»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.
00:04 Derperperd joined
AlexDaniel is aaronsherman here? 00:13
00:17 pierre_ joined
AlexDaniel ah nvm, just realized that I actually have a reddit account… 00:21
00:24 lizmat left 00:25 Actualeyes joined
seatek where can i submit grammatical corrections to the docs.perl6.org stuff? should I just point them out here? just ran across one. 00:41
00:47 aborazmeh joined, aborazmeh left, aborazmeh joined
MasterDuke seatek: a PR here would be much appreciated: github.com/perl6/doc 00:52
00:53 rurban left 00:56 yqt left 00:57 BenGoldberg joined
seatek MasterDuke, done -- 186 open on that page.. didn't search for it to see if it already was reported. 00:57
00:58 Jesmaybe joined
MasterDuke looks like you're in the perl6 team, you can probably directly edit the file from github 01:00
seatek MasterDuke, ok I'll try 01:01
dalek c: fe5f6d7 | adaptiveoptics++ | doc/Language/typesystem.pod6:
Update typesystem.pod6 in Class section for minor grammar fix
01:03
01:03 Jesmaybe left
seatek MasterDuke, well, it worked. 01:05
MasterDuke great, glad to see people contributing 01:06
seatek MasterDuke, I suppose I'm expected to resolve my own bug now... ;)
01:06 Jesmaybe joined
seatek MasterDuke, do you guys like those tracked in the bug system, or just making the changes in version control ok? 01:07
MasterDuke something small like that is fine to just change 01:08
seatek k
01:09 grondilu left 01:12 Jesmaybe left, Jesmaybe joined 01:20 Jesmaybe left, Jesmaybe joined
AlexDaniel seatek++ 01:28
dalek c: f78c962 | adaptiveoptics++ | doc/Language/typesystem.pod6:
Update typesystem.pod6, clarifying comma and clarifiying negation added
01:31
01:31 Jesmaybe left, Jesmaybe joined 01:36 pierre_ left 01:38 pierre_ joined 01:42 w4and0er96 joined, Jesmaybe left
dalek c/spellcheck: 9398f28 | coke++ | doc/Language/traps.pod6:
fix typos
01:43
c/spellcheck: d2574de | (Zoffix Znet)++ | doc/Type/Numeric.pod6:
Clarify special cases in Numeric.log

don't hide words file
01:43 dalek left 01:44 Jesmaybe joined
seatek does anyone know of any good documention on the syntax for dispatching on attribute handles? 01:47
01:48 dalek joined, ChanServ sets mode: +v dalek 01:56 amalia_ left 01:57 cibs left 01:58 cibs joined 02:00 amalia_ joined 02:01 flexibea` joined 02:02 noganex joined 02:04 flexibeast left, cdg joined 02:05 noganex_ left 02:06 cdg left, cdg joined 02:08 olinkl joined 02:17 Jesmaybe left 02:18 Jesmaybe joined 02:23 pierre_ left 02:31 pierre_ joined 02:35 pierre_ left
[Coke] github.com/perl6/doc/blob/master/d....pod6#L627 - what's the N<> here? 02:51
02:51 Jesmaybe left 02:52 Jesmaybe joined
[Coke] Looks like we could remove the N<> and move that whole sentence to ethe end of the P. 02:52
m: say nextcallee.perl; 02:56
camelia rakudo-moar 906719: OUTPUT«nextsame is not in the dynamic scope of a dispatcher␤ in block <unit> at <tmp> line 1␤␤»
02:56 cdg left 02:58 Jesmaybe left, Jesmaybe joined 03:00 noganex_ joined 03:02 noganex left 03:03 pierre_ joined 03:15 khw left, pierre_ left 03:17 skids left 03:22 pierre_ joined 03:23 Jesmaybe left
seatek m: say '12ab' ~~ m/<[a .. f 0 .. 9]> ** 4/ ?? 'good' !! 'bad'; 03:25
camelia rakudo-moar 906719: OUTPUT«good␤»
seatek say '12a' ~~ m/<[a .. f 0 .. 9]> ** 4/ ?? 'good' !! 'bad'; 03:26
m: say '12a' ~~ m/<[a .. f 0 .. 9]> ** 4/ ?? 'good' !! 'bad';
camelia rakudo-moar 906719: OUTPUT«bad␤»
seatek m: say '12abc' ~~ m/<[a .. f 0 .. 9]> ** 4/ ?? 'good' !! 'bad';
camelia rakudo-moar 906719: OUTPUT«good␤»
03:27 pierre_ left
seatek it's not wanting exactly 4 of that list... less than 4 is bad. more than 4 is ok. I understand why.... 03:27
but is there way to enforce exactly 4 of the list, like you could use {4} in perl5 ?
now i can't remember if perl5 actually did enforce exactly 4 in a list 03:31
MasterDuke you need to wrap it in ^ $
seatek MasterDuke, oh man. duh. yeah. thanks :) 03:33
MasterDuke m: say '12a' ~~ m/<xdigit> ** 4/ 03:34
camelia rakudo-moar 906719: OUTPUT«False␤»
MasterDuke m: say '12ab' ~~ m/<xdigit> ** 4/
camelia rakudo-moar 906719: OUTPUT«「12ab」␤ xdigit => 「1」␤ xdigit => 「2」␤ xdigit => 「a」␤ xdigit => 「b」␤»
MasterDuke <xdigit> is a builtin character class for <[a .. f 0 .. 9]>
seatek oo that's handy 03:35
c/spellcheck: 5111425 | coke++ | / (5 files):
Use 2 dict files -

one for words, jargon, programs, methods, classes, etc. one for code that looks like words
track more words
seatek that' can't be subset typed though can it?
MasterDuke not sure what you mean? 03:38
03:39 aborazmeh left
seatek it's not a "type" that can have subsets created of that type... 03:40
dalek c: 34ca6b4 | coke++ | doc/Language/ (2 files):
fix typo
c: 0fce86b | coke++ | doc/Language/ (4 files):
alternate forms.
seatek m: subset hexword of xdigit where *.chars == 40;
camelia rakudo-moar 906719: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Invalid typename 'xdigit'␤at <tmp>:1␤------> 3subset hexword of xdigit7⏏5 where *.chars == 40;␤»
dalek c/spellcheck: 34ca6b4 | coke++ | doc/Language/ (2 files):
fix typo
c/spellcheck: 0fce86b | coke++ | doc/Language/ (4 files):
alternate forms.
c/spellcheck: ebefa37 | coke++ | doc/Language/ (6 files):
Merge branch 'master' into spellcheck
seatek m: subset hexword of xdigit where *.chars == 4; 03:41
camelia rakudo-moar 906719: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Invalid typename 'xdigit'␤at <tmp>:1␤------> 3subset hexword of xdigit7⏏5 where *.chars == 4;␤»
MasterDuke m: subset Hex of Str where /^ <xdigit>..5 $/; my Hex $a = "ab3"; say $a 03:42
camelia rakudo-moar 906719: OUTPUT«Type check failed in assignment to $a; expected Hex but got Str ("ab3")␤ in block <unit> at <tmp> line 1␤␤»
seatek ok so yeah -- it stays in regexes then :)
much clearer 03:43
weird i seem to remember that from ages ago.. a long list of those in regexes 03:44
i'm having flashbacks
MasterDuke typo in my example though, should have been <xdigit>**5 03:46
seatek yeah
MasterDuke++ thanks 03:47
MasterDuke np. btw, i'm outa here now, but if you've got a min or two could you create a docs issue that the builtin char classes aren't documented? 03:48
geekosaur I think you can add one, but yes, needs docs 03:49
regexes are a built-in grammar which can have terms and such added to it
seatek MasterDuke, sure thing
[Coke] yay, down to 85 files with spelling issues. 03:54
geekosaur oj, the implementation isn't composable :/ 04:02
04:15 brillenfux joined 04:16 wamba joined
AlexDaniel m: my regex xdigzit5 { [ <xdigit> | ‘z’ ]**5 }; say ‘12z3b’ ~~ /^ <xdigzit5> $/ 04:19
camelia rakudo-moar 906719: OUTPUT«「12z3b」␤ xdigzit5 => 「12z3b」␤ xdigit => 「1」␤ xdigit => 「2」␤ xdigit => 「3」␤ xdigit => 「b」␤»
AlexDaniel seatek: ↑ ?
geekosaur thought they wanted a subset
although actually that should be doable to 04:20
AlexDaniel sure
geekosaur [ <xdigit> - whatever [
er
]
AlexDaniel ummm don't think this is going to work
<[\w] - [\d]> works, but <xdigit> probably won't 04:21
ShimmerFairy geekosaur: you mean like <+xdigit + [z]> ?
geekosaur hm. speculations said it should; I don't know if the impl does it though
AlexDaniel ShimmerFairy is right, it does work 04:22
ShimmerFairy m: say "CEF" ~~ /<+alpha - [Ff]>+/
camelia rakudo-moar 906719: OUTPUT«「CE」␤»
AlexDaniel yea
cool
04:23 BenGoldberg left
AlexDaniel anyway, just wanted to say that you can create your own regexes/rules/tokens if you need that kind of stuff. And these basically give you infinite freedom, so you can have subsets or whatever else you wish 04:25
04:33 AlexDaniel left 04:38 Cabanossi left 04:40 dj_goku joined, dj_goku left, dj_goku joined 04:41 Cabanossi joined 04:45 dataf4l joined
dataf4l hey guys 04:45
I just posted my first perl6 bug
rt.perl.org/Public/Bug/Display.html?id=129909
I wonder if I’ll ever be able to actually compile perl6 programs to binary executables or at least jar files so my clients can run them without hassle. 04:46
seatek ooo, yeah, very cool Shimmer, Alex & geek 04:56
04:56 Derperperd left
seatek yeah my thing was wanting a custom "variable type" of Sessid where it's a 40-character long hex thang.... settled on: subset Sessid of Str where m/^<xdigit> ** 40$/; 04:57
the xdigit replaced my <[a..f 0..9]> 04:59
05:00 rurban joined, rurban left
seatek say '123abcg' ~~ /<xdigit + [g]>+/ 05:03
m:say '123abcg' ~~ /<xdigit + [g]>+/ 05:04
m: say '123abcg' ~~ /<xdigit + [g]>+/
camelia rakudo-moar 906719: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Quantifier quantifies nothing␤at <tmp>:1␤------> 3say '123abcg' ~~ /<xdigit +7⏏5 [g]>+/␤»
seatek m: say '123abcg' ~~ /<+xdigit + [g]>+/
camelia rakudo-moar 906719: OUTPUT«「123abcg」␤»
seatek that is interesting -- needs the + up front...
m: say '123abch' ~~ /<+xdigit + [g]>+/
camelia rakudo-moar 906719: OUTPUT«「123abc」␤»
seatek very cool you can modify those inline 05:05
m: say '123abch' ~~ /<0 .. 9 a .. f + [g]>+/
camelia rakudo-moar 906719: OUTPUT«5===SORRY!5===␤Unrecognized regex metacharacter < (must be quoted to match literally)␤at <tmp>:1␤------> 3say '123abch' ~~ /<7⏏050 .. 9 a .. f + [g]>+/␤Regex not terminated.␤at <tmp>:1␤------> 3say '123abch' ~~ /<7⏏050 .. 9 a .. …»
seatek m: say '123abch' ~~ /<[0 .. 9 a .. f] + [g]>+/ 05:06
camelia rakudo-moar 906719: OUTPUT«「123abc」␤»
seatek m: say '123abch' ~~ /<+[0 .. 9 a .. f] + [g]>+/
camelia rakudo-moar 906719: OUTPUT«「123abc」␤»
seatek hmm
05:15 mohae_ left 05:16 mohae joined 05:17 flexibea` left 05:19 seatek left 05:22 Cabanossi left 05:24 Cabanossi joined 05:36 dataf4l left 05:38 Actualeyes left, pierre_ joined, Actualeyes joined 05:43 pierre_ left 05:47 msanderson joined 05:50 wamba left 05:51 domidumont joined 05:52 msanderson left 05:53 circ-user-8mgj7 joined, aborazmeh joined, aborazmeh left, aborazmeh joined 05:55 domidumont left 05:56 domidumont joined 05:57 RabidGravy joined
circ-user-8mgj7 NICK msanderson 05:58
about
ABOUT
sigh 05:59
06:01 circ-user-8mgj7 is now known as msanderson 06:05 pierre_ joined 06:15 salva joined, salva left 06:17 msanderson left 06:21 grondilu joined 06:23 neuraload joined 06:27 wamba joined 06:28 darutoko joined, firstdayonthejob joined 06:37 FROGGS joined, firstdayonthejob left 06:39 abraxxa joined 06:45 wamba left 06:46 Xliff joined 06:47 kjs_ joined 06:50 bjz joined 06:53 wamba joined 06:56 bjz left 06:57 aborazmeh left 07:00 rindolf joined 07:02 ufobat joined, bjz joined 07:03 bjz left
andrzejku hello 07:03
good morning
FROGGS morning andrzejku
07:03 rurban joined 07:07 kjs_ left 07:10 pierre_ left 07:15 domidumont left 07:16 domidumont joined 07:17 jcallen_ joined 07:18 Timbus_ left, Timbus joined 07:19 pierre_ joined, bjz joined 07:20 Actualeyes left, Actualeyes joined 07:23 bjz left, pierre_ left 07:25 zakharyas joined, pdcawley joined 07:26 bjz joined, CIAvash joined, flexibeast joined 07:30 canopus left 07:31 lizmat joined 07:35 nadim joined 07:38 kjs_ joined, canopus joined 07:43 wamba left, dakkar joined 07:45 kjs_ left 07:46 wamba joined 07:50 pierre_ joined 07:52 user9 joined 07:53 yqt joined 07:54 jargan is now known as jast 08:01 jonas3 joined
El_Che morning 08:02
andrzejku El_Che: communist? 08:04
08:04 rindolf left 08:07 espadrine joined, wamba left 08:08 bjz left 08:09 dakkar left 08:11 dakkar joined 08:12 vendethiel- left, vendethiel joined 08:16 zakharyas left 08:18 abraxxa left 08:21 abraxxa joined 08:30 abraxxa left, abraxxa joined 08:36 ufobat left, lizmat left 08:37 lizmat joined 08:47 espadrine left
lizmat clickbaits p6weekly.wordpress.com/2016/10/17/...ease-time/ 08:58
09:03 xinming_ joined 09:04 ufobat joined
El_Che lizmat: could interest some people on the weekly: github.com/nxadm/rakudo-pkg/releas...ag/2016.10 (deb and rpms for 2016.10) 09:05
lizmat El_Che: thanks, will do so next week :-) 09:06
09:07 xinming left 09:08 pierre_ left 09:10 pierre_ joined 09:23 geekosaur left, bjz joined 09:24 bjz left 09:25 geekosaur joined, rudolfochrist joined 09:27 kjs_ joined 09:33 kjs_ left, pdcawley left 09:39 labster left 09:49 rudolfochrist left 09:51 skids joined 10:06 kjs_ joined 10:09 pierre_ left 10:13 skids left 10:14 pierre_ joined 10:16 rindolf joined 10:19 pierre_ left 10:22 pdcawley joined, bjz joined 10:24 rurban1 joined 10:26 dalek left 10:27 pdcawley left, mindos left, jnap_ left, TEttinger left, rurban left 10:28 dalek joined, ChanServ sets mode: +v dalek, notbenh_ joined, mindos joined, jnap_ joined 10:31 pdcawley joined 10:34 zakharyas joined 10:38 kjs_ left
moritz El_Che: fwiw I'm still planning to create a Debian repo hosted somewhere on *.perl6.org; I've tried the simplest thing that could possibly work on the past weekend, but it didn't work :/ 10:40
10:40 Gauurof joined 10:41 Gauurof left 10:44 kjs_ joined 10:47 bjz left 10:52 grondilu left 10:54 pdcawley_ joined 10:55 ocbtec joined, pdcawley left 10:57 pierre_ joined, pierre_ left 10:58 rurban joined, rurban left 11:01 rurban1 left, bjz joined 11:03 pmurias joined
pmurias hi 11:03
moritz \o pmurias 11:04
masak oh hai pmurias 11:06
11:11 sftp left 11:12 pierre_ joined
masak today's word: "gnaborretni" 11:12
(something people apparently sometimes call the inverted interrobang)
11:16 sftp joined 11:17 araujo joined, araujo left, araujo joined 11:21 kjs_ left 11:22 araujo left
El_Che moritz: what do you mean? a fully integrated debian/ubuntu pkg? My approach for the monthly rakudo releases differs by having no dependencies and installing everything in /opt/rakudo. The idea is not to have a conflct with a future debian/ubuntu/centos package 11:25
11:26 pierre_ left, pierre_ joined
moritz El_Che: the charm would be to only have to configure the repo once, and then being able to upgrade (or install new versions) with apt 11:27
El_Che I think the packages are repo friendly -> they have a version and revision number 11:29
11:41 pdcawley_ left, pierre_ left 11:47 pierre_ joined
moritz yes 11:51
I just need to do a bit of scripting to teach aptly to have several published repos, not just one 11:52
11:57 lizmat_ joined, lizmat left 12:00 CIAvash left 12:03 bjz_ joined, bjz left 12:14 rindolf left 12:17 vendethiel left 12:18 vendethiel joined 12:36 itaipu joined, Actualeyes left 12:38 kjs_ joined 12:46 sufrostico joined, Derperperd joined 12:50 kjs_ left, Actualeyes joined 12:57 raiph joined
Woodi_ hi #perl6 :) 12:58
iBakeCake \o
Woodi_ after git was invented it's possible, that someday we would like to have "versions" being like switching branches. hmm, or maybe just drop thet fs thing asap ? ;) 12:59
raiph hi all 13:00
13:00 kjs_ joined
iBakeCake hi 13:00
Woodi_ yo raiph :)
13:01 kjs_ left
[Coke] m: StructObj.new 13:02
camelia rakudo-moar 20d37a: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared name:␤ StructObj used at line 1␤␤»
raiph iBakeCake: clicking "Run this code" buttons on perl6.party/post/IRC-Client-Perl-6-...IRC-Module yields "Could not find IRC::Client"
iBakeCake raiph: yeah, the widgets only run anything glot.io can run. In the alternate universe with non-lazy Zoffix unrunnable code samples have the feature disabled :) 13:04
dalek c: 420e381 | coke++ | doc/Language/nativecall.pod6:
use full word
13:07 seatek joined
pmurias Woodi_: just using the HEAD of a git repo is not a replacement for having versions despite what a lot of people want to do 13:09
13:12 Derperperd left 13:29 cognominal joined
Woodi_ pmurias: of course that that is just idea. actually for versions like this you need git repo for every file or some set of related files. so fs working like git with set of heads :) 13:40
13:46 aborazmeh joined, aborazmeh left, aborazmeh joined
Woodi_ ...per user :) there was some plans for fs as db but probably still too futuristic :) 13:47
iBakeCake What's "fs"? File system?
13:50 raiph left, raiph joined
Woodi_ yes 13:50
[Coke] m: '{'.uniname.say # docs refer to these as curleys, curlys, curlies, braces... 13:51
camelia rakudo-moar 20d37a: OUTPUT«LEFT CURLY BRACKET␤»
iBakeCake curleys and curlys look like incorrect spelling of curlies :)
[Coke] I agree. I'd be happy to use 'curly brackets' everywhere we have the other version. 13:52
13:53 skids joined
iBakeCake sounds good to me 13:53
13:53 grondilu joined
seatek i think they can't really be called curly unless they loop around themselves. they're curvy brackets. :) 13:53
gfldex [Coke]++ # for many words
jnthn
.oO( moustache brackets )
13:54
seatek that works!
dalek c/spellcheck: 420e381 | coke++ | doc/Language/nativecall.pod6:
use full word
c/spellcheck: fb7ddd2 | coke++ | doc/Language/nativecall.pod6:
Merge branch 'master' into spellcheck
c/spellcheck: c61c7d1 | coke++ | xt/ (2 files):
more words
c: cbaed78 | coke++ | doc/Language/rb-nutshell.pod6:
curley -> curly
13:55
[Coke] (curly)? brace appears many times, that's some editing to update that. 13:56
I at least killed the 'e'
El_Che I like the sound of curley
sound french
s
13:56 wamba joined 13:58 rudolfochrist joined
dalek c: df250da | coke++ | doc/Type/Parameter.pod6:
fix typo
13:59
seatek I love how integrated types are into the DBIish stuff. You can have variable defined as a DateTime type, and it JUST WORKS for timestamp column types. :) 14:00
14:00 khw joined
iBakeCake Isn't that simply 'cause it stringifies to a timestamp? 14:01
m: DateTime.now.Str.say
camelia rakudo-moar 20d37a: OUTPUT«2016-10-18T16:01:31.664285+02:00␤»
14:02 cp_ joined
seatek Could be... but it's not something I'm used to. I'm used to having all kinds of hassle with timestamps 14:02
usually you have to treat them as strings, and insert them in
14:02 brillenfux left
seatek but it's all transparent now ;) 14:03
14:05 aborazmeh left, kjs_ joined 14:07 zakharyas left, zakharyas joined
seatek wow -- it even works backward -- when you load the colum from the database into a variable of a type DateTime, it auto-creates a DateTime object for you based on the string in the database. very Cadillac. :) 14:11
14:11 zakharyas left, dataf4l joined
Woodi_ taking lots of fuel ;) 14:11
seatek hehe :) 14:12
14:13 raiph left, raiph joined
iBakeCake sub is-approx-rel (|c) { is-approx :rel-tol(1e-15), |c } 14:13
Is that how to curry in Perl 6 or is there some better way?
14:14 kurahaupo__ joined
raiph m: (&say.assuming: 'foo')() # iBakeCake 14:15
camelia rakudo-moar 20d37a: OUTPUT«foo␤»
iBakeCake awesome. raiph++ 14:16
14:17 kjs_ left, kjs_ joined 14:18 pierre_ left 14:21 raiph left, raiph joined 14:24 kjs_ left 14:29 wamba left 14:36 bjz_ left, cdg joined 14:40 jonas3 left 14:46 abraxxa left 14:48 rindolf joined 14:51 raiph left, raiph joined 14:53 rudolfochrist left, neuraload left 14:55 timeless joined 14:56 raiph left, raiph joined 14:58 w4and0er96 left 15:02 giftnuss_ left, domidumont left 15:04 vendethiel left, vendethiel joined 15:06 zakharyas joined 15:07 w4and0er96 joined 15:09 giftnuss_ joined, Spot__ joined 15:14 Lucas_One left 15:17 Spot__ left 15:19 clkao joined, Spot__ joined 15:20 pdcawley joined 15:23 zakharyas left 15:24 khisanth__ joined 15:26 pdcawley left 15:28 khisanth_ left 15:29 TEttinger joined 15:31 eliasr joined 15:32 Lucas_One joined 15:33 fumlead joined
[Coke] can we pick a standard tempfile name in doc/Language/exceptions.pod6 that might pass a spell checker? 15:33
iBakeCake Yes! 15:34
s|/tmp/0IeSX_XSWO|my-script.p6| 15:35
[Coke] roger. 15:44
15:45 dataf4l left, Derperperd joined
dalek c: 6a285c2 | coke++ | doc/Language/glossary.pod6:
fix typo
15:46
c: d7ecb33 | coke++ | doc/Language/exceptions.pod6:
use readable file names
c/spellcheck: cbaed78 | coke++ | doc/Language/rb-nutshell.pod6:
curley -> curly
c/spellcheck: df250da | coke++ | doc/Type/Parameter.pod6:
fix typo
c/spellcheck: 6a285c2 | coke++ | doc/Language/glossary.pod6:
fix typo
c/spellcheck: e8a3012 | coke++ | doc/ (3 files):
Merge branch 'master' into spellcheck
c/spellcheck: d7ecb33 | coke++ | doc/Language/exceptions.pod6:
use readable file names
c/spellcheck: 882da05 | coke++ | doc/Language/exceptions.pod6:
Merge branch 'master' into spellcheck
16:03 pmurias left 16:07 dataf4l joined 16:17 harmil_wk left 16:30 cdg_ joined 16:33 cdg left 16:39 sufrostico left 16:40 matiaslina joined 16:47 dakkar left 16:50 sufrostico joined 16:55 kent\n left, kent\n joined 16:56 kent\n left 16:59 kent\n joined 17:05 rindolf left 17:13 rindolf joined 17:14 Actualeyes left
El_Che moritz: the repo idea would be a good thing. In my opinion standalone rakudo packages (maybe including zef/panda, now a install script) cover a different base than rakudo star (get a full distribution, but you have to compile it) and real OS packages (best QA and integration, but old). 17:17
17:18 pdcawley joined
El_Che moritz: I think specifically of my use case: docker, and applications with limited local dependencies 17:18
17:27 cp_ left, pdcawley left 17:30 hobbs joined, hobbs left, hobbs joined 17:32 pdcawley joined 17:33 pdcawley left, Cabanossi left, Bucciarati left, gypsydav15 joined, rjbs- joined, hoelzro_ joined, nemo joined, Ulti_ joined, Cabanossi joined, damnlie joined, baest joined, jsimonet joined, Kaffe joined, facetious joined, Jonis joined, robins joined, ponbiki joined, pdcawley joined, Bucciarati joined, BooK joined, ugexe joined, stmuk joined 17:34 perigrin joined, domm1 joined, moritz joined, hahainternet joined, go|dfish joined, nemo is now known as Guest32147, Peter_R joined, ponbiki is now known as Guest4979, spider-mario joined, luis joined 17:35 emdashcomma joined 17:36 protium joined, ssm_ joined
skids
.oO(everyone must've had to leave until the firemarshal was gone.)
17:36
17:36 girafe joined 17:37 Hotkeys joined, darutoko left, chansen_ joined 17:38 domidumont joined, wamba joined 17:39 zakharyas joined, Guest4979 is now known as ponpon, zostay joined 17:41 inokenty left, ggherdov joined, mithaldu_ joined 17:42 kipd joined, peteretep joined 17:43 SmokeMachine____ joined, Guest32147 is now known as nemo 17:46 tbrowder joined 17:50 pdcawley left 17:51 pecastro joined 17:53 chienjo joined 17:54 chienjo left 18:00 JimmyZ joined 18:20 dataf4l left 18:21 labster joined, sufrostico left 18:24 itaipu left 18:25 AlexDaniel joined 18:40 itaipu joined 18:44 breinbaas joined
dalek c: 31fa161 | (Zoffix Znet)++ | doc/Language/testing.pod6:
Correct is-approx behaviour with no tolerances

The function has been made[^1] to DWIM when no explicit tolerances are provided instead of assuming a default absolute tolerance of 1e-5, which proved nearly useless in real-world use cases.
  [1] github.com/rakudo/rakudo/commit/82...444d9991d5
18:44
18:47 itaipu left 18:50 fumlead left 18:54 itaipu joined 18:55 inokenty joined, fumlead joined
dalek c: 6a93543 | (Zoffix Znet)++ | doc/Language/testing.pod6:
Mention DWIMMY is-approx uses *absolute* $expected

  ... to figure out what sort of tolerance to use
18:57
samcv is there a way to make perl6 let me redefine variables with 'my'? I know that shouldn't be done, i'm just trying to run a bunch of example code and want to check it to make sure nothing else is wrong with it 19:02
iBakeCake samcv: well, it does let you. It simply issues a warning 19:03
samcv: I think if you add CONTROL { when CX::Warn {.resume} }; somewhere at the top of the program it'll do the trick to silence them
m: CONTROL { when CX::Warn {.resume} }; my $foo = 42; my $foo = 88; say $foo 19:04
camelia rakudo-moar 82432a: OUTPUT«Potential difficulties:␤ Redeclaration of symbol '$foo'␤ at <tmp>:1␤ ------> 3:Warn {.resume} }; my $foo = 42; my $foo7⏏5 = 88; say $foo␤88␤»
iBakeCake m: CONTROL { default {.^name;} }; my $foo = 42; my $foo = 88; say $foo
camelia rakudo-moar 82432a: OUTPUT«Potential difficulties:␤ Redeclaration of symbol '$foo'␤ at <tmp>:1␤ ------> 3fault {.^name;} }; my $foo = 42; my $foo7⏏5 = 88; say $foo␤88␤»
iBakeCake ... or not :/
19:04 sufrostico joined
samcv yeah that didn't fix it. i get a ton of warnings and it doesn't run 19:04
i'm working to get the learnxinyminutes for perl6 much clearer and make sure 100% of the code is actually valid and has the correct response 19:05
iBakeCake m: class Foo {}; class Foo {}; say 42
camelia rakudo-moar 82432a: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Redeclaration of symbol 'Foo'␤at <tmp>:1␤------> 3class Foo {}; class Foo7⏏5 {}; say 42␤ expecting any of:␤ generic role␤»
samcv would be nice to run the whole thing at once, though at least the sections i've worked on i've checked sections of code.
iBakeCake Yeah, it doesn't let you redefine just any old thing
Split on lines that start with # and stick stuff into separate files :) 19:06
(or just .EVAL
)
m: my $code = Q:to/E/;␤class Foo {};␤# tis just a class␤#hehe␤class Foo {}␤say "hi"␤E␤$code.split(/(^^'#'\N+\n)+/)».EVAL 19:09
camelia rakudo-moar 82432a: OUTPUT«===SORRY!===␤Could not locate compile-time value for symbol Foo␤»
iBakeCake heee :}
samcv well # don't always seperate things. i could just seperate it at the spots it redefines things there's like 7 of them or rename the variables. i think it'd be nice to make it able to run 19:12
because the site lets you download a .pl file that has all the code in it 19:13
so not repeating variable names may be the way to go
iBakeCake +1
19:13 itaipu left
samcv i think i've fixed the multiple places where pseudocode was used as an example and replaced that with a working example. but i'm sure i'll find more when i fix the variable redefinition 19:14
gfldex samcv: you may want to peek into the following tool that we use to extract examples from the docs. It doesn't use EVAL to make compile time errors explode early. github.com/perl6/doc/blob/master/u...xamples.p6 19:16
samcv oh nice. great 19:17
fumlead Hi #perl6, could you please help me with: 19:20
method calc-op:sym<add> ($/) { make [+] $<num>; }
What is the [+] ?
Is it circumfix operator?
gfldex m: say [+] 1,2,3; # reduction + operator AKA sum 19:21
camelia rakudo-moar 82432a: OUTPUT«6␤»
19:21 domidumont left
[Coke] fumlead: docs.perl6.org/language/operators#..._Operators 19:22
fumlead [Coke]: thanks, exactly what I want 19:23
Searching in docs lead me to the array constructor 19:24
[Coke] someone should open a ticket in perl6/docs to make [] point to that. 19:26
gfldex i will fix it right away 19:27
iBakeCake gfldex++
19:27 pyrimidi_ joined
dalek c: 6eddfd2 | gfldex++ | doc/Language/operators.pod6:
index reduction meta operators
19:29
19:31 canopus left 19:32 acrussell joined
dalek c: 47bbf0a | gfldex++ | doc/Language/operators.pod6:
index cross meta operator
19:33
c: c58aa6f | gfldex++ | doc/Language/operators.pod6:
index zip meta operator
19:35 pyrimidine joined, skids left 19:37 pyrimidi_ left, zakharyas left 19:43 canopus joined 19:48 bjz joined
rindolf Hi all! Is there any work being done on optimising the performance of sprintf in latest rakudo perl6? Can I help? 19:55
19:56 dataf4l joined
timotimo i don't think anybody has been working on that recently 19:58
[Coke] no one is actively working on sprintf, nope.
timotimo you can, of course, work on it if you like
[Coke] let us know if you need pointers on where to dig in. 20:02
20:11 Derperperd left 20:15 itaipu joined
samcv is it correct to say a variable declared with my $var is globally scoped in perl6? idk what do you call the part of the code not in MAIN and just by itself. like how it's scoped 20:21
mst it's still lexically scoped, just to the file 20:22
samcv yeah.
mst I often refer to that, in perl5, semi-innacurately, as "file scoped"
since it's not accessible from outside the file so it's arguably not global qua global
samcv yeah. i think that's good i'll call it $file_scoped. that makes it very clear the scope of it 20:23
20:27 Ven_ joined 20:32 cooper_ joined 20:34 acrussell left, cooper_ left 20:35 acrussell joined 20:42 ufobat left 20:43 cooper_ joined 20:45 bjz left 20:48 cpage_ left 20:50 pecastro left 20:56 bjz joined 20:57 matiaslina left 21:04 Ven_ left 21:05 itaipu left 21:07 Lucas_One left, bjz left 21:08 labster left 21:10 labster joined 21:14 dalek left 21:15 mindos left, mindos joined 21:16 firstdayonthejob joined, bjz joined 21:17 Matias[m] left, M-Illandan left, enick_966 left, ilmari[m] left, tadzik left, Matthew[m] left, solarbunny left 21:19 enick_966 joined, dalek joined, ChanServ sets mode: +v dalek 21:20 ilmari[m] joined, Matias[m] joined, tadzik joined, cibs left 21:21 cibs joined 21:22 M-Illandan joined, Matthew[m] joined 21:23 acrussell left 21:25 solarbunny joined 21:26 Lucas_One joined 21:30 cpage_ joined 21:37 bjz left 21:42 fumlead left 21:45 espadrine joined 21:46 yqt left 21:48 sufrostico left 21:49 lostinfog joined 22:04 RabidGravy left, wamba left
rindolf [Coke]: hi! Sorry for the late response, and I could use some pointers. 22:16
samcv i think i'm correct about this, but in perl6 hashes are stored similarly to arrays unlike perl 5 22:17
and hashes in perl6 retain their order but perl5 does not. i think in perl 5 it's literally a 'hashtable' in the computer science definition making lookup very fast and deduplicating.
ok perl6 seems to remove duplicates too, but are the other things i said correct or are they not completely correct? 22:18
geekosaur I think perl6 hashes do not retain order, but the default presentation sorts the keys? 22:20
(for some reason)
samcv ah, so they are not stored the same as arrays right 22:22
that's what i would expect. learnxinyminutes says ‘Hashes are actually arrays of Pairs’ but i don't think that's correct
want to double check before i change it :) 22:23
geekosaur does not recall actual implementation, sadly
samcv i know they can be used ‘as’ arrays more easily than perl5 but i *think* they are converted by perl 6
geekosaur might be more a question for #perl6-dev 22:24
samcv i will try to dig a little more
i'll ask in there
22:25 FROGGS left
cognominal m: $_ = 'abcd'; my ($b, $d) = @( m/a(b)c(d)/ ); say $b; say $d 22:26
camelia rakudo-moar eb6907: OUTPUT«「b」␤「d」␤»
cognominal m: $_ = 'abcd'; my ($b, $d) = m/a(b)c(d)/; say $b; say $d
camelia rakudo-moar eb6907: OUTPUT«「abcd」␤ 0 => 「b」␤ 1 => 「d」␤(Any)␤»
22:26 cibs left
cognominal I thought the @() would be unecessary 22:27
22:27 cibs joined
geekosaur perl5 does that, but perl 6 needs to be more careful or it gets some other use cases wrong 22:28
more specifically, m// returns an item ($/), not a list/array-type thing 22:29
22:29 BenGoldberg joined 22:31 sufrostico joined
cognominal geekosaur, I thought the my ($b, $d) would set an array context 22:31
geekosaur, what use cases ? 22:32
geekosaur any time you have a thing that is an object that presents different scalar vs. array results (as $/ does), you need to provide some way to get either one
cognominal but again, the my ( ) should set an array context. Or is this Perl 5 thinking ? 22:33
geekosaur if you treat $/ as a single thing, you get the matched string and a list of Pair for captures. if you treat as an array then you just get the captures.
that's p5
p6 doesn't have contexts in that sense
cognominal I have not done Perl 6 in months, that's my perl 5 background taking over :( 22:34
geekosaur so, a problem with the p5 behavior is that the context at the callsite has action-at-a-distance on the function (cf. wantarray) 22:35
22:35 lostinfog left
geekosaur whereas you'd kinda like functions to be consistent. so p6 does not magically propagate context, you need to tell it what you want (in this case with @()). (but then you end up with things like the single arg rule to produce behavior that people expect that context did in p5...) 22:36
timotimo rindolf: well, first of all you really need to have a benchmark so that you can measure not only improvements that you make, but also to find out what exactly is slow 22:44
rindolf timotimo: OK, sorry, but I need to go to sleep. 22:45
timotimo: good night.
22:46 itaipu joined, rindolf left 22:56 aries_liuxueyang joined 22:58 firstdayonthejob left 23:01 pecastro joined 23:09 cdg_ left 23:10 pdcawley joined, kurahaupo__ left 23:22 espadrine left 23:29 ocbtec left 23:30 girafe left 23:33 skids joined 23:41 pdcawley left 23:44 xinming_ left 23:58 pierre_ joined, pierre_ left