany, near the drielandenpunt
iH2O ah ah
holyghost jast : I once went ot Eupen
My 2 daughters don't do alcohold either, 15 and 11
I am happy with that 12:14
jast makes sense :)
iH2O but in France the approach is different
holyghost There's this biochemistry you must understand for alcohol
iH2O they teach young children to drink moderate amounts of wine
holyghost ah ok 12:15
iH2O and to keep it moderate
holyghost To be honest, wine is far too expensive for me sometimes
iH2O esp French wine
have California wine instead
much cheaper 12:16
its made by siphonning out the underground water
Gods knows if theres anything left
(underground water)
cuz California is a desert, after all
holyghost understands
desert wine
iH2O well, in a way
holyghost vinyards are ok in such climates 12:17
Main problem is when selling OSS, I really do not want to push far on the credits 12:18
iH2O im in Canada, and thats a day in/day out fight to prevent California from draining clear the Great Lakes
the Great Lakes are the greatest fresh water reserve in the world
holyghost AFAIK Canada is trees and foresters
iH2O we lauch satellites too :-P 12:19
holyghost ok
iH2O *lauch
holyghost *lol*
jast that's the German word for leek :)
12:20 kurahaupo left
holyghost ok, thanks for the talk, 14:20 over here 12:21
I am working on an Inference Engine in perl6
with objects as operator, prefix, postfix, things like that
I'll post it when it's usable to perlmonks.org 12:22
CUFP
12:26 wamba joined 12:33 buharin joined
buharin hi :) 12:33
holyghost hello, what's up 12:34
masak .oO( greater y coordinates )
yoleaux 03:37Z <Zoffix> masak: can this ticket be closed, after adding the tests for current behaviour? It was suggested I ask you: rt.perl.org/Ticket/Display.html?id=70297
holyghost Well, I am off codingfor today 12:35
There's this Gaussian Support Vector Machine I am going to plug into PaganVisions2 12:36
PaganVisions2 is not online yet, but I need to work on this SVM to make a good AI for this game
It's quite clear from the math :-)
PaganVisions is SDL1 but that library is broken in panda AFAIK 12:37
12:37 pmurias joined
gfldex m: printf("%x", 1073741824); 12:38
camelia rakudo-moar 94a31c: OUTPUT«40000000»
holyghost m: printf("%x", 0xff) 12:39
camelia rakudo-moar 94a31c: OUTPUT«ff»
holyghost coolnes
psch m: say 0xff.base(2)
masak Zoffix: looks good to me! feel free to close that one :)
camelia rakudo-moar 94a31c: OUTPUT«11111111␤»
holyghost m: if only it opened X11 windows :-) 12:42
camelia rakudo-moar 94a31c: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Missing block␤at <tmp>:1␤------> 3if only it7⏏5 opened X11 windows :-)␤ expecting any of:␤ new name to be defined␤»
holyghost +1 psch 12:43
iH2O someday, perl6 will have a goto 12:44
:-P 12:45
masak somehow haven't missed `goto` all that much
iH2O i dont like using tons of flags 12:46
re- :-P
masak I mean, people are sometimes overly Dijkstra about it... but I can't think of a case where I went "oh, this program would have been better with `goto`"
12:47 freezerburnv left
masak I do love the `next LABEL` constructs, though 12:47
also, factoring things out into small subs ;)
iH2O bash, not only doesnt have a goto, but you cant exit at all a script from within a command substitution without using real dirty flags 12:48
12:48 zacts left
iH2O i.e. "exit" doesnt work in this case 12:48
holyghost exit 1 12:51
BrokenRobot exit 2
jast goto does make sense in C for error handling 12:52
haven't really ever used it in anything that had suitable alternatives
masak right 12:53
holyghost I used it in NES assembler, cc65.org, cannot think of a beter example than exiting 2 loops 12:54
masak holyghost: but that's what `last LABEL` does
holyghost understands
last as in perl5
last label for as you say 12:55
perlpilot good $localtime #perl6
BrokenRobot \o
12:56 abraxxa left
perlpilot reads the scrollback and sees that masak used "Dijkstra" as an adjective. 12:56
I wonder if he's turning in his grave for this legacy 12:57
iH2O in bash, exit [N] isnt the same as break [N]. exit [N] is only the exit status, not the nesting depth to exit from 12:58
i.e. the respective [N}s refers to completely different things 12:59
perlpilot iH2O: That's why they're spelled different :)
12:59 neilhwatson joined
iH2O maybe my explanation wasnt clear, but the meaning is 13:00
pmurias masak: goto is usefull for emitting code 13:02
yoleaux 6 Jul 2016 16:41Z <jnthn> pmurias: Yes, in NQP - modulo fixups post-compile (I think) - it stays stable. 13:03
6 Jul 2016 16:42Z <jnthn> pmurias: and in the case where it isn't then the thing in $!do is marked as a stub, iirc
13:04 zacts joined 13:13 yqt left 13:14 zakharyas joined 13:17 rgrinberg joined
stmuk why has all the doc been renamed to pod6? 13:18
yoleaux 5 Jul 2016 13:52Z <BrokenRobot> stmuk: This potentially blocks R*, as JSON::PRC is part of Task::Star: github.com/bbkr/jsonrpc/issues/28
13:19 wamba left
BrokenRobot stmuk: because GitHub was attempting to parse.pod as Perl 5 pod and showing "Error cannot parse" or whatever when viewing files 13:20
github.com/perl6/doc/issues/167 13:21
stmuk ah
13:24 skids joined 13:29 AlexDaniel joined 13:31 holyghost left 13:35 loren joined
loren Night ~~ 13:35
ugexe too bad .gitattributes wouldnt work for that
13:36 holyghost joined
loren How can i use a class type as hash key ? 13:37
ugexe what have you tried?
BrokenRobot Aren't hash keys always strings?
Is there a way to represent ^D in a string?
loren I don't want they compare as string
psch m: my %h := :{}; %h{Int} = Int; %h{Num} = Num; .WHAT.say for %h.keys 13:38
camelia rakudo-moar 94a31c: OUTPUT«(Num)␤(Int)␤»
psch m: my %h{Any}; %h{Int} = Int; %h{Num} = Num; .WHAT.say for %h.keys # the other syntax 13:39
camelia rakudo-moar 94a31c: OUTPUT«(Num)␤(Int)␤»
BrokenRobot Ah
Nice
loren m: class A { }; my %h = A.new() => 1; %h{A.new()} = 2; say %h.perl;
camelia rakudo-moar 94a31c: OUTPUT«{"A<51651808>" => 1, "A<51651840>" => 2}␤»
psch loren: you have to initialize the hash as i showed, otherwise you get Stringy keys 13:40
loren m: class A { }; my %h := :{};%h{A.new()} = 1; %h{A.new()} = 2; say %h.perl;
camelia rakudo-moar 94a31c: OUTPUT«:{(A.new) => 1, (A.new) => 2}␤»
holyghost Coolnes 13:41
loren I except result like this : :{ A.new => 2 }
BrokenRobot Specifically, I'm trying to run(...).in.print: "..." a ^D to a program, but I'm unsure how to represent that in that string 13:42
psch loren: but they are different As
loren But these value are equal ..
dalek c: e3efa24 | (Steve Mynott)++ | bin/p6doc:
don't assume extension is .pod since it isn't
psch loren: no, not unless you cheat with the .WHICH 13:43
m: class A { method WHICH { "A" } }; my %h := :{};%h{A.new()} = 1; %h{A.new()} = 2; say %h.perl;
camelia rakudo-moar 94a31c: OUTPUT«:{(A.new) => 2}␤»
psch note, i *strongly* recommend you don't do that
but if you have to, you can do it like that
loren So, i have to overwrite .WHICH ?
psch i personally would recommend asking a question that clarifies the problem you want to solve
ugexe m: class Foo { }; say Foo.new === Foo.new; say Foo.new eqv Foo.new # they are equivalent, not equal 13:44
camelia rakudo-moar 94a31c: OUTPUT«False␤True␤»
psch not a question that asks for help with the solution you already decided on
13:45 grondilu joined
loren I want use a class as Hash key, so i can uniq some data 13:46
psch m: class A { }; my %h := :{};%h{A.new().WHAT} = 1; %h{A.new().WHAT} = 2; say %h.perl; # probably the most obvious way
camelia rakudo-moar 94a31c: OUTPUT«:{(A) => 2}␤»
psch but that way you obviously don't carry whatever A knows around in the Hash
loren Sorry my EN bad ..
BrokenRobot m: say "\c[004]".uninames 13:47
camelia rakudo-moar 94a31c: OUTPUT«(END OF TRANSMISSION)␤»
BrokenRobot tries that
loren class A { has $.value; }
assume class A has a value 13:48
psch m: class A { has $.value }; my @as; @as.push: A.new(:value($_)) for 1,1,2,3,4,4; say @as.unique: :by(*.value) 13:49
camelia rakudo-moar 94a31c: OUTPUT«(A.new(value => 1) A.new(value => 1) A.new(value => 2) A.new(value => 3) A.new(value => 4) A.new(value => 4))␤»
psch ergh
BrokenRobot Doesn't work :/
psch is that again a positional..?
m: class A { has $.value }; my @as; @as.push: A.new(:value($_)) for 1,1,2,3,4,4; say @as.unique: as => *.value
camelia rakudo-moar 94a31c: OUTPUT«(A.new(value => 1) A.new(value => 2) A.new(value => 3) A.new(value => 4))␤»
loren en
psch there
loren: ^^^
loren thanks .. 13:50
psch, BrokenRobot i understand ..
psch well, unless the values you wanted to use in your hash also carry some other significance... 13:51
loren: i think BrokenRobot was talking about something else :)
loren e, ^_^ .. haha don't mind
BrokenRobot Yeah, I was talking about my CTRL+D thing and I think I found the solution: $proc.in.close; does the job 13:52
mst you're trying to get the thing you started to see EOF? 13:53
if so, then, yes, that's how you'd do it in basically any language
BrokenRobot when you run `perl6` it starts REPL and if you press CTRL+D it exits it. I'm trying to emulate doing so programmatically so I could test the output. 13:54
13:55 iH2O left
BrokenRobot And .in.close seems to be doing the trick. 13:55
mst that would be the obvious simple solution, yeah 13:56
13:58 zacts left
[Coke] mst: irclog.perlgeek.de/perl6/2016-07-08#i_12804851 - that's incorrect, yes? 13:59
ugexe if you're using a file handle insetad of the IO::Pipe you could probably do something like nqp::eoffh($proc.in)
mst [Coke]: freenode #perl and irc.perl.org #perl-help for newbie help, then the rest of irc.perl.org for module-specific stuff 14:00
though buharin seems to be successfully annoying #perl, #perl-help, -and- irc.perl.org #perl, depending on the day
BrokenRobot man, I've not been on irc.perl.org/#perl. GumbyBRAIN is probably missing me...
*in ages
mst irc.perl.org #perl is basically "network generic/social channel", which means it's either really busy, or really quiet, or we end up helping people in there because we can't be arsed sending them to #perl-help first 14:01
buharin mst: why annoying?
mst buharin: generally people don't like having to argue with you to get you to actually read the documentation :) 14:02
buharin mst: omg, I just asked
that's cruel 14:03
ugexe there is no documentation, only Allah
buharin mst is so rude 14:04
14:05 mr-fooba_ joined
buharin you should respect your friends 14:05
ugexe I dont think you've seen rude mst
gfldex is it ok if i add a :has() selector hack via javascript to the docs?
mst buharin: I do. that's why I dislike your tendency to waste their time. 14:06
buharin mst: I am yours friend too
mst buharin: you are not my friend. you need to be more respectful of the time of the people helping you before that will happen.
buharin: now this is no longer on topic for #perl6 so please drop it
14:06 khw joined
stmuk "The only requirement is that you know how to be nice to all kinds of people (and butterflies)" 14:07
mst gfldex: I suspect that if, sans javascript, the page still works fine, but is merely slightly less pretty, nobody will have any serious complaints
(by which I mean "if anybody does then complain, I'd be happy to tell them to please stop being silly" :) 14:08
masak pmurias: generated code as a use case for `goto`? interesting.
gfldex mst: the pretty yellow box border would have to go without :has() or a hack
or i rip the thing apart completely, what I would like to avoid 14:09
[Coke] m: sub foo {}; my $x = foo() // 'meow' # RT #128572, zoffix
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128572
camelia ( no output )
pmurias masak: a lot of compilers like using jumps/conditional jumps internally
mst gfldex: what I'm basically saying is "it should still be moderately pretty without the javascript, but if the javascript makes it more pretty, that's cool" - the documentation being essentially static, I think people are allowed to expert it ti *work* without javascript, but that doesn't mean javascript can't be used to make it shinier, if you see what I mean?
BrokenRobot [Coke]: yeah, pmichaud++ explained the error in my ticket. 14:10
m: "\x[1F98B]".uninames.say 14:11
camelia rakudo-moar 94a31c: OUTPUT«(<reserved>)␤»
BrokenRobot aww
gfldex mst: i moved to the ToC as a float to the right of the viewport size is >80em, what works nicely if there is a ToC. If not there is a nice big hole. Most sites with a sidey ToC solve that problem by not having box margins. 14:12
14:12 zacts joined
gfldex s/right/left/ 14:12
mst I'm sure that would've made sense to somebody with any understanding of CSS whatsoever. 14:13
BrokenRobot :)
gfldex let me find a nice friendly wall to talk to :) 14:14
mst basically "remember there are curmudgeons who will turn javascript off, and make sure it's not broken for them" is what I'm trying to say
BrokenRobot What creates that hole?
mst exactly how you make any of that work, I have no idea, I defer to people who have some idea what the heck they're doing :)
gfldex i got a margin on #content for the #ToC to float into. If there is no #ToC, there is still that margin. #content:has(#ToC) would solve that problem. IIRC that selector is proposed for the last 5 years, so there is still hope. 14:15
[Coke] "not broken for them". it can be ugly as hell, though. 14:17
skids m: my $f = "fie foo"; say (qqw|fee $f fum|).perl; # how to get that to be "fee", "fie foo", "fum" ? 14:21
camelia rakudo-moar 94a31c: OUTPUT«("fee", "fie", "foo", "fum")␤»
skids I was actually suprised that "recursed"
BrokenRobot m: my $f = "fie foo"; say (qqww|fee "$f" fum|).perl; 14:22
camelia rakudo-moar 94a31c: OUTPUT«("fee", "fie foo", "fum")␤»
[Coke] docs.perl6.org/language/quoting#Wo...tion:_qqww
"quote protection" is the thing you wanted. 14:23
14:23 cdg joined
[Coke] m: my $f = "fie foo"; say <<fee "$f" fum>>.perl; # also works 14:24
camelia rakudo-moar 94a31c: OUTPUT«("fee", "fie foo", "fum")␤»
psch q is "quote", qq is "quote and interpolate", qqw is "quote, interpolate, and then split on <ws>", qqww is "quote, split on <ws>, then interpolate"
14:24 cdg left
skids Huh. Well, great to know, but considering the nonpresence of any actual quotes, no wonder I didn't think it was what I needed when reading the docs. 14:24
psch at least that's how i remember them
14:25 cdg joined
skids thanks 14:25
psch qqww is probably better as "quote, interpolate, keep nested quotes together"
because that's what it actually does 14:26
psch .oO( we have too many features! )
14:26 pmurias left
skids will maybe improve those docs later 14:26
[Coke] skids: it helps if you imagine it as a literal instead of a variable.
better docs, always good 14:27
skids won't say there are too many features considering he wants to use the one under question :-) 14:29
Oh dear. Now I know why I was confused. 14:32
$ perl6 -e 'sub MAIN ($f) { say (qqww|fee $f fum|).perl }' "fie foo"
("fee", "fie", "foo", "fum")
BrokenRobot gfldex: yeah, makes sense, since we can't move TOC outside of #content here. I don't know what's involved with adding a :has selector, but maybe if ( ! $('#content #TOC).length ) { $('#content').css('margin', ...) } is simpler to implement. 14:33
psch m: my $f = "fie foo"; say qqww|fee '$f' fum|
camelia rakudo-moar 94a31c: OUTPUT«(fee $f fum)␤»
BrokenRobot m: my $f = "fie foo"; say qqww|fee qq/$f/ fum|
camelia rakudo-moar 94a31c: OUTPUT«(fee qq/ fie foo / fum)␤»
BrokenRobot m: my $f = "fie foo"; dd qqww|fee qq/$f/ fum|
camelia rakudo-moar 94a31c: OUTPUT«("fee", "qq/", "fie", "foo", "/", "fum")␤»
BrokenRobot heh
psch m: my $f = "fie foo"; say qqww|fee '\qq[$f]' fum| # /o\ 14:34
camelia rakudo-moar 94a31c: OUTPUT«(fee fie foo fum)␤»
gfldex BrokenRobot: i think i found a way by adding a no-toc class in Pod::To::HTML. Useful to have anyways.
BrokenRobot gfldex++
psch m: my $f = "fie foo"; say qqww|fee \qq[$f] fum|
camelia rakudo-moar 94a31c: OUTPUT«(fee fie foo fum)␤»
psch m: my $f = "fie foo"; say qqww|fee \qq/$f/ fum|
camelia rakudo-moar 94a31c: OUTPUT«(fee fie foo fum)␤»
skids m: my $f = "fie foo"; sub f ($f) { say (qqww|fee $f fum|).perl }; f($f);
camelia rakudo-moar 94a31c: OUTPUT«("fee", "fie", "foo", "fum")␤»
psch maybe "ww means 'respect quotes', so if you don't use any it's equivalent to w"..? 14:36
(minus the usual \qq escape hatch, of course :S )
skids Right so there is no such feature I guess.
psch well, no
the only thing ww gains you over w is that you can quote to keep interpolated things with spaces grouped 14:37
m: my $f = "fie foo"; sub f ($f) { say (qqw|fee "$f" fum|).perl }; f($f);
camelia rakudo-moar 94a31c: OUTPUT«("fee", "\"fie", "foo\"", "fum")␤»
psch w only gets you .words, while ww gets you (the imaginary) .words-but-check-if-it-was-quoted 14:38
14:38 wamba joined 14:39 rgrinberg left, rgrinberg joined 14:43 zacts left, wamba left 14:55 luiz_lha joined, luiz_lha is now known as Guest59918 14:56 Guest59918 is now known as luiz_lha 15:00 huggable left, molaf joined 15:05 zacts joined 15:13 loren left 15:14 zacts_pi joined 15:16 zacts left 15:21 zacts_pi left
dogbert17 .seen womble 15:23
yoleaux I haven't seen womble around.
dogbert17 it's a bit silent around here atm
15:23 zacts_pi joined
BrokenRobot .seen asquare 15:24
yoleaux I haven't seen asquare around.
dogbert17 :)
have a new doc gist, what else, this time for Capture.pairs. Is it correct? gist.github.com/dogbert17/226e7adb...6b2e8e1bc0
holyghost talking to yourslef again ? :-) 15:25
hello btw 15:26
dogbert17 hello holyghost
gfldex dogbert17: s/L<pairs/L<Pairs/ 15:28
dogbert17 gfldex: consider it fixed :)
dalek c: f577b11 | (Wenzel P. P. Peppmeyer)++ | / (2 files):
move ToC to the left if viewport >110em
15:29 itaipu joined
dalek c: 99e2e37 | (Jan-Olof Hendig)++ | doc/Type/Capture.pod6:
Added docs for Capture.pairs. gfldex++
 15:31
15:34 ssotka joined, domidumont joined 15:37 cdg_ joined 15:38 cdg left
dalek c: d92efe7 | (Wenzel P. P. Peppmeyer)++ | / (2 files):
with a leftist ToC send ghosty Camelia into hiding
 15:39
15:45 stmuk_ joined, cdg_ left 15:46 cdg joined
dalek c: ea4191a | (Steve Mynott)++ | bin/p6doc:
revert to working old-locate-curli-module for time being
 15:47
15:48 stmuk_ is now known as stmuk
dalek c: bdf6756 | (Wenzel P. P. Peppmeyer)++ | / (2 files):
The internet seams to agree that leftist ToCs don't need blue links. All

less nervous to me and I hope to anyone else too.		 15:48
stmuk .tell ugexe github.com/perl6/doc/commit/ea4191a2e4 maybe of interest
yoleaux stmuk: I'll pass your message to ugexe.
15:51 loren joined
arnsholt So, Smalltalk has made an interesting trade-off in how it allows user-defined operators: all operators have the same precedence and are evaluated from left to right 15:54
harmil It just occurred to me as I was transcribing some formulas into p6 that, similar to "token" in grammars, "formula" would be a wonderful alias for "sub" if it enforced "is copy" as the default for all parameters. The intuitive semantics of a formula are that they will modify their parameters, but without side-effects. 15:56
ugexe stmuk: ah i see why it broke. but your solution seem just as good though. (fwiw the fix is something like s/{$module}<pm><file>/{$module}.value<file>/)
yoleaux 15:48Z <stmuk> ugexe: github.com/perl6/doc/commit/ea4191a2e4 maybe of interest
15:57 travis-ci joined
travis-ci Doc build failed. Steve Mynott 'revert to working old-locate-curli-module for time being' 15:57
travis-ci.org/perl6/doc/builds/143366707 github.com/perl6/doc/compare/d92ef...4191a2e43b
15:57 travis-ci left
harmil Of course, once we have Lispish macros, that would be something we could write in the language directly. 15:57
stmuk travis-ci sssh
15:58 rgrinberg left
stmuk ugexe: I'm not sure my solution handles multiple versions of the same module 15:58
BrokenRobot trailing whitespace strikes again :P
15:58 darutoko joined
stmuk ffs 15:59
[Coke] make test is pretty quick. Running it before pushing is a good plan.
(even if we remove the trailing ws test.)
16:02 canopus joined
ugexe stmuk: ah, yeah it will give you whatever it finds first (but thats all mine did too). Just needs an additional :$version to pass as :$version-matcher to CU::DS I believe 16:02
16:02 domidumont left
dalek c: ff9cfd0 | (Jan-Olof Hendig)++ | doc/Type/Capture.pod6:
Added code examples to methods list and elems
 16:02
16:03 setty1 joined 16:04 zakharyas left
dalek c: 4109e90 | (Steve Mynott)++ | bin/p6doc:
delete one space to shut up pointless test
 16:05
stmuk passive agressive moi? :P 16:06
BrokenRobot It's not aggressive if the statement is true :P
16:07 itaipu left 16:13 travis-ci joined
travis-ci Doc build failed. Jan-Olof Hendig 'Added code examples to methods list and elems' 16:14
travis-ci.org/perl6/doc/builds/143370637 github.com/perl6/doc/compare/bdf67...9cfd041ddc
16:14 travis-ci left, Herby_ joined
Herby_ o/ 16:14
BrokenRobot \o
16:16 kurahaupo joined, zacts_pi left
pmichaud good morning #perl6 16:18
Herby_ \o/
BrokenRobot It's morning already!? :P
tbrowder mornin' pmichaud 16:19
ref table pod handling: next set of changes, builds okay, table tests pass okay, please see PR github.com/rakudo/rakudo/pull/817
pmichaud It will stop being morning in about 41 minutes.
kurahaupo is never quite sure whether the period between midnight and dawn counts as night or morning. 16:20
BrokenRobot It depends on how long ago you woke up :) 16:21
kurahaupo suspects the answer is "yes"
BrokenRobot: about 18 hours
BrokenRobot Then it's night :) 16:22
kurahaupo and when I wake up again in about 3h45 (when it'll still be dark)?
BrokenRobot It'll be morning :) 16:23
16:27 domidumont joined 16:30 loren left, kurahaupo_ joined
holyghost is a bit drunk 16:31
I really need to stop the beer-sugar-coding
BrokenRobot Ah, you remind me of a yonger version of me :)
holyghost well, I really need to learn to buy a lemonade instead of beer 16:32
BrokenRobot Meh, coffee is where it's at ;)
holyghost right
coffee is good indeed 16:33
some dude must have changed those beans somehow
too good to be true
16:34 kurahaupo left 16:35 sortiz joined 16:36 yqt joined
BrokenRobot Well... conclusive proof that fewer cores, less RAM, and non-SSD drives are better: gist.github.com/zoffixznet/cb36f42...79ff7c71aa 16:37
gfldex BrokenRobot: setting the number of parallel jobs to the number of DDR channels may provide the best result 16:38
BrokenRobot And it looks like S10-packages/precompilation.t is the the bottle neck test. The rest of the spectest finishes before the last 5 tests in that file finish 16:40
ugexe probably lock files then 16:41
BrokenRobot And by "bottle neck" I mean like 6-10 seconds :)
16:47 rgrinberg joined 16:58 kurahaupo_ left 16:59 zacts joined 17:06 hankache joined
BrokenRobot .ask tbrowder can this be closed? You can modify the roast checkout in t/spec to make the PR for tests toort.perl.org/Ticket/Display.html?id=128519 17:09
yoleaux BrokenRobot: I'll pass your message to tbrowder.
17:11 iH2O joined 17:13 lostinfog joined 17:16 firstdayonthejob joined
BrokenRobot .ask DrForr can this be closed? It's mentioned to be a dupe of a recently-fixed ticket. I tried to test myself, but Linenoise is failing. rt.perl.org/Ticket/Display.html?id=128371 17:24
yoleaux BrokenRobot: I'll pass your message to DrForr.
17:27 holyghost left, acrussell joined 17:28 holyghost joined 17:31 neilhwatson left, rgrinberg left 17:32 neilhwatson joined 17:36 rgrinberg joined 17:40 iH2O left 17:46 yqt left
holyghost BrokenRobot, the coffee helped, I feel better now 17:48
BrokenRobot Great.
17:50 girafe joined
nemo slt 17:51
17:51 hankache left 17:52 acrussell left 17:55 yqt joined
[Coke] drinks a Coke Zero 17:55
timotimo don't dereference the zero, coke
17:56 xtreak joined
harmil It's okay, it generates a lazy exception. *burp* 17:56
xtreak What is the use of adverb syntax in hashes and other places?
timotimo it's shorter than using => 17:57
BrokenRobot :$foo is way shorter than foo => $foo ;) 17:58
psch 'adverb syntax' generally is a shorthand for Pair construction, which automatically slips into argument lists 17:59
...as a more fundamental explanation
xtreak Sorry was asking about :kv, :delete on hashes. I don't know that colonpair syntax was adverb also.
timotimo oh 18:00
gfldex m: sub f(:$foo){ given :$foo { when .value > 10 { say 'BIG' }; default { say 'SMALL' } } }; f(:foo(11));
camelia rakudo-moar 94a31c: OUTPUT«BIG␤»
psch i think those are parser special cases..? 18:01
geekosaur yes, :delete is short for :delete(True) (which means you can use a variable there!)
there's something where it becomes a parameter to the "closest" method call, remembering that things like subscripts are postcircumfix methods
psch right, that's the specific 'adverb' case for colonpairs 18:02
i mean, for postcircumfix
18:02 espadrine joined
holyghost gfldex, looks like CLOS 18:05
Common Lisp Object System 18:06
18:10 zacts left
gfldex m: my $bar = 10; my %h = :$bar; dd %h 18:11
camelia rakudo-moar 94a31c: OUTPUT«Hash %h = {:bar(10)}␤»
18:12 yqt left
xtreak ircbrowse.net/browse/haskell looks interesting with searchable logs and code examples 18:13
dalek c: 9567745 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Pair.pod6:
be explicit about the power of :$bar
 18:15
18:16 espadrine left
dalek c: d1fee79 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Pair.pod6:
hide index entries and add :$
 18:20
18:21 domidumont left
tbrowder BrokenRobot: yoleaux: You may close RT #128519. 18:23
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128519
yoleaux 17:09Z <BrokenRobot> tbrowder: can this be closed? You can modify the roast checkout in t/spec to make the PR for tests toort.perl.org/Ticket/Display.html?id=128519
BrokenRobot Closed. Thanks. 18:26
18:26 itaipu joined
dalek c: fe89ae2 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Signature.pod6:
link from /type/Signature to /type/Pair
 18:29
18:30 pochi joined, cdg left 18:32 rgrinberg left, rgrinberg joined 18:34 noganex joined 18:36 sno joined 18:39 dupek joined 18:40 dupek is now known as sucker
sucker hello 18:40
[Coke] hallo
18:42 TEttinger joined 18:48 itaipu left
gfldex m: my package P:ver<1.2.3>:auth<me> {}; say [ P.^ver(), P.^auth ]; 18:50
camelia rakudo-moar 94a31c: OUTPUT«Method 'ver' not found for invocant of class 'Perl6::Metamodel::PackageHOW'␤ in block <unit> at <tmp> line 1␤␤»
gfldex m: my package P:ver<1.2.3>:auth<me> {};
camelia ( no output )
gfldex m: my class C:ver<1.2.3>:auth<me> {}; say [ C.^ver(), C.^auth ];
camelia rakudo-moar 94a31c: OUTPUT«[v1.2.3 me]␤»
gfldex is there a ticked for that?
18:53 cdg joined 18:54 cdg left, FROGGS joined, cdg joined
gfldex m: my module M:ver<1.2.3>:auth<me> {}; say [ M.^ver(), M.^auth ]; 18:55
camelia rakudo-moar 94a31c: OUTPUT«[v1.2.3 me]␤»
18:55 zacts joined
BrokenRobot m: my module P:ver<1.2.3>:auth<me> {}; say [ P.^ver(), P.^auth ]; 18:57
camelia rakudo-moar 94a31c: OUTPUT«[v1.2.3 me]␤»
18:59 cdg left 19:00 huggable joined
BrokenRobot Seems an easy 1-line fix. 19:00
gfldex BrokenRobot: please fix if you can, I already doced that. :) 19:01
BrokenRobot Sure.
gfldex also there is no spectest for $package.^ver|auth 19:02
also for module, class is there tho
19:04 pmurias joined, pmurias left 19:06 pmurias joined
pmurias morepypy.blogspot.ca/2016/07/rever...ython.html - awesome pypy feature that hopefully we will steal someday 19:07
19:07 labster joined 19:08 itaipu joined 19:11 xtreak left
dalek c: d1ed93b | skids++ | doc/Language/quoting.pod6:
Document order of operations in qqw and qqww
 19:20
[Coke] .seen colomon 19:24
yoleaux I saw colomon 10 Jun 2016 01:44Z in #perl6: <colomon> and the optimizations are built on the proxy level
colomon o/
moritz what a prompt reply after three weeks of silence! \o/ colomon
[Coke] colomon: hey, can I kill this: github.com/rakudo/rakudo/wiki/NG-issues 19:25
Looks like it might be old. :)
colomon moritz: irc is almost always open on my MBP
[Coke]: it’s a bit old, perhaps. ;) 19:26
19:26 addison joined
colomon if it were a kid, it would be ready for kindergarten 19:26
19:26 wamba joined
[Coke] killing. 19:26
colomon [Coke]++
moritz: though I have been insanely busy, alas. 19:27
BrokenRobot m: my package P:ver<1.2.3>:auth<me> {}; say [ P.^ver(), P.^auth ]; 19:28
camelia rakudo-moar 94a31c: OUTPUT«Method 'ver' not found for invocant of class 'Perl6::Metamodel::PackageHOW'␤ in block <unit> at <tmp> line 1␤␤»
BrokenRobot RT: rt.perl.org/Ticket/Display.html?id=128579
PR: github.com/rakudo/rakudo/pull/818
Tests: github.com/perl6/roast/pull/134y
19:29 maybekoo2 joined 19:30 darutoko left
BrokenRobot I meant github.com/perl6/roast/pull/134 19:32
moritz looks legit. Merged 19:33
at least the rakudo one; you can merge the roast one yourself, I think
BrokenRobot \o/ moritz++
moritz and reap the karma
BrokenRobot++
moritz experiments with cron-apt for unattended security upgrades 19:35
if it works fine, I'll do that on the *.p6c.org hosts
dalek c: 44a60a7 | (Jan-Olof Hendig)++ | doc/Type/Capture.pod6:
Added docs for Capture.antipairs
19:40 itaipu left 19:42 domidumont joined 19:43 gcole_ joined
dalek c: 845d9dd | (Wenzel P. P. Peppmeyer)++ | doc/Language/typesystem.pod6:
doc .^ver and .^auth for class, role, grammar, module and package
 19:45
19:48 yqt joined 19:50 holyghost left 19:51 ssotka left
dogbert17 m: my $capture = \(2, 3, apples => 2); say $capture.Str; # hmm, is this correct behavior? 19:54
camelia rakudo-moar ae36a1: OUTPUT«This representation (VMIter) cannot unbox to a native string (for type BOOTIter)␤ in block <unit> at <tmp> line 1␤␤»
[Coke] Based on the discussion on RT #128520, started updating github.com/rakudo/rakudo/wiki with notes about RT.
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128520
[Coke] Please feel free to add placeholders if there's a topic you want covered, or feel free to ping me. 19:55
I just did some cleanup, will do more, and will plan to add more content.
moritz dogbert17: looks like a bug to me
gfldex m: my $capture = (2, 3, apples => 2).Capture; say $capture.perl;
camelia rakudo-moar ae36a1: OUTPUT«\(2, 3, :apples(2))␤»
gfldex dogbert17: the .Str problem is a bug 19:56
dogbert17 uh oh
psch not really sure what .Str means on a Capture though 19:57
gfldex .perl basicly
psch i mean, yeah, VMIter throwage is definitely LTA
gfldex Capture is the data structure that holds parameter lists (among other things), turning that into a Str sounds like introspection to me. So forwarding to .perl makes sense. 19:59
psch m: my %h = :1a, :2b; say %h.Str
camelia rakudo-moar ae36a1: OUTPUT«a 1␤b 2␤»
psch i'm not convinced, but i don't have a better idea 20:00
dogbert17 at least .gist calls .perl while .Str has a different implementation (if I read the code correctly)
gfldex dogbert17: i can't find any RT for that. Please rakudobug. 20:01
dogbert17 gfldex: I will try, third time might be a charm (all my previous mails have failed to arrive for some reason) 20:03
gfldex dogbert17: if it fails again, feel free to poke me
psch $ ./perl6-m -e'(1, 2, :2c).Capture.Str.say' 20:05
1 2 c 2
that's what the current code intented to afaict
basically, call .Str on all arguments in the Capture and concat with spaces
20:06 addison left, domidumont left
[Coke] any point in saving this: github.com/rakudo/rakudo/wiki/Real-World-Usage ? 20:07
gfldex m: my $l = (1,2, :2c); $l.Str.put;
camelia rakudo-moar ae36a1: OUTPUT«1 2 c 2␤»
perlpilot [Coke]: I don't think so.
gfldex m: my $l = (1,2, :2c); $l.perl.put;
camelia rakudo-moar ae36a1: OUTPUT«$(1, 2, :c(2))␤»
dogbert17 gfldex: will do
psch m: my $l = \(1,2, :2c); $l.perl.put; 20:08
camelia rakudo-moar ae36a1: OUTPUT«\(1, 2, :c(2))␤»
[Coke] gone.
gfldex psch: i take my introspection expection back. A listy thing should behause like List.
psch the Capture.perl is really about \() - because .perl wants to try and get something .EVAL-able
gfldex: well, a Capture is, depending on contained arguments, listy or hashy or both 20:09
that's the one thing i'd be a bit hesitant about, fwiw
Hash.Str adds \n per kv
20:09 pmurias left
psch m: say %(:1a, :2b, :3c).Str 20:09
camelia rakudo-moar ae36a1: OUTPUT«a 1␤b 2␤c 3␤»
psch m: say <a b c>.Str 20:10
camelia rakudo-moar ae36a1: OUTPUT«a b c␤»
gfldex m: my %h = <a b c> Z=> 1,2,3; %h.Str.say;
camelia rakudo-moar ae36a1: OUTPUT«a 1␤b 2␤c 3␤»
psch but mixing that in Capture.Str would probably be weird
gfldex m: my %h = 1,2,3 Z=> «a "\n" c»; %h.Str.say; dd %h; 20:12
camelia rakudo-moar ae36a1: OUTPUT«1 a␤2 ␤␤3 c␤Hash %h = {"1" => "a", "2" => "\n", "3" => "c"}␤»
gfldex blindly calling .Str seams not to make much sense in any case. 20:13
psch well, it's just about "make it fit into a Str" 20:14
it's explicitely not intended to carry programmatically semantic meaning
...i'm probably not really intended to talk complicated anymore :P 20:15
anyway, i'll push the Capture.Str fix as-is
can still be ticketed if we want anything more useful there
20:16 efdetonator joined 20:17 efdetonator left
hoelzro Zoffix: are you still seeing Linenoise build failures? it works for me on both zef and panda 20:19
[Coke] m: (:A<a> :B<b>) # just LTA, or should this do *something* ? 20:26
camelia rakudo-moar ae36a1: OUTPUT«===SORRY!===␤Unknown QAST node type NQPMu␤»
psch dimly remembers something about colonpairs not needing commas
i don't know if that got tossed though
AlexDaniel [Coke]: how can I change my default queue? 20:28
20:28 neilhwatson left 20:29 rindolf left
FROGGS m: say (:A<a> :B<b>) 20:30
camelia rakudo-moar ae36a1: OUTPUT«===SORRY!===␤Unknown QAST node type NQPMu␤»
[Coke] AlexDaniel: rt.perl.org/Prefs/Other.html - first choice 20:31
AlexDaniel [Coke]: ah shit, I am blind 20:32
[Coke] tbrowder: See that page for how it displays your username.
20:33 itaipu joined
[Coke] ah, as am I, as I now see where "new ticket in" is. :) 20:34
for the longest time, you *couldn't* new ticket that way!
20:34 kaare_ left 20:36 rgrinberg left 20:39 rindolf joined, gcole_ left, djbkd joined 20:42 addison joined 20:43 djbkd left
AlexDaniel [Coke]: can't edit my comment, but don't want to write another one, so I'll just mention it here: “I believe as is, no, it's plain text.” – nope, see this for example: rt.perl.org/Ticket/Display.html?id=128531 20:45
[Coke]: it has colors, headings, links and so on
20:46 _dolmen_ joined
dalek c: 2196187 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Mu.pod6:
doc Mu.return and Mu.return-rw
 20:46
[Coke] AlexDaniel: that's a text/html attachment.
sure, you can email in attachments.
AlexDaniel [Coke]: oh, indeed! Proving again that I am blind
[Coke] you can't edit them in the web ui
no worries. 20:47
20:47 addison left
[Coke] my comment was just about the web text inputs, but I wasn't clear. 20:47
AlexDaniel [Coke]: hm, but you can still use > for quotes
[Coke] probably from the email heritage. 20:48
timotimo psch: we still have some bugs when you use colon pairs without commas, like you'd get a 1 instead of True 20:50
pmichaud Good afternoon, #perl6 20:53
moritz \o pmichaud
lizmat pmichaud o/ 20:54
dalek c: f49f74e | (Wenzel P. P. Peppmeyer)++ | doc/Language/operators.pod6:
add ... and … to index
 20:55
c: b5e4d86 | (Wenzel P. P. Peppmeyer)++ | doc/Language/operators.pod6:
add lazy list reference for … to index
c: da63e22 | (Wenzel P. P. Peppmeyer)++ | doc/Language/control.pod6:
add lazy list index reference for gather and take
pmichaud Only here for a short bit today... workers behind my house have apparently servered my landline Internet connection and repairs can't be made until Monday at the earliest. 20:56
So I'm stuck with my mobile wifi hotspot.
20:56 itaipu left
[Coke] zoinks. 20:56
pmichaud: rt.perl.org/Ticket/Display.html?id=128423 if you haven't seen that. (https issue on rakudo.org) 20:58
moritz pmichaud: that sucks
[Coke] let us know if we pass the work to someone. ISTR you had seen it
lizmat pmichaud: also, core devs generally hang out on #perl6-dev nowadays :-)
pmichaud [Coke]: I've seen it. I've tried to get the certificate fixed before but I'm not quite sure how to do that. 20:59
Also, I've been told (by my cpanel instance) that I've got to do a big migration to a new server anyway for all of my domains on that host, so I'm likely to resolve the certificate issue as part of that work. 21:00
I still know that I need to upgrade the WordPress instance, and will do that this weekend if the internet connectivity doesn't end up being too onerous. 21:01
21:01 skids left 21:02 addison joined
AlexDaniel m: ++ 21:03
camelia rakudo-moar 574519: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Prefix ++ requires an argument, but no valid term found␤at <tmp>:1␤------> 3++7⏏5<EOL>␤ expecting any of:␤ prefix␤»
AlexDaniel expecting any of: prefix
huggable: dunno
huggable AlexDaniel, ¯\_(ツ)_/¯
moritz pmichaud: since hack.p6c.org occasionally freezes, there's now an irc.p6c.org which should be more stable, and which you can use for your IRC session 21:05
pmichaud: same password and ssh key as hack
pmichaud good to know -- switching.
AlexDaniel m: say Any + 1
camelia rakudo-moar 574519: OUTPUT«Use of uninitialized value of type Any in numeric context in block <unit> at <tmp> line 1␤1␤»
AlexDaniel m: my $x = Any; say $x + 1 21:06
camelia rakudo-moar 574519: OUTPUT«Use of uninitialized value of type Any in numeric context in block <unit> at <tmp> line 1␤1␤»
AlexDaniel m: my $x = Any; say $x++
camelia rakudo-moar 574519: OUTPUT«0␤»
AlexDaniel m: my $x = Any; $x *= 5; say $x
camelia rakudo-moar 574519: OUTPUT«5␤»
AlexDaniel OK, but why? 21:07
+= acts like if it was 0, but *= acts like if it is 1 21:08
psch m: say [+]; say [*]
camelia rakudo-moar 574519: OUTPUT«0␤1␤»
pmichaud AlexDaniel: S03:435 has a partial answer
synopsebot6 Link: design.perl6.org/S03.html#435_has_a...ial_answer
dalek c: 03c49a9 | (Wenzel P. P. Peppmeyer)++ | doc/Language/list.pod6:
doc lazy lists
 21:09
pmichaud (synopsebot6 gives a weird anchor link there.)
moritz S03:435
synopsebot6 Link: design.perl6.org/S03.html#line_435
moritz better
pmichaud I'm curious why synopsebot6 chose to add my text though
21:09 FROGGS left
psch S03:it adds everything unfortunately 21:09
synopsebot6 Link: design.perl6.org/S03.html#it_adds_e...ortunately
pmichaud Also S03:4151 21:10
synopsebot6 Link: design.perl6.org/S03.html#line_4151
pmichaud which really answers the question :)
moritz pmichaud: it's an instance of producing heading anchors, like S02:Literals
synopsebot6 Link: design.perl6.org/S02.html#Literals
moritz just overeager
time for sleep here
psch well, it's hard to guess well there
either you ask for _ from the user or you check what anchors exist on the side
*site
pmichaud moritz: okay, that makes more sense. Perhaps it could recognize starting digits as meaning a line number instead of anchor target
moritz sounds like a good heuristic 21:11
pmichaud since I don't think we have anchor targets that start with digits.
21:11 xtreak joined 21:13 sucker left 21:14 itaipu joined, setty1 left
pmichaud [Coke]: a comment on RT #128520 (RT versus other ticketing systems) -- I disagree with the remark that lack of progress on old tickets has little to do with the ticketing system used. For a long period of time I found my frustration with working with RT to be a serious impediment to working on bug reports. 21:14
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128520
pmichaud to the point that I just didn't want to do them.
I expressed this to masak++ and others at one of our gatherings (YAPC::NA 2013, I think) as being a very real issue for me. 21:15
dalek c: e402f7b | (Wenzel P. P. Peppmeyer)++ | / (2 files):
Shady Camelia is also getting in the way when the ToC is hidden. As we

  shady Camelia has to go.		 21:16
Zoffix A more recent example of that: irclog.perlgeek.de/perl6/2016-07-04#i_12781927
AlexDaniel [Coke]: it seems like it is possible to turn on wysiwyg editor on RT 21:17
mspo I'd like to explain gather take on a whatever to a nine y/o 21:18
see how it goes
pmichaud my nine year olds would have no problem with it.
AlexDaniel I feel a bit sad to see shady camelia go… :(
the decision to remove it is probably right though 21:19
pmichaud can I see an example of shady camelia?
gfldex pmichaud: see github.com/perl6/doc/issues/681 21:20
AlexDaniel pmichaud: files.progarm.org/2016-07-09-00200..._scrot.png
Zoffix gfldex, FYI. It'd need to be removed from docs too: github.com/rakudo/rakudo/pull/818#...-231468771
psch that looks really bad with the lines
AlexDaniel yeah…
pmichaud just make her smaller, I think. 21:21
mspo pmichaud: my 3 y/o might
gfldex Zoffix: on it 21:22
21:22 travis-ci joined
travis-ci Doc build errored. Wenzel P. P. Peppmeyer 'doc lazy lists' 21:22
travis-ci.org/perl6/doc/builds/143442335 github.com/perl6/doc/compare/da63e...c49a92f677
21:22 travis-ci left
pmichaud anyway, I'm afk again to preserve mobile bandwidth. 21:23
21:23 travis-ci joined
travis-ci Doc build errored. Wenzel P. P. Peppmeyer 'Shady Camelia is also getting in the way when the ToC is hidden. As we 21:23
travis-ci.org/perl6/doc/builds/143443847 github.com/perl6/doc/compare/03c49...02f7b04abb
21:23 travis-ci left
pmichaud oh, wait, switching hosts first 21:23
AlexDaniel what about this? files.progarm.org/2016-07-09-00232..._scrot.png
dalek c: b42b053 | (Wenzel P. P. Peppmeyer)++ | doc/Language/typesystem.pod6:
remove .^auth and .^ver from package as the PR to roast was rejected
 21:24
AlexDaniel one camelia on top left is probably enough… 21:25
21:26 rgrinberg joined
gfldex the problem with combining zebra tables and background images could be solved tho 21:26
but then the problem of black text on dark grey background would still be there
21:27 pmichaud joined
pmichaud AlexDaniel: I like the smaller Camelia much better (than the original, even) 21:28
21:29 RabidGravy joined
pmichaud really afk this time 21:29
21:31 xtreak left, travis-ci joined
travis-ci Doc build errored. Wenzel P. P. Peppmeyer 'remove .^auth and .^ver from package as the PR to roast was rejected' 21:31
travis-ci.org/perl6/doc/builds/143445444 github.com/perl6/doc/compare/e402f...2b05375855
21:31 travis-ci left 21:32 Actualeyes left
AlexDaniel gfldex: maybe we can try smaller camelia? What do you think? 21:38
grondilu wonders why nqp ops are faster than pure Perl 6 and thus why lizmat and others put so many of them in src/core 21:41
I mean for instance why can't Perl 6's "if" be just as fast as nqp::if? 21:42
21:42 freezerburnv joined
lizmat grondilu: because if foo { } creates a scope, and nqp::if doesn't 21:43
psch Perl 6 level if attaches to a Block, yeah
which gives you e.g. the capacity for ENTER and LEAVE phasers
lizmat also, a whole block of nqp::if counts as a single statement 21:44
psch also scoping as lizmat says
lizmat so error reporting sucks if something goes wrong in there
grondilu ideally shouldn't the optimizer be able to downgrade a if statement to "whatever nqp::if" does if it's possible?
lizmat hmmm... it appears that tbrowders last changes borked a lot of spectest files
grondilu: ideally, yes 21:45
psch grondilu: a sufficiently smart optimizer would, yes :)
grondilu ok, so that means that in the long term it will. I'm satisfied with that.
lizmat indeed...
but we need a faster rakudo *now* 21:46
:-)
grondilu I'm patient, but I was worrying that the strategy to improve performance was just to rewrite everything in nqp :/
psch it's an intermediate strategy, and a very good for that
lizmat++
s/a / 21:47
timotimo agreed
21:47 Actualeyes joined
Zoffix "Cannot invoke this object (REPR: P6opaque; NQPMu)" 21:48
timotimo that's when you have misspelt something in nqp code
lizmat yeah, checking now if that really was the commit
psch m: use nqp; nqp::nosuchop() 21:50
camelia rakudo-moar 58c262: OUTPUT«===SORRY!===␤No registered operation handler for 'nosuchop'␤»
psch m: use nqp; foo()
camelia rakudo-moar 58c262: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared routine:␤ foo used at line 1␤␤»
psch nqp-m: foo()
camelia nqp-moarvm: OUTPUT«Cannot invoke this object (REPR: Null; VMNull)␤ at <tmp>:1 (<ephemeral file>:<mainline>)␤ from gen/moar/stage2/NQPHLL.nqp:1505 (/home/camelia/rakudo-m-inst-2/share/nqp/lib/NQPHLL.moarvm:eval)␤ from gen/moar/stage2/NQPHLL.nqp:1708 (/home/camelia/rakudo-m…»
tbrowder [Coke] I will look later (for some reason I can't log in on my iPad
lizmat tbrowder: I reverted your commit 21:52
21:52 kurahaupo joined
psch really non-obvious what in there might break something... 21:54
21:54 itaipu left
tbrowder okay, i don't see that yet--is there a problem or recommendation? 21:54
21:54 zacts left
psch tbrowder: i'd assume building with your patch and running spect 21:55
tbrowder: then checking if anything breaks, then checking *why* it breaks
well, that's my recommendation i suppose :) 21:56
21:59 freezerburnv left 22:01 rindolf left
lizmat tbrowder: this is a stacktrace of the failure I saw: gist.github.com/lizmat/7bf37d24e9d...3fb892c217 22:01
pochi m: my Array[Numerix] $x
camelia rakudo-moar 897c22: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤An exception occurred while parameterizing Array␤at <tmp>:1␤Exception details:␤ 5===SORRY!5=== Error while compiling ␤ Cannot invoke this object (REPR: Null; VMNull)␤ at :␤»
pochi It would be nice if it told me it couldn't find the type Numerix 22:02
lizmat pochi: please RT is by sending a bug report to [email@hidden.address] 22:03
*it
pochi ok
lizmat pochi: thanks!
tbrowder psch: I did do the table tests but no other, didn't think it would affect anything else. 22:04
psch tbrowder: well, aparently you were wrong. :) lizmat mentioned affected tickets in the revert
lizmat psch: affected *files* 22:05
psch ...right
psch blames the time
lizmat tbrowder: please always do a full spectest :-)
if you set TEST_JOBS=8 on an i7, it takes around 4 minutes nowadays 22:06
22:08 Evan joined, Evan is now known as Guest64213, lostinfog left
Guest64213 what are the equivalent of perl 5's system(), $!, and $? in perl6? 22:09
timotimo m: say $ 22:10
camelia rakudo-moar 897c22: OUTPUT«(Any)␤»
timotimo sorry
m: say $!
camelia rakudo-moar 897c22: OUTPUT«Nil␤»
timotimo m: say $?
camelia rakudo-moar 897c22: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unsupported use of $? variable; in Perl 6 please use $! for handling child errors also␤at <tmp>:1␤------> 3say $?7⏏5<EOL>␤»
psch m: say $*EXECUTABLE
camelia rakudo-moar 897c22: OUTPUT«"./rakudo-m-inst/bin/perl6-m".IO␤»
psch $! is still around 22:11
timotimo there's system, run, Proc::Async, and qx
Guest64213 ok, cool.
psch m: try 2 / 0; say $!.perl
camelia rakudo-moar f583f2: OUTPUT«Any␤»
timotimo bad example :) 22:12
psch eh, Rats
Guest64213 what is the best way to find documentation of built-in functions?
psch m: try 2 div 0; say $!.perl
camelia rakudo-moar f583f2: OUTPUT«X::Numeric::DivideByZero.new(using => "div", numerator => 2)␤»
Guest64213 I was looking at tablets.perl6.org/appendix-a-index.html
timotimo oh, the tablets?
Guest64213 but it doesn't mention system
Zoffix Guest64213, docs.perl6.org/
timotimo i wonder how you found them
Zoffix Guest64213, there's a page on suboruines list: docs.perl6.org/routine.html
22:12 TakinOver joined
Guest64213 tablets are linked from perl6.org/documentation/ 22:12
timotimo oh
Zoffix generally just searches
timotimo maybe the perl6 documentation ought to get a big button 22:13
Zoffix And other stuff moved to ... I dunno, some "Historical" section?
Speculations listed 2nd after document is LTA. And that Periodic Table is probably outta date too, I recall. Isn't it?
psch i skimmed it and didn't see anything wrong 22:14
timotimo yeah, it's a bit outdated
psch although precedence is probably underspecific in at least a few cases
22:14 _dolmen_ left
psch i'm thinking specifically of method invocation 22:14
Zoffix m: my @a = ^10; say @@a[^3] 22:16
camelia rakudo-moar f583f2: OUTPUT«(0 1 2)␤»
Zoffix m: my @a = ^10; say @a[^3]
camelia rakudo-moar f583f2: OUTPUT«(0 1 2)␤»
psch m: my $x = 1; $x ,= 2; say $x.perl # uhh
camelia rakudo-moar f583f2: OUTPUT«(my \List_81150664 = $(List_81150664, 2))␤»
Zoffix What's "@@ slice context"?
timotimo like LoL?
psch oh, yeah, that's definitely gone
22:16 RabidGravy left
psch i think that was supposed to be non-flattening list coercion 22:17
Zoffix Ah
m: my $x ::= 42; BEGIN say $x 22:18
camelia rakudo-moar f583f2: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤"::=" not yet implemented. Sorry. ␤at <tmp>:1␤------> 3my $x ::= 427⏏5; BEGIN say $x␤»
psch m: my $x = 1; $x [R,]= $x ,= 2; say $x
camelia rakudo-moar f583f2: OUTPUT«(\List_64149696 = (List_64149696 List_64149696))␤»
22:18 Sgeo joined
psch somehow that's terribly entertaining to me 22:18
Zoffix :) 22:19
22:20 _dolmen_ joined 22:23 eugen left 22:29 kurahaupo left 22:35 jack_rabbit joined 22:40 cdg joined 22:42 gtodd joined 22:48 djbkd joined
harmil Closed that <$0> bug... man, I wish someone had told me earlier that plain $0 was the solution. 22:53
22:58 grondilu left 23:00 huggable left 23:02 firstdayonthejob left
pochi m: sub x(*@x) { say @x.perl }; x( [1,2,3], [4,5,6] ) 23:10
camelia rakudo-moar f583f2: OUTPUT«[1, 2, 3, 4, 5, 6]␤»
pochi is there a slurpy signature that will give me this result: [ [1,2,3], [4,5,6] ]? 23:11
sortiz m: sub x(**@x) { say @x.perl }; x( [1,2,3], [4,5,6] ) 23:12
camelia rakudo-moar f583f2: OUTPUT«[[1, 2, 3], [4, 5, 6]]␤»
pochi oh, double star
does it have a name?
sortiz I use "Non-flatten slurpy" 23:14
23:16 yqt left 23:17 killbill joined 23:26 BenGoldberg joined, kid51 left 23:30 _dolmen_ left 23:37 sortiz left, Zoffix joined 23:50 z3ndrag0n left
pochi m: my Array[Int] @a; say @a[0]; @a[0] = Array[Int].new; say @a[0]; @a[0][0] = 4 23:50
camelia rakudo-moar f583f2: OUTPUT«(Array[Int])␤[]␤»
23:55 maybekoo2 left

Powered by Cro and the Raku® Programming Language.

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

Thank you!