»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend! | YAPC::NA talk submissions due end of Month. Have you submitted your talk?
Set by toddr_ on 18 March 2013.
00:00 Maddingue joined 00:01 fgomez_ left, raiph joined, fgomez joined 00:02 jtpalmer_ joined, risou joined 00:03 genehack joined, krunen joined, awwaiid joined, clkao joined 00:04 ingy joined, prammer_ joined, leprevost joined, Exodist joined, pmichaud joined, Ayiko joined, Tene joined, simcop2387 joined 00:08 pmichaud left, pmichaud_ joined, Exodist left, Tene left 00:10 Tene joined, raiph left 00:13 Ayiko left, Exodist joined, _sri joined, Ayiko joined 00:14 TimToady_ joined, __sri left 00:15 pjcj joined, rom1504 left, rom1504 joined 00:16 raiph joined 00:17 simcop2387_ joined 00:18 bonsaikitten joined 00:20 arnsholt_ joined 00:21 nebuchad` joined 00:22 nebuchad` is now known as nebuchadnezzar, krunen left, awwaiid left 00:23 awwaiid joined, krunen joined 00:24 clkao left, simcop2387 left, Yappo______ left, rking left, mangala_ left, BinGOs left, simcop2387_ is now known as simcop2387 00:27 PerlJam joined 00:28 s0rear joined 00:29 integral joined, integral left, integral joined, Patterner left, BinGOs joined, clkao joined, rjbs joined 00:30 jaldhar joined, genehack_ joined, Psyche^ joined, Psyche^ is now known as Patterner 00:31 raiph_ joined 00:33 GlitchMr- joined 00:34 integral left, Chillance left, integral_ joined, integral_ left, integral_ joined, raiph left, Exodist left, genehack left, Gothmog_ joined 00:35 xinming left 00:36 robinsmidsrod joined, breinbaas joined, xinming joined, prammer_ left, mtk left 00:37 prammer_ joined, leprevost left 00:38 dalek left 00:39 dalek joined, ChanServ sets mode: +v dalek, sivoais joined 00:41 s0rear is now known as sorear, Exodist joined, mangala joined, avar joined, avar left, avar joined, salv0 joined 00:42 Chillance joined, thou_ left 00:43 Yappo______ joined, jfried joined, mj41 joined 00:45 p6eval joined, tadzik joined, cameron.freenode.net sets mode: +v p6eval, leprevost joined 00:47 jerome_ joined 00:48 Gruber joined 00:49 mtk joined 00:50 Chillance left, Exodist left 00:51 salv0 left, salv0 joined, lustlife joined, Exodist joined
diakopter . 00:59
01:03 Exodist left, anuby joined, exodist_ joined 01:05 GlitchMr joined 01:07 rking joined, Trey joined, GlitchMr- left, Gruber left 01:08 groky left, colomon joined 01:09 gcole joined, groky joined 01:10 Trey is now known as TreyRust, TreyRust left 01:13 Grrrr joined 01:16 gcole_ joined 01:19 gcole left 01:21 lustlife1 joined, raiph_ left 01:25 preflex left 01:26 jtpalmer_ left 01:27 dayangkun left, GlitchMr- joined, eiro_ joined 01:28 GlitchMr left, lustlife left, mtk left, ingy left 01:29 raiph joined 01:30 leprevost left 01:31 exodist_ left, eiro left, preflex joined 01:32 Exodist joined, ingyfoo joined, Exodist left 01:33 gcole_ left 01:39 leprevost joined, jfried_ joined 01:41 jtpalmer joined, jtpalmer left, jfried left, jfried_ is now known as jfried, mtk joined 01:42 raiph left 01:43 vaelxon joined 01:44 Yappo______ left, mj41 left 01:45 gcole joined, mj41 joined 01:48 Yappo______ joined 01:50 jtpalmer joined, lustlife joined, rking left, Exodist joined 01:53 raiph joined 01:55 sitaktif1 joined 01:57 jtpalmer_ joined 02:00 anuby left, lustlife2 joined, anuby joined 02:02 Exodist left 02:07 mtk left, leprevost left 02:09 Exodist joined, salv0 left, lustlife1 left, leprevost joined 02:10 salv0 joined 02:11 lustlife left 02:12 mtk joined, jtpalmer left 02:16 pmurias left 02:18 anuby_ joined 02:21 jtpalmer_ left 02:23 salv01 joined 02:24 mtk left, salv0 left 02:26 raiph left 02:27 salv01 left 02:28 loops joined 02:31 raiph joined 02:33 anuby__ joined, Exodist left, Yappo______ left, lustlife joined, loops left, TimToady_ is now known as TimToady 02:34 Exodist joined, TimToady left, TimToady joined, anuby left, Yappo______ joined 02:35 salv01 joined, mtk0 joined, anuby_ left 02:36 lustlife2 left 02:38 jtpalmer joined 02:41 baest joined, Exodist left 02:42 lustlife left 02:43 baest_ left 02:44 FROGGS_ joined 02:45 baest left 02:46 baest joined, FROGGS left, Pleiades` joined 02:49 Exodist joined, lustlife joined 02:59 baest_ joined 03:05 baest left 03:08 orafu left, orafu joined 03:12 skyheights joined
skyheights offers www.destroyallsoftware.com/talks/wat to masak++ and moritz++ to cheer them up after yesterday's #parrot episode 03:15
03:16 skyheights left 03:17 dwoldrich joined 03:24 preflex left, preflex_ joined, preflex_ is now known as preflex 03:31 raiph left 03:37 raiph joined 03:40 preflex left, raiph left, preflex_ joined 03:41 preflex_ is now known as preflex 03:46 preflex left, raiph joined, sevin joined
sevin t 03:46
03:47 rking joined 03:51 skids left
raiph sevin: ? 04:03
04:03 preflex joined
sevin had my cursor on the wrong window, my bad 04:05
04:08 preflex left 04:15 colomon left 04:18 thou joined 04:20 sevin left 04:23 drbean left 04:25 drbean joined 04:30 loops joined 04:37 preflex joined 04:42 preflex left, colomon joined 04:44 leprevost left 04:51 SamuraiJack joined 05:04 raiph left 05:14 raiph joined 05:16 preflex joined
moritz_ \o 05:17
05:25 preflex left 05:30 crab2313 joined 05:41 BinGOs_ joined, atrodo_ joined 05:42 preflex joined, markov1 joined, dmol joined 05:43 nebuchad` joined, apejens_ joined 05:44 jlaire_ joined, LoRe_ joined 05:46 swarley-freenode joined, yeltzooo6 joined, BinGOs left, benabik left, jlaire left, atrodo_ is now known as atrodo, yeltzooo left, nebuchadnezzar left, sunnavy_ left, crazedpsyc joined 05:47 jlaire_ is now known as jlaire, sunnavy joined 05:49 dmol left
moritz_ fwiw ilbot6 ran through the night without problems 05:50
05:56 kurahaupo left 05:57 vishaltaware joined 05:59 vishaltaware left 06:04 kaleem joined 06:10 cognominal left 06:24 lue joined
diakopter moritz_: did you see my idea about ilbot counting karma from your db? :) 06:25
moritz_ diakopter: yes, it's not new 06:26
diakopter: I thought about an extra bot that reads the db and reports karma when asked for
06:26 dwoldrich left
moritz_ but so far it's just idle pondering, no real plans to do anything 06:26
diakopter ok... I really didn't care whether it was new or not 06:27
seems you're implying you're the only one who could do it 06:28
moritz_ not at all 06:30
it's open source
06:31 kaare_ joined 06:42 apejens_ is now known as apejens 06:55 domidumont joined 07:02 FROGGS_ left 07:09 domidumont left 07:10 domidumont joined 07:17 FROGGS_ joined 07:20 renormalist joined 07:22 domidumont left
GlitchMr- while '' ne my $answer = prompt 'Temperature: ' { 07:24
I like that one
07:25 kaare_ left
moritz_ while chars(my $answer = prompt 'Temperature: ') { } #also an option 07:43
07:46 preflex left 07:47 nebuchad` is now known as nebuchadnezzar
dalek : ff7d3c3 | (Tobias Leich)++ | t/v5/ (12 files):
added "use perl5" where it was missing
 07:48
07:49 domidumont joined 08:02 preflex joined 08:07 cognominal joined 08:11 Beltechs joined 08:14 wk left 08:19 cognominal__ joined 08:20 cognominal left
dalek ecs: eac99e2 | (Stéphane Payrard)++ | S12-objects.pod:
Twigil in a BUILD call must be '!'
 08:22
08:23 preflex left, cognominal joined 08:24 cognominal__ left 08:31 nebuchad` joined 08:33 cognominal left, jlaire_ joined 08:35 nebuchadnezzar left, nebuchad` is now known as nebuchadnezzar, kaare_ joined, SamuraiJack left, crab2313 left 08:36 __rnddim__ joined 08:37 markov joined 08:38 daxim joined 08:39 atrodo left, SamuraiJack_ joined, colomon left 08:40 renormalist left, jlaire left, markov1 left, lue left, mj41 left, eiro_ left, eiro joined 08:41 kaleem left, ingyfoo left 08:42 atrodo joined, SamuraiJack_ left 08:43 SamuraiJack_ joined, renormalist joined 08:44 cognominal joined, kaleem joined 08:45 colomon joined 08:46 ingy joined, preflex joined
dalek : 302e259 | (Tobias Leich)++ | lib/Perl6/P5 (2 files):
dispatch dot infix to concat operation
 08:47
08:48 mj41 joined 08:52 preflex left 08:54 colomon left, kresike joined
kresike hello all you happy or unhappy perl6 people 08:55
08:56 Heather joined
Heather is removing KDE 08:57
bonsaikitten Heather: heretic! 08:59
Heather bonsaikitten I need something lighter, xmonad now 09:00
bonsaikitten i.imgur.com/IbKnZqc.png
bonsaikitten pff. just disable whatever you don't need
09:00 hoelzro_ is now known as hoelzro
Heather bonsaikitten I don't need kdelibs 09:01
09:01 fhelmberger joined 09:02 kaare_ left
Heather bonsaikitten to be honest I need KDE libs -__-' I still use KDevelop but I don't need plasma, for example 09:02
bonsaikitten pfff.
sorear xmonad eh. do the docs still mention me?
Heather sorear huh? 09:03
bonsaikitten funniest thing is that I don't know how to remove KDE now, I afraid to affect KDevelop
sorear Heather: i was one of the 3 members of the original xmonad team 09:04
it's _weird_ seeing how much it took off since I left (0.3-ish)
Heather sorear why left? 09:05
daxim `rpm -q --requires kdevelop4` says it depends on kdebase4-runtime. it's not possible to remove KDE and keep KDevelop 09:07
Heather daxim cool -__-'
bonsaikitten daxim: amusing how people invent package names unrelated to upstream 09:08
Heather daxim but I need to clean other things )
daxim yeah, I hate it when debian does that. can't they follow LSB/RPM naming conventions like everyone else? 09:09
bonsaikitten "everyone" ?
Heather I need to remove meta package but I afraid that some fonts / icons etc... let call it unnecessary things, will be removed too
bonsaikitten Heather: try it and see 09:10
Heather bonsaikitten just emerge -C xxx-meta ?
bonsaikitten Heather: yes
or add -a if you're a scared cat 09:11
09:11 cognominal left
Heather bonsaikitten though I need -a for next depclean 09:13
09:14 Beltechs left
sorear sleep& 09:15
09:16 sqirrel joined 09:18 cognominal joined 09:30 cognominal__ joined, cognominal left 09:31 sqirrel left
tadzik hello #perl6 09:33
Heather tadzik hey
FROGGS_ hi tadzik
09:33 dakkar joined
masak good forenoon, #perl6 09:36
moritz_ good am, masak :-) 09:37
tadzik hi masak
tadzik is teaching Perl today
masak \o/ 09:38
loops is learning Perl today
masak tadzik: powodzenia!
Heather ...
moritz_ !!!
masak ???
moritz_ r: say WHAT do { ... } 09:39
loops \@$
p6eval rakudo b972ca: OUTPUT«Stub code executed␤current instr.: 'throw' pc 350303 (src/gen/CORE.setting.pir:152936) (src/gen/CORE.setting:9353)␤called from Sub 'sink' pc 381956 (src/gen/CORE.setting.pir:165283) (src/gen/CORE.setting:10618)␤called from Sub 'MAIN' pc 381 (src/gen/perl6.pir:147) …
Heather it's polish !
moritz_ huh, where did that 'sink' come from?
looks like a bug to me
masak! you are needed here!
FROGGS_ pokes masakbot 09:41
masak huh! huh!
FROGGS_ tadzik: polish perl workshop?
masak I looked at that and didn't see anything odd.
what's the bug with 'sink'?
tadzik FROGGS_: nah, not yet
just an Advanced Perl class in spare time
FROGGS_ tadzik++ # cool! 09:42
tadzik I'm almost as frightened as excited :)
masak tadzik++ # uses his spare time to teach Perl :)
tadzik okay, I'm not that frightened once I started
cognominal__ Advanced Perl, that's Perl6 ? :)
tadzik haha, I wish :)
FROGGS_ tadzik: tell them that you can write Peril5 in Peril6 now :o) 09:43
well, at least almost 09:44
Heather what about open courses?
but not courseria...
tadzik perl6maven, mayhaps?
Heather hates courseria with weirdo reasons
tadzik is off again
masak Heather: dude, you were the one who brought up this "courseria" thing... :P 09:47
masak is off for some errands 09:48
cognominal__ coursera? 09:49
09:49 snearch joined 10:02 BinGOs_ is now known as BinGOs, BinGOs left, BinGOs joined, preflex joined
Heather courseria ! :) 10:05
hoelzro coursera? 10:07
hoelzro starts his coursera course today
(also, goede morgen #perl6)
Heather hoelzro link? 10:12
hoelzro to the course? 10:13
Heather hoelzro yep
hoelzro Heather: www.coursera.org/course/algs4partII
10:22 snearch left
Heather I want this www.coursera.org/course/mentalhealth but it's closed 10:24
hoelzro interesting 10:26
10:30 nwc10_ is now known as nwc10 10:32 anuby__ left 10:33 rindolf joined 10:36 thou left
dalek : 9b86017 | (Tobias Leich)++ | lib/Perl6/P5 (2 files):
remove tokens which are methods or subs
 10:42
10:49 erkan joined, erkan left, erkan joined 10:54 wk joined
daxim tadzik, status plpw2013? post some /news or something 11:04
hoelzro of the keywords that can introduce a new scope (ex. grammar, module, etc), is package the only one that *must* be followed by a block? 11:08
ex. package Foo::Bar; is now illegal 11:09
(I'm not considering -> { ... } or { ... } atm)
11:10 betterworld joined
FROGGS_ hoelzro: I think so, yes 11:11
hoelzro cool
FROGGS_ std: package FooBar;
p6eval std 86b102f: OUTPUT«===SORRY!===␤Semicolon form of package definition indicates a Perl 5 module; unfortunately,␤ STD doesn't know how to parse Perl 5 code yet at /tmp/4xSPjXCqb4 line 1:␤------> package FooBar⏏;␤ expecting any of:␤ name␤ trait␤Par…
FROGGS_ std: grammar FooBar;
p6eval std 86b102f: OUTPUT«ok 00:00 41m␤»
FROGGS_ std: class FooBar;
p6eval std 86b102f: OUTPUT«ok 00:00 41m␤»
FROGGS_ right
hoelzro cool 11:12
11:12 betterworld left
masak STD is the only parser I know to use the wording "unfortunately" :) 11:14
FROGGS_ masak: and the sentence after that is pretty wrong 11:15
std: use v5; package FooBar;
p6eval std 86b102f: OUTPUT«ok 00:00 52m␤»
masak std: use v5; my $foo; $foo->method(); print $foo . "bar" 11:27
p6eval std 86b102f: OUTPUT«ok 00:01 53m␤»
masak huh. wow. :)
11:29 betterworld joined, betterworld left
FROGGS_ masak: that is STD_P5.pm6, I use this grammar 11:32
11:40 betterworld joined
masak I don't know about you, but I'm every so slightly awestruck. 11:41
it's like, I know where the Perl 6 grammars thing is heading, and has been heading all along. but seeing it actually get there step by step is... very fulfilling. :)
FROGGS_ well, I'm just happy that there already is a Perl 5 grammar, writing one from scratch might be a lifetime job for me 11:51
but twiddling a bit so that the action methods fit the grammar rules is a job I can do within time :o)
diakopter std: use v5 { }; 11:52
p6eval std 86b102f: OUTPUT«===SORRY!===␤Cannot locate module v5 at /tmp/PEOjPBaicE line 1:␤------> use v5 { }⏏;␤Check failed␤FAILED 00:00 50m␤»
diakopter p5eval: use v5 { };
p5eval diakopter: ERROR: syntax error at (eval 7) line 1, near "use v5 { "
FROGGS_ the v5.pm might be a preset for other languages like cobol some day, but I won't dare to write the grammar
diakopter p5eval: use v5 qw(); 11:53
p5eval diakopter: ERROR: syntax error at (eval 7) line 1, near "use v5 qw()"
diakopter slow
FROGGS_ diakopter: use v5 takes no params
masak: do you know cobol? 11:54
tadzik daxim: ayes, there will be updates this week 11:55
FROGGS_ well, I guess such an old language has not that kind of perlish syntax (perlish =:= hard to implement)
daxim make it so
tadzik I saw your talk submissions too, good stuff 11:56
masak FROGGS_: I've never used COBOL. only seen enough of it.
tadzik there was supposed to be an update last week, it's just that $life got in a way
FROGGS_ r: $life
p6eval rakudo b972ca: OUTPUT«===SORRY!===␤Variable '$life' is not declared␤at /tmp/znUry6ZClg:1␤------> $life⏏<EOL>␤ expecting any of:␤ postfix␤»
FROGGS_ hmmm, what's that?
masak tadzik: how does one submit a talk for the Polish Perl Workshop? 11:57
tadzik masak: through a talk submission link, I suppose :
isn't there one?
masak let's just pretend my fingers are glued together and I can't search for it :)
tadzik act.yapc.eu/plpw2013/newtalk
there we go :)
masak \o/
tadzik oh yes, please submit
++masak
tadzik food &
diakopter r: my $a = './'; require $a x 1000; 11:58
p6eval rakudo b972ca: OUTPUT«Could not find ./././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././././…
diakopter r: my $a = '../'; require $a x 1000;
p6eval rakudo b972ca: OUTPUT«Could not find ../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../../.…
diakopter r: my $a = ';'; require $a x 1000;
p6eval rakudo b972ca: OUTPUT«Could not find ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;…
diakopter r: my $a = ' '; require $a x 1000; 11:59
p6eval rakudo b972ca: OUTPUT«Could not find …
masak there should be a "this is an insane usage -- diakopter, is that you?" error message. :P
"Witaj na stronie Polskiego Perl Workshopu 2013!"
"Workshopu"! :D
sounds almost Japanese...
FROGGS_ :op('callmethod'), :name('&lunch'), :node(self) 12:00
diakopter r: my $a = '.0'; require "v6{$a x 1000}";
p6eval rakudo b972ca: OUTPUT«Could not find v6.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0…
masak diakopter: I think we should release the frozen spec with that version number. 12:01
diakopter r: my $a = '0'; require "v6{$a x 1000}";
p6eval rakudo b972ca: OUTPUT«Could not find v600000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000…
diakopter masak: yes.
masak by the time Perl 6{'0' x 1000} is out, sorear and I will already have absorbed a few stars.
diakopter r: require "v5"; 12:02
p6eval rakudo b972ca: OUTPUT«Could not find v5 in any of: /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/site/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/vendor/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/lib, /home/p6eval/.perl6/201…
diakopter r: require "your mom";
p6eval rakudo b972ca: OUTPUT«Could not find your mom in any of: /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/site/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/vendor/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/lib, /home/p6eval/.per…
diakopter r: require "anything"; 12:05
p6eval rakudo b972ca: OUTPUT«Could not find anything in any of: /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/site/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/vendor/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/lib, /home/p6eval/.per…
diakopter r: require "jack";
p6eval rakudo b972ca: OUTPUT«Could not find jack in any of: /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/site/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/vendor/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/lib, /home/p6eval/.perl6/2…
masak r: require "diakopter"; 12:06
p6eval rakudo b972ca: OUTPUT«Could not find diakopter in any of: /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/site/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/vendor/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/lib, /home/p6eval/.pe… 12:07
masak r: requires "sense or even intelligence";
p6eval rakudo b972ca: OUTPUT«===SORRY!===␤Undeclared routine:␤ requires used at line 1␤␤»
masak r: requires "sense or even intelligence";
p6eval rakudo b972ca: OUTPUT«===SORRY!===␤Undeclared routine:␤ requires used at line 1␤␤»
masak ugh
r: require "sense or even intelligence";
p6eval rakudo b972ca: OUTPUT«Could not find sense or even intelligence in any of: /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/site/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/vendor/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/lib,…
diakopter hee
cognominal__ r: my (&nique, &ta, &mère); nique ta mère 12:09
p6eval rakudo b972ca: OUTPUT«No such method 'Any' for invocant of type 'Parcel'␤ in block at /tmp/72HX4mW0RB:1␤␤»
12:09 daniel-s_ joined 12:13 preflex left
diakopter r: my (&nique, &ta, &mère) = ({;},{;},{;}); nique ta mère 12:14
p6eval rakudo b972ca: ( no output )
diakopter r: my (&nique, &ta, &mère) = {;} xx 3; nique ta mère
p6eval rakudo b972ca: ( no output )
12:17 preflex joined 12:25 preflex left 12:27 preflex joined 12:40 Heather left 12:41 snearch joined 12:42 SmokeMachine joined 12:44 sqirrel joined, SmokeMachine left
FROGGS_ r: require "my keys" 12:48
p6eval rakudo b972ca: OUTPUT«Could not find my keys in any of: /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/site/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/vendor/lib, /home/p6eval/nom-inst/lib/parrot/4.10.0-devel/languages/perl6/lib, /home/p6eval/.perl…
FROGGS_ should be "your keys" then 12:49
nwc10 masak: I'm probably biased, but the reasons I like perl 6 are the gradual typing and the first class grammars. And the additional reason I like Rakudo is that it's self hosting 12:52
heck. Probably a third reason is that the language is designed/intended to be lexically extendible
there are things that I don't like, which will probably remain even after the principal dislike is resolved (I have plenty of snow, but it's not Christmas) 12:53
12:55 cognominal joined, cognominal__ left
moritz_ having everything lexically (like imports etc.) is really great 12:59
nwc10 having the ability to change the grammar lexically should help future-proofing 13:00
hoelzro +1
(lexical stuff)++
nwc10 likely that one is key in the long term
lexical imports is useful in the medium term 13:01
jnthn Lexical stuff also works out very nicely implementation wise. 13:02
To the degree that I find anything that doesn't work lexically mostly a nuisance to have to deal with. :)
Leading example: packages :) 13:03
moritz_ this is kinda amusing, because Back In The Days, lexical classes din't work at all :-) 13:04
(when we still put methods into parrot namespaces) 13:05
13:06 shlomif joined, cognominal left
jnthn moritz_: Yeah, one of the biggest things that we've changed over time is making the toolchain good at lexical stuff 13:06
13:06 cognominal joined, moritz_ is now known as moritz
jnthn OTOH, we didn't actally get namespace stuff very right back then...since we didn't have the ability to do the separate compilation bits properly. 13:07
moritz which is kinda connected to the invention of Perl6::World
jnthn *nod*
13:07 rindolf left
nwc10 actually, heck, that's another useful thing - proper compilation, and linking 13:07
jnthn: nqp-jvm-prep still works (apart from the two cross-compile known problems) 13:10
arnsholt_ jnthn! o/ 13:11
jnthn nwc10: OK :)
Will look more at that after vacation :)
arnsholt_: o/
arnsholt_: How's things going?
13:14 snearch left 13:15 arnsholt_ is now known as arnsholt
arnsholt Pretty good 13:15
I pushed some commits yesterday (I think)
Think I've gotten the most important parts of the HLL config stuff in 13:16
jnthn I see you've been plugging away at that, yes
I'll take a closer look and probably contribute some in a week or so
masak jnthn! \o/
arnsholt Had to tweak the regex codegen stuff a bit. That was a bit scary, I must admit
jnthn Since getting that right will also be a Rakudo porting blocker
arnsholt: oh, where ooc? 13:17
o/ masak
arnsholt jnthn: github.com/perl6/nqp/commit/95084d...c2e023b366 13:19
13:19 SamuraiJack_ left
arnsholt Essentially, the code tried to fetch from an nqp::list directly into S and I registers 13:20
13:20 sahadev joined
jnthn arnsholt: which nqp::list out of curiosity? 13:20
If it's the bstack it's explicitly meant to be a list_i
arnsholt: oh, no, it's just the cursor start thing 13:22
arnsholt: yes, that's fine
13:24 bluescreen10 joined
arnsholt jnthn: Excellent. I was a bit wary, but it seemed like the only option 13:25
I also added a set_elems to REPROps_Positional (and dispatch to that from set_integer_native in sixmodelobject.pmc)
Loosely based on what I found in the JVM stuff
jnthn *nod*
Need an nqp:: op for that
arnsholt Troo 13:26
jnthn (has been on my todo list...didn't work out what to call it... :))
arnsholt Hehe
13:27 shlomif is now known as rindolf
arnsholt I could do that. Why not nqp::set_elems, since that's what the REPR method is called? 13:27
jnthn We never put _ in nqp op names ither than the _i, _s, _n suffixes.
But nqp::setelems would wfm
arnsholt Oh, right 13:28
Now I see why the Parrot ops and NQP ops differ like that =D
But I can implement that
Should I remove the set_integer_native VTABLE as well? (I guess I should)
jnthn I think you'll fidn a couple of places in Cusror.nqp that do the assign op on an array 13:29
And htat'd turn into an nqp::setelems too
arnsholt Yup
jnthn Yeah, no need to introduce more v-table things.
arnsholt Yeah, it was an assign in Cursor that made me implement it 13:30
jnthn yeah, I still have those commant out in nqp-jvm :)
Feel free to update things there too, if you have interest. :) 13:31
OK, back to afk vacation mode :)
arnsholt Really? I thought I found a call to set_elems in the same place in the JVM stuff
Enjoy!
13:35 baest_ is now known as baest, cognominal__ joined, jlaire_ is now known as jlaire 13:36 cognominal left
FROGGS_ use perl5; my $x = (1, 2); say %{ $x }; 13:36
("1" => 2).hash
\o/
arnsholt It's not supposed to do that in Perl 5, though, is it? 13:38
huf no, that should error out in p5 13:39
FROGGS_ damn
huf my $x = (1, 2); is my $x = 2; ()-s do nothing and , is the scalar comma
moritz p5eval: my $x = (1, 2); %{ $x }
p5eval moritz: 0
FROGGS_ I'm currently on the way to support scalar/array/hash dereferencing
moritz p5eval: my $x = (1, 2); [ %{ $x } ]
p5eval moritz: []
moritz huh, it doesn't error out?
huf p5eval: use strict; use warnings; my $x = (1, 2); [ %{ $x } ] 13:40
p5eval huf: ERROR: Can't use string ("2") as a HASH ref while "strict refs" in use at (eval 7) line 1.
moritz ah, strict.
huf i was assuming sanity :)
moritz not an entirely bad assumption
FROGGS_ huf: right now I have no hash-/arrayrefs, so it is just half-true
but at least @{ } calls .list 13:41
13:43 rindolf left 13:46 kaleem left 13:50 census joined
dalek : 5b9b18e | (Tobias Leich)++ | lib/Perl6/P5 (2 files):
added scalar-/hash-/array-dereferencing and removed debug cruft
 13:51
13:52 drbean left 13:53 cognominal__ left 13:54 drbean joined, cognominal__ joined
FROGGS_ I guess one of the important modules for Perl 5 is Exporter? 13:58
nwc10 what Exporter does is very important. But it could have a lot better interface 13:59
that it does it in pure Perl is also nice
FROGGS_ nwc10: I dont want to improve all modules that are on cpan, I just want to make a few of them work
nwc10 ah OK. Have fun with typeglobs then :-) 14:00
FROGGS_ hehe
>.<
daxim that better interface is p3rl.org/Sub::Exporter - if you want to apply it to an existing Exporter module, use p3rl.org/Sub::Import 14:01
14:01 skids joined, xalbo_ is now known as xalbo 14:05 PacoAir joined 14:23 Psyche^ joined 14:24 Patterner left, Psyche^ is now known as Patterner 14:31 xilo joined 14:37 cognominal__ left
masak just suddenly had an idea for a game: gist.github.com/masak/5237570 14:41
discuss. ;)
(I've played Nomic twice in my life. it has made me interested enough in the game to come back to it now and then.)
FROGGS_ masak: a 300-point-bonus when you already win with 200 points? 14:43
14:43 cognominal joined
masak yes. 14:43
better get consensus to change that rule... :) 14:44
or someone will win just by showing up and voting early.
FROGGS_ masak: so, the game is about playing and modifying the game? 14:45
masak indeed.
FROGGS_ so there is no 3d-gfx at all?
:P
masak heh.
the patches voted on determine what slightly different variant will be played in the next round.
FROGGS_ do I get extra points for being the first participant? 14:46
moritz so there will be a language that describes the game rules and patches to the rules
masak more like an API. written in Perl 6. 14:47
I don't see why patches couldn't conceptually be git commits.
the game could be backed by a git repo somewhere.
moritz ah, haven't read to the end yet
masak proposing a patch would be comitting in a newly-created branch.
winning a vote would merge the branch. (always a fast-forward) 14:48
moritz sounds geeky :-)
FROGGS_ moritz: -10 points!
masak oh, it is.
FROGGS_ # read to the end
:o)
moritz btw I'm against making stuff immutable 14:49
FROGGS_ do I loose points for being sarcastic?
moritz FROGGS_: not at all!
FROGGS_ \o/
masak the original Nomic had the problem of inexactitude of language. so it had a complicated-ish judging process. a Perl 6 nomic, running the rules, wouldn't have that problem or the need for such a process.
moritz: yeah, "against immutable" is my feeling too. ooc, why are you against it?
FROGGS_ I'm in btw 14:50
moritz masak: because the reviewing process should decide whether stuff can be changed or not
masak FROGGS_: good to know. it's just a glint in my eye so far. but would be a nice stress-test for Perl 6. :)
moritz masak: all the power to the players!
masak :)
I think trying without an immutable layer makes a lot of sense.
now, given what you know about the game, how would a session of this game play out? 14:51
how would someone win?
as I said, I've been in two live games, so I know a bit about what happens.
14:51 cognominal left
moritz probably by others growing tired of monitoring everybody's score 14:51
masak heh. 14:52
no, assume sufficiently interested players. that isn't always the case, but the assumption makes the question more interesting.
I almost said "sufficiently rational players", but that's not quite it.
moritz well, one option to win is to slowly accumulate points, and have all the others dead-lock each other so that they can't reach an agreement to stop you 14:53
masak yup.
basically "divide and conquer" / "pit everyone against each other".
anything else come to mind? :) 14:54
moritz I also wonder if there should be a bonus/malus for accepted/rejected patches
masak ooh. yes.
masak adds that
moritz well, somebody could also add it as a patch :-)
that would have been my first patch, fwiw :-)
oh, and I'm in.
14:54 wingfold joined
masak hah. :) 14:55
dang, now I have to build this ;)
three players is kind of a minimum.
masak adds "Perl 6 Nomic" to his growing TODO list 14:56
I think I could codify the *rules* as Perl 6 code in an evening or so. 14:57
no more guesses for how to win? I know two more player types that I've seen in the wild.
moritz well, there can be aliances
there can be harakiri
masak yes, but I've yet to see a winning strategy emerge out of an alliance or a harakiri. 14:58
moritz oh, and one more thing (not quite related to winning): there needs to be a deadline for patch proposal too
masak I thought about that.
decided not to add it, and to make it an exploitable problem in the game. 14:59
moritz ok
masak a two-player alliance could easily vote through anything at 23:50 on a Sunday.
moritz that would make it kinda time consuming on Sunday nights
masak heh. maybe make it mid-day some better day... :) 15:00
anyway... the two remaining strategies I know of are: (1) gradually increase the complexity of the game, hoping to win by out-complexifying everyone else (let's call this strategy "the yudkowsky"), and (2) playing just for lulz, intentionally proposing/approving things that are more amusing than strategic, hoping to get people off-balance ("chaotic") 15:02
moritz submit patches too late for anybody to vote on, for the not-voted penalty on them 15:03
15:03 cognominal joined
masak yep. 15:03
moritz *force
masak no, I won't fix that one.
moritz masak++ "the yudkowsky"
well, there can always be patches that try to fix that. 15:04
also interesting questions: can submitted patches be refined?
(ie push more patches the same branch/pull request) 15:05
15:05 dwoldrich joined
FROGGS_ I'd say yes, as long as there is no vote 15:05
masak not in the initial ruleset.
but I guess I should make it easy to add such logic.
cognominal masak, you game is more funnier if you patch the RNA code of a virus, use the Venter's "digital to biological converter" and inoculate the result to another player. the winner is the last player alive edge.org/conversation/what-is-life
masak cognominal: :P 15:06
cognominal s/more//
15:06 xantamoc joined
FROGGS_ hehe 15:07
I'd be out
masak the challenge in Nomic is making the game "take off", giving it a personality that makes people interested in the game's evolution.
cognominal real world virus are slow to act because it is bad to kill the host before infecting others. In this game you want fast acting viruses
FROGGS_ .oO( R.I.P. Perl 6 devs. They all died on a fast killing virus they self created ) 15:11
15:13 cognominal left
geekosaur latest incarnation of core wars? 15:14
15:14 cognominal joined
masak geekosaur: Nomic and core wars definitely have commonalities. but I'd say Nomic is more loopy than core wars. 15:15
anyone know what's going with Philippe's 'is export' problem on p6l?
geekosaur aside from the typo? 15:17
(if there has been a followup, it hasn't reached me yet) 15:18
masak oh, there was a typo?
geekosaur SalesReportGram called, SalesExportGram exported 15:19
masak yes, how about that.
if no-one has replied that, I could do so.
geekosaur I did already
masak oki
geekosaur++
geekosaur maybe the list is slow
masak yes. it is.
geekosaur ok, they fixed that, now get another error I wouldn't know about (missing parsefile method) 15:23
moritz r: say dir 15:25
p6eval rakudo b972ca: OUTPUT«star src .bashrc nom-inst1 toqast test3.pl .profile t nom-inst2 nom-inst toqast-inst toqast-inst2 examples Makefile .gitignore lib test2.pl .local bin rakudo p1 VERSION simple-tests .perlbrew std obj main.pl .lesshst nom niecza test.pl .cpanm log .bash_logout run T…
moritz r: grammar A { token TOP { .* } }; say A.parsefile('.bashrc')
p6eval rakudo b972ca: OUTPUT«｢# ~/.bashrc: executed by bash(1) for non-login shells.␤# see /usr/share/doc/bash/examples/startup-files (in the package bash-doc)␤# for examples␤␤# If not running interactively, don't do anything␤[ -z "$PS1" ] && return␤␤# don't put duplicate lines in the history.…
moritz sounds like he's calling the method on the module, not on the grammar 15:26
masak yes. 15:27
hoelzro are IO::Socket and friends guaranteed to be provided by a conforming Perl 6 implementation?
15:29 rindolf joined
hoelzro ah, nvm 15:30
15:31 cognominal left 15:35 dwoldrich left 15:37 cognominal joined 15:50 lichtkind joined
sahadev hello #perl6. i have a question about implicit coercion (of Str to Int) in case of arithmetic operators. In case of most operators the coercion does take place, but div and mod seem to be exceptions. ("42" + "2" is ok, but "42" div "2" errors). why were these two operators designed as such? especially considering their symbolic counterparts (/ and %) seem to DWIM? 15:54
moritz sahadev: Perl 6 has two types of operators. Most (like % and +) coerce their arguments 15:55
sahadev: and then there are a few operators that always take input of the same type, and produce output of the same time (div and mod, for example)
masak r: 007
p6eval rakudo b972ca: OUTPUT«Potential difficulties:␤ Leading 0 does not indicate octal in Perl 6; please use 0o07 if you mean that␤ at /tmp/fASOXE1sUd:1␤ ------> 007⏏<EOL>␤ Leading 0 does not indicate octal in Perl 6; please use 0o07 if you mean that␤ at /tmp…
moritz so, if that's not what you want, use / and % instead
masak why do I get the warning twice? :) 15:56
moritz masak: once for every 0?
r: 0007
p6eval rakudo b972ca: OUTPUT«Potential difficulties:␤ Leading 0 does not indicate octal in Perl 6; please use 0o007 if you mean that␤ at /tmp/JqbucXY9nL:1␤ ------> 0007⏏<EOL>␤ Leading 0 does not indicate octal in Perl 6; please use 0o007 if you mean that␤ at /…
moritz afk
sahadev moritz: i have observed the same. i am curious about the reason why div and mod were designed to be non-coercive. 15:58
arnsholt Whoa. My NQP compiled itself! =C 15:59
s/C/D
cognominal arnsholt, what does you new NQP?
arnsholt makes test and sees what falls out
I'm working on making nqp::list be a sixmodel object, rather than a ResizablePMCArray 16:00
Bringing the Parrot side of things more in alignment with how things look on the JVM side, as it turns out. Although my reasons for working on it are not actually related to that
Lots of failed tests, though =) 16:01
FROGGS_ arnsholt++ 16:02
16:04 cognominal left 16:05 FROGGS_ left
masak _sri++ posted, in the middle of yesterday's failed discussion on #parrot, nodejs.org/api/ -- a possible goal-post for Perl 6 to reach in terms of "production-ready" features. just thought I'd repost it here so it doesn't get lost in the noise. 16:05
tadzik hello hello
16:05 arlinius joined
hoelzro masak++ 16:05
I had an idea for the REPL yesterday 16:06
masak also,
<sri> speaking as an outsider... i get the impression rakudo development is too focused on polishing the core language, and doesn't pay enough attention to the stdlib
PerlJam sahadev: whenever I notice things like that, I always wonder if there's some easy way to tell the difference a priori (without creating a test case).
hoelzro masak: I echo that concern
however
that's not a problem if others (like myself) are willing and able to polish the stdlib
which is why I want to write a POSIX module 16:07
masak oh, it's still a problem :)
it's a thing which lots of people have to spend lots of time solving.
and awareness of the specifics of the problem, and its magnitude, is a necessary first step.
PerlJam masak: did you notice how many of the things in the nodejs api relate to IO?
hoelzro s/not/not as much of/
perigrin as PHP has shown you can simply make the stdlib part of the core language. 16:08
perigrin hides.
masak PerlJam: and we're very weak on IO.
PerlJam yep :(
16:08 cognominal joined
masak we got DateTime/Temporal settled. it works. it's a success. 16:08
next up: IO.
let's do this, people.
sahadev PerlJam: that ("easy way to tell the difference") would be useful. 16:09
PerlJam sahadev: yeah ... I don't have an answer in this partcular case because I don't think I ever noticed that quality of div/mod 16:10
hoelzro I'm also working on IO::Socket::SSL
(which ties into my earlier question on IO::Socket and friends)
arnsholt hoelzro++ # stdlib 16:11
16:11 jeffreykegler joined
hoelzro do we have a "Perl 6 production readiness" clock yet? 16:11
if so, I think we should start documenting what parts of the stdlib need work
ex. LWP::Simple is...nice, but far too rudimentary for real world usage
masak my first thought is that a clock would be nice. my second thought is that it may do everyone a disservice to project production readiness down into one dimension. 16:12
hoelzro ok
jeffreykegler masak, moritz: I do not think it correct to call yesterday's #parrot exchange "failed". You and allison stuck to technical issues, and stated them clearly.
hoelzro s/clock/checklist/ then =)
masak maybe we'd be better off presenting a few orthogonal dimensions. 16:13
jeffreykegler The result serves the reader well, though I imagine it was painful for participants
PerlJam masak: as long as we have a clear definition from our perspective what "production readiness" means and can articulate that to others, I think it will still be a win.
perigrin masak: two dimentions ... a clock is a vector no?
masak jeffreykegler: oh, it was failed. neither one got through to the other, and all that was created was blood pressure.
perigrin people are expecting you to boil it down to one dimension already .... a boolean.
masak jeffreykegler: I have more things to add to that discussion, but I'm consciously staying out of re-opening it.
PerlJam perigrin++ indeed
masak PerlJam: yeah, that's the goal. 16:14
hoelzro masak: didn't you post a gist of "to dos" for production readiness?
masak I was going to :/
others did.
it's in the backlog somewhere.
jeffreykegler Sometimes agreement is not the best outcome. Differing approaches can serve the community.
16:14 domidumont left
PerlJam jeffreykegler: YOU'RE WRONG! ;-) 16:15
masak autopun ;)
jeffreykegler: I'm the guy who wrote strangelyconsistent.org/blog/perl6-...ld-as-perl -- that blog has been described as "sad but hopeful". I believe I am not "in denial" about anything.
hoelzro ugh, backlog searching =/
masak hoelzro: I could find it for you.
hoelzro already grepping =) 16:16
16:16 colomon joined
hoelzro aw, twas moritz 16:18
16:18 cognominal left
masak jeffreykegler: allison tries to take a pragmatic approach to the state of Perl 6. in doing so, she oversimplifies and uses old data in a way that misses what we've accomplished already. which isn't a lot, but it's something. 16:18
PerlJam hoelzro: you can almost guarantee that if you think masak did something but it wasn't him, it was probably moritz :)
hoelzro: vice versa too
masak or tadzik.
hoelzro PerlJam: my thoughts exactly =)
its because all(<masak moritz>) ~~ /^m/ 16:19
masak hoelzro: I had this idea for a site: github.com/masak/production-readiness
hoelzro I forked moritz' gist: gist.github.com/hoelzro/5238329
masak hoelzro++
tadzik tadzik what?
oversimplifies and uses old data?
ah 16:20
masak no, realizes my ideas :)
tadzik: but for all I know, you oversimplify and use old data, too. you inconsiderate clod! :P 16:21
masak .oO( by a 2004 census, virtually everyone oversimplifies and uses old data ) 16:22
16:22 simcop2387 left
arnsholt masak: Do they make up statistics on the spot as well, by any chance? =D 16:22
masak only about 77%
arnsholt ^_^ 16:23
masak and the remaining 43% are really bad at match.
math*, g'ah
PerlJam that was an autopun of sorts 16:24
tadzik :D
hoelzro fortunately, I'm going on a two week vacation in the US tomorrow, so I'll have to work on stuff to stay sane =) 16:25
that, or play (Star|Mine)craft
kresike bye folks
16:25 kresike left 16:30 simcop2387 joined
census hi masak! someone say census? :) 16:30
16:30 cognominal joined
masak haha :) 16:31
census: hi :) nice to see you.
rurban masak: diakopter submitted a YAPC::US talk for you. Cannot you login by yourself? www.yapcna.org/yn2013/talk/4697 Zero to Perl 6 [3 hour training] 16:33
And more for you and TimToady as well 16:34
raiph r: module A { grammar A {} }; say A # (A) -- what's best way to find out if A is a module or a grammar? 16:37
p6eval rakudo b972ca: OUTPUT«(A)␤»
16:38 cognominal left, cognominal joined
PerlJam raiph: look at it's HOW maybe? 16:38
s/it's/its/
raiph d'oh. thanks 16:39
masak rurban: I'm aware of the talk. I haven't gotten around to filling it in yet.
rurban: I learned on Friday that I'm going, and have been busy/distracted since.
rurban good
masak rurban: is your criticism limited to my ability to log in? because I can log in. ;)
(I have the talk edit page open in a tab, waiting for me to get to it) 16:40
PerlJam
.oO( criticism? )
masak "Cannot you login by yourself?"
rurban no criticism. diakopter submitted a whole bunch of p6 talks so I wasn't sure.
masak yeah, I'm fine with delegating the handling of our track to him.
diakopter++
rurban 19 p6 all on diakopter talks I believe 16:41
19 p6 talks all by diakopter
masak I wasn't aware that they showed up as being "by diakopter".
mine seems to have the right speaker -- me. 16:42
so not sure what you mean.
rurban The problem is only the verification. the author should verify, not the submitter
good, all fixed. so it seems. thanks for checking 16:43
masak ok.
moritz ashleyde1: if div and mod were coercive, they'd be / and %
ashleyde1: so they were designed to be different, to give you options
16:43 thou joined
PerlJam moritz: so ... how can you tell the coercive ops from the non-coercive ones? Do you just have to memorize a list? 16:44
moritz: (or read docs, or perform test, or ... )
anocelot PerlJam: They're the ones that have the optional "That's some nice code you have there; it would be a shame if something were to happen to it" clause. 16:45
moritz PerlJam: there are heuristics, but it boils down to memorizing
masak anocelot++ # :) 16:46
16:46 SamuraiJack_ joined
masak anocelot: nice snowclone you've got there. would be a shame if something were to... happen to it. 16:46
anocelot masak: XD 16:48
masak .oO( nice Y combinator you've got there. would be a shame is something were to &?ROUTINE() # nooooooo ) 16:53
suck in an endless tailcall of mobster intimidation. 16:54
stuck*
my "spell correctly" key is broken today, it seems. 16:55
arnsholt Sigh. GCC's handling of undeclared types is bad 17:02
17:04 jeffreykegler left
moritz develop with clang instead? 17:05
17:06 FROGGS joined
arnsholt I should probably look into that 17:09
pmichaud_ good afternoon, #perl6
moritz good pm, pm
masak pmichaud_! \o/
17:09 pmichaud_ is now known as pmichaud
pmichaud ooh, looks like interesting backscroll to read. reading. 17:12
moritz and, interestingly enough, I've discovered two sink-related bugs ince our last longer discussion 17:13
though neither related to emptying arrays
FROGGS hi pmichaud
moritz (one in Net::IRC::Bot, one in rakudo)
pmichaud oops, I have to leave. bbiaw
arnsholt Remind me, what're the proper Perl 6 semantics for deleting the last element of an array/list? 17:17
I know Perl 5 shrinks the array in that case, but ISTR Perl 6 not having that special case...
moritz nr: my @a = <a b c>; @a.delete(2); say @a.elems
masak r: my @a = 1..10; @a.pop; say @a
p6eval rakudo b972ca, niecza v24-35-g5c06e28: OUTPUT«2␤»
rakudo b972ca: OUTPUT«1 2 3 4 5 6 7 8 9␤»
moritz ashleyde1: it seems that Perl 6 shrinks too, in imitation of perl 5.
sorry, meant arnsholt
masak it seems I misunderstood the question ;) 17:18
moritz++ # didn't
arnsholt Right. Thanks!
17:20 Targen_ left 17:25 wk left 17:27 fhelmberger left 17:33 xantamoc left 17:35 spider-mario joined
arnsholt Three more test files to go! 17:38
17:39 xantamoc joined
FROGGS Is here somebody who would volunteer to add the test-fudge-mechanism to v5? 17:44
arnsholt Oh. The NQP test suite reveals that delete should actually resize in any case! 17:45
FROGGS arnsholt: well, if it really resize the allocated memory area is another story (if your questions was about that) 17:46
17:47 cognominal left
PerlJam Perl 5 doesn't resize the allocated memory when you shrink an array, so I don't think he was talking about that. 17:48
arnsholt No, not the allocated area. After deleting at index 1 in an array with 4 elements, it expects the number of elements to be 3 17:51
pmichaud back again 17:53
17:54 cognominal joined
PerlJam arnsholt: I don't think Perl 5 does any kind of optimization there either. 17:55
huf arnsholt: delete how? 17:56
pmichaud ...delete should resize?
PerlJam well ... what did you mean by "delete"?
moritz nr: my @a = <a b c>; @a.delete(2); say @a.elems
p6eval rakudo b972ca, niecza v24-35-g5c06e28: OUTPUT«2␤»
pmichaud I don't think delete resizes for elements removed from the middle.
moritz only at the end
huf and even that feels a bit bonusy to me :)
arnsholt pmichaud: Then t/nqp/59-nqpop.t is wrong 17:57
huf but then, delete on arrays feels bonusy to me :)
moritz it feels like taking the Perl 5 semantics without thinking about it
17:57 kaleem joined
arnsholt This is in NQP mind, not real Perl 6 17:57
pmichaud I mean in Perl 5. Checking.
arnsholt Oh, right. Sorry
Yeah, I'm pretty sure Perl 5 doesn't resize when you delete in the middle 17:58
moritz p5eval: my @a = <a b c>; delete $a[1]; \@a
p5eval moritz: ERROR: Can't locate File/Glob.pm in @INC (@INC contains: /lib) at (eval 7) line 1.BEGIN failed--compilation aborted at (eval 7) line 1.
moritz p5eval: my @a = qw<a b c>; delete $a[1]; \@a
p5eval moritz: ["a",undef,"c"]
arnsholt You'd have to splice to do that
moritz aye
PerlJam I wonder why Perl 5 doesn't shrink the array from both ends since either way it's just a pointer move 18:00
pmichaud I agree that the nqp test looks wrong to me. 18:01
huf PerlJam: because shrinking from the end doesnt disturb the other elements
moritz OTOH, on the NQP level, what would be the point of having a delete opteration that doesn't shrink?
pmichaud Hope I'm not the one that wrote it.
arnsholt Hehe
pmichaud moritz: it can result in a shrink of it's the last element.
*if 18:02
arnsholt To be fair, when that test was written, it expected the list to be backed by an RPA, which I assume has those semantics
moritz after all you can implement Perl 6 level deletion (resetting) with simple element assignment
pmichaud and jnthn and I were discussing that deletepos might not be needed (but that wasn't ever resolved) 18:03
oh, and nwc10, also.
arnsholt Oh, I see. So that's why the JVM code doesn't have a delete_pos =)
pmichaud okay, I'm not the one who wrote those tests. phew. 18:04
now, the question is, why does it pass?
diakopter elven magic 18:05
arnsholt I'm assuming that's what RPA does
Gimme a sec, and I'll check the code
18:05 daxim left
arnsholt Yup. RPA's delete_keyed_int copies all the elements above the deleted element down a slot 18:06
And since nqp::list historically was just an RPA in disguise, the tests are like that 18:07
18:07 dmol joined
pmichaud yeah, apparently that's what RPA does. 18:08
PerlJam do those semantics blead through to rakudo? (I would guess so) 18:09
pmichaud well, not really. 18:10
arnsholt AFAICT, there's no code in NQP (apart from the tests) or Rakudo that uses deletepos
pmichaud right.
and I'm not sure what QRPA does.
arnsholt And nqp-jvm-prep has no mentions of deletepos anywhere that I can find
QRPA doesn't implement delete =) 18:11
pmichaud anyway, deletepos, when/if it exists, will have the p5/p6 semantics and not the parrot ones.
arnsholt Right, right
pmichaud returns to reading backscroll.
18:15 dakkar left
pmichaud looks like I need to read "yesterday's #parrot episode", too. 18:17
moritz not much to gain there, IMHO
diakopter yeah, nothing novel there
18:18 SamuraiJack_ left, SamuraiJack__ joined
PerlJam you'll just end up frustrated like masak I'd warrent 18:19
er, warrant
18:19 cognominal left 18:20 kaleem left 18:21 Chillance joined 18:24 cognominal joined, Targen joined
diakopter rurban: :P to be fair, I submitted only 14, and 4 of them are just lightning talks, another is an official scheduled training, and 2 of them aren't really about Perl 6 (masak's non-training ones) and only 2 of the rest are with me as speaker... sorry it's odd-looking; hopefully it's cleared up now 18:25
er, 13 18:26
arnsholt Last test file! 18:27
rurban diakopter: no prob, just wanted to crosscheck :)
pmichaud PerlJam: no, I'm not frustrated by it, because I've already explored the inconsistencies in allison's statements before. :) 18:28
arnsholt All green! \o/ 18:30
Apart from the ones that are skipped. Best check those first
pmichaud whenever I ask people "What is missing from Rakudo (Perl 6) to become 'production usable'?" I get back the answers "speed, stabilility, libraries, and a transition path from p5". We've definitely focused on the first two or three of these over the past 2.5 years.
masak aye. 18:36
I guess I'm a latecomer learning about allison's stance. :/ 18:37
pmichaud but yes, some of the comments on #parrot are really.... bizarre. 18:39
masak tl;dr: "Perl 6 has no users" -- "yes, it does" -- "let's not argue semantics. Perl 6 doesn't ship" -- "yes, it does!" -- "you're arguing semantics again" 18:42
geekosaur wat 18:43
diakopter masak: I detect a certain amount of semantics in the fact that you're quoting. :)
geekosaur "semantics" does not mean "doesn't agree with me"
sorear good * #perl6
masak sorear! \o/
sorear pmichaud: what's up, down, and sideways today? o_O
masak I guess the real conclusion to draw is that we should get that production-readiness page up, and soon.
pmichaud the "Perl 6 has no users" and "Perl 6 doesn't ship" arguments come from come from wanting to judge p6 by p5 standards of success. That's not necessarily wrong, however there's also a tendency in those quotes to want to treat epsilon the same as zero.
moritz geekosaur: after that exchange I left #parrot to cool down 18:44
[Coke] I think a target goal of "MOAR USERS" is a good thing we can all agree on.
moritz [Coke]: no doubt
pmichaud [Coke]: I agree that "more users" is important. But that has to be translated into smaller milestones.
masak pmichaud: yes.
moritz also "more possible use cases", which is nearly the same
masak pmichaud: (to treating epsilon as zero)
moritz which is where speed, modules, stability and IO come in again
pmichaud and so far the smaller milestones I hear are "faster, more stable, more libraries"
diakopter [Coke]: I think I can agree on "MOAR USERS" 18:45
masak pmichaud: I see what allison is doing there, and the point she wants to make. I just keep holding her to higher standards than a rounding-down of epsilon.
hoelzro (moar users)++
masak (moar everything)++
pmichaud masak: I suggest not doing that. :)
masak noted.
FROGGS well, I am working on the "more libs" with v5
moritz FROGGS++
pmichaud to look at it from the other side, they're wanting to hold us to a higher standard than "we're making progress". 18:46
moritz currently writes an article for a German perl magazine
masak pmichaud: yes, and I see that too.
pmichaud: and I agree. I want to hold us to a higher standard than that, too.
FROGGS jnthn does the speed thing (JVM), who does the more users?
moritz FROGGS: it's not something you can tackle directly 18:47
pmichaud FROGGS: there are marketing things that can be done to increase users, but until you solve some of the other basic hurdles any marketing efforts will tend to backfire
moritz FROGGS: it's a mixture of providing the prerequisites, and of motivating people to try out Perl 6
FROGGS moritz: well, I made two, and you at least one from what I know :o)
ya right, what was that quote again? "If you build it, they will come." 18:48
moritz 's goal is more like 1e5
18:48 Liz joined
sorear i think the operant phrase here is "tough act to follow" 18:49
pmichaud it's also to important to remember that for those of us working on Rakudo, we're not (yet?) at a stage where we feel our investments are likely to be lost. Parrot has a bit more of an existential crisis than Perl 6 and Rakudo do.
18:49 Liz is now known as Guest66032
FROGGS brb 18:49
18:49 Guest66032 is now known as lizmat
masak pmichaud: fair point. 18:51
I bet when the Roman empire was crumbling, senators went around muttering "well, I really wish those Greek dudes would own up to their part of the problems, too" 18:52
pmichaud :) 18:53
sorear pmichaud: you've already thrown away everything twice, you're just more used to starting over with a better plan? 18:55
pmichaud sorear: I think I'm out of the "starting over" phases, finally. :) 18:56
that said, I'm absolutely convinced that the places where we did "start over" it was absolutely best to go ahead and "start over". Maybe that cost us some users for the short-and-medium term, but long-term it was the right thing to do. (more) 18:57
sticking with what we had before would've been a recipe for "you have to start lines with a TAB in Makefiles because we already have a userbase that expects it".
I think we have a reasonable story these days on speed improvements. Where we fall short, story-wise, is something that demonstrates the "stability" (to end-users) of the language and implementations. 18:59
18:59 GlitchMr- is now known as GlitchMr
pmichaud being able to declare a 6.0.0 spec would probably help in that respect. 18:59
moritz pmichaud: re stability story, I'm currently exploring doing the IRC logging through Perl 6 19:00
pmichaud: having a continuous production setting will force me to react to stability issues with my stuff more directly
pmichaud agreed.
moritz pmichaud: say hello to ilbot6 19:01
:-)
it logs into a dev database
19:01 cognominal left
moritz and once I fix some minor issues, and it has run stable for a week, I'll use it for production 19:01
pmichaud +1
19:03 wingfold left
timotimo_ i'm considering doing a perl6 talk at the GulaschProgrammierNacht this year, but i don't know about what exactly 19:06
19:06 cognominal joined
timotimo_ but it would be fantastic, if i could get tthe iperl6kernel project going until then so that i can present using the ipython notebook 19:06
bbl
19:11 Gruu_ joined
pmichaud alas, I have to disappear for a while again -- bbl 19:11
moritz GPN++
moritz wants to discuss a completely unrelated topic 19:12
one of my favorites: Error messages!
hoelzro *sigh*
I haven't read the transcript of the parrot talk
but now chromatic has referenced it on Twitter =/
19:13 cognominal left
moritz one of the things that have always bothered me is that most error messages don't give the user enough context 19:13
like they say "Cannot open ~/.somepipe: permission denied"
now the user wonders: why the </$!#@@!> did the program try to open that pipe? 19:14
I'd want to know "I tried to open that pipe, and failed with "permission denied". I tried that because I want to talk to DCOP to get some configuration data from gnome" or something along these lines
of course that's way too much text to burden the lazy programmer with 19:15
19:15 cognominal joined
hoelzro how does the "default" EXPORT for a UNIT work? 19:15
moritz now if a programmer gets the same error, he reads the backtrace, and can roughly deduce what happens (if he happens to know the program)
19:16 SamuraiJack__ left
hoelzro how does the module inject definitions into the calling context? 19:16
moritz (phone, bbiab)
arnsholt moritz: I think I agree. Compiler/runtime diagnostics should be as helpful as possible 19:17
pmichaud (just read chromatic's tweets).... it would be far more helpful if chromatic and allison could point out the specific "unstable" features, rather than simply carping about them generically.
arnsholt I'm quite used to deciphering cryptic diagnostics from GCC (and don't get me started about TeX/LaTeX), but not having retarded diagnostics is definitely a plus IMO
moritz so, maybe we should give our routines :dba adverbs 19:18
lizmat :dba ?
moritz which could appear in user-readable backtraces
hoelzro moritz: that's a neat idea 19:19
moritz lizmat: "do business as", a feature we have in the regex engine
lizmat knows about database administrators, but wonders how they would become an adverb
aha
ok
19:19 Gruu_ left
lizmat has a backlog of catching up on synopses :-( 19:19
arnsholt moritz: That's actually a very neat idea. I like it! 19:20
moritz (me previous idea was to use contextuals, like my $*BUSINESS = 'reading config file') 19:21
(I think I like :dba better)
pmichaud gone again
19:25 cognominal left
lizmat $*DOING ? 19:26
moritz (back from phone) 19:27
lizmat dba seems too much like a job description to me
moritz so, the idea is to have an end user's view on backtraces, which only considers routines marked in some way as being of interest to the user
lizmat: naming is up for discussion, as always :-) 19:28
lizmat but that could well be my mind needs reprogramming in that respect
moritz but first I want to get more general feedback on the idea
lizmat I like the idea
to an extent, this could replace at least partly internal documentation
moritz so, that's 2 +1s
lizmat I don't think my plusses count as 1 yet 19:29
more like 0.01 :-)
moritz good that we round up to the next integer :-) 19:30
timotimo_ i wonder if perl6 parsing could be made faster by either estimating the right path better to backtrack less often or make it cheaper to throw away data that was generated as side-effects of parsing
hoelzro so this "general dba" extension isn't just for compiler authors, right? you could do it in regular code?
moritz hoelzro: sure 19:31
hoelzro so what would that look like?
moritz something like sub write-config($config, $file) dba('Safing user config file') { # code here }
*saving
timotimo_ i like the idea fwiw
moritz and then if that sub dies, the backtrace would read
sorear
.oO( not my $*BUSINESS )
 19:32
moritz "Error while opening file $file for writing: permission deninend\nWhile saving user config file")
hoelzro ok, neat
that's an excellent idea
moritz++
19:32 census left 19:33 domidumont joined, sahadev left 19:35 sqirrel left
lizmat wonders how to handle multiple DBA's 19:35
moritz arrays exist
hoelzro looking at make $<pairlist>.ast.hash in JSON::Tiny...where does that .hash method come from? 19:36
Any?
19:36 cognominal joined
timotimo_ multiple DBA's would probably be more like a stack actually 19:36
moritz hoelzro: $<pairlist>.ast is a List
hoelzro ahhhh
moritz hoelzro: so likely from class List
timotimo_ how about assigning dba's to blocks inside fucniots,too?
hoelzro understood
moritz timotimo_: seems like overkill
timotimo_ fair enough
moritz timotimo_: considering that we usually don't want blocks to show up in backtraces 19:37
dalek nqp/vmarray-list: b754127 | (Arne Skjærholt)++ | src/ (3 files):
nqp/vmarray-list: Add the set_elems method to the Positional REPR table.
nqp/vmarray-list:
nqp/vmarray-list: Also implements that method in VMarray and dispatches the set_integer_native
nqp/vmarray-list: VTABLE to set_elems.
19:37 dalek left
arnsholt Yeah, I was afraid that was going to happen 19:37
19:38 dalek joined, ChanServ sets mode: +v dalek
hoelzro can we get generic dba in Rakudo before 2013.04? I want to use it already =) 19:38
moritz hoelzro: I'm working on a prototype :-)
hoelzro ;)
timotimo_ subscribed to p6u
hoelzro could it perhaps be defined in a module? 19:39
timotimo_ moritz: yes, that's a good argument
moritz hoelzro: that's what I'm working on :-)
hoelzro \o/
moritz++
moritz I don't touch the compiler when modules work fine
hoelzro amen
moritz because it takes so long to compile
lizmat wonders whether :dba shouldn't be in nqp
arnsholt Which repository does dalek live in? 19:40
:dba is probably best suited to be implemented in Perl 6
Or, the Perl 6 compiler 19:41
moritz arnsholt: iirc there's a "botnix-plugins" repo somewhere
lizmat I was wondering whether :dba could be helpful in debugging NQP problems
timotimo_ moritz: can you just define a sub trait_mod:<dba> and it'll work like that or how is that done without relying on "is"?
diakopter dalek could easily be reimplemented in something else, since sorear++ simplified it so all it does is send lines from a file to channels 19:42
moritz timotimo_: iirc rakudo doesn't do user-defined trait_mod's yet 19:43
timotimo_: so I have to hijack 'is' or 'does' or 'handles' or whatever
skids wonders at the etymology of dba (aside from the acronym meaning)
hoelzro $/ is lexical, right?
skids Does it hail from an old language?
arnsholt github.com/Infinoid/dalek-plugins perhaps?
hoelzro I can't address function a's $/ from b if a calls b? 19:44
arnsholt $/ is lexical, yeah
hoelzro hmm
arnsholt So if you want to use it from something your function calls, pass it as an argument
hoelzro is still trying to figure out how the heck to create his own EXPORT 19:45
arnsholt Or put it in a dynamically scoped var, I guess
moritz gist.github.com/moritz/5240038
hoelzro: through CALLER::<$/> maybe
but you shouldn't, if it can be avoided
[Coke] PERL 6 SCRIpT 19:46
arnsholt moritz: Cool!
moritz anyway, gist.github.com/moritz/5240038 is the first user-level backtraces prototype
hoelzro moritz: well, if I'm implementing EXPORT, how do I inject values into the calling context?
CALLER::<fork> = &fork; ?
19:47 FROGGS left
moritz hoelzro: by merging the EXPORT_hash branch from [email@hidden.address] into nom, and then returning a hash of (symbol_name => $symbol) pairs 19:47
timotimo_ hoelzro: there's a branch that makes (is supposed to make) EXPORT able to return a hash of symbols that will be installed
19:48 FROGGS joined
hoelzro oh 19:48
timotimo_ last time i checked, it only works well for types (or was that subs?)
hoelzro that's handy
moritz timotimo_: only for subs
and I'm supposed to work on it
hoelzro is it possible to do in mainline rakudo?
timotimo_ :)
moritz instead I prototype dba for routines
moritz must resist putting the hacking into the article he's writing just now 19:49
timotimo_ as long as you get both to work for the next rakudo compiler release ... ;)
19:49 japhb_ joined
timotimo_ at least I don't mind 19:49
hoelzro I prefer the return hash 19:50
but I want this damn POSIX module to start working!
moritz hoelzro: the problem with the CALLER appraoch is that the lexical scope you're in isn't strictly the caller of EXPORT
hoelzro ah, good point
moritz it's called from within the compiler for that scope 19:51
19:51 myname_ joined
hoelzro so my current EXPORT looks like this: sub EXPORT(*@imports) { for @imports -> $import { say "importing $import"; } } 19:51
and the printing never happens
I'm guessing that for does some sort of lazy magic?
or...?
moritz needs to be our sub EXPORT
and yes, add a 1; or so after it, just to be sure
hoelzro our sub EXPORT fails 19:52
moritz how?
hoelzro moritz: is that the reason, though? laziness?
Merging GLOBAL symbols failed: duplicate definition of symbol &EXPORT
moritz confused
hoelzro too
moritz oh
are you using 'lib' ?
hoelzro CALLER::<CALLER><fork> = &fork 19:53
?
moritz eg use lib 'some path';?
hoelzro from the importing script, yes
moritz ah
we need a better solution for EXPORT
hoelzro agreed =/
this is actually pretty painful
moritz I'd like to discuss this with jnthn++ soon
r: module A { method x() { } }; A.x 19:55
p6eval rakudo b972ca: OUTPUT«Useless declaration of a has-scoped method in module␤No such method 'x' for invocant of type 'A'␤ in block at /tmp/BqgbWp_mFD:1␤␤»
moritz that's why we can't make EXPORT a method
FROGGS timotimo_: GPN sounds nice
timotimo_ it's an excellent event, indeed 19:56
pmichaud I was going to attempt some troll hugging with chromatic++... but after seeing his latest tweets I'm not sure it's worth the effort. He's still complaining about the state of things as of two years ago. :-/
timotimo_ hoelzro: since you're the pygmens-perl6 guru; have you ever looked at CodeMirror? do you have an idea how that works? i believe it's the piece of ipython notebook that does the syntax highlighting and it probably lacks a perl6 highlighter (not asking you to do it, though. not just yet ;) ) 19:57
pmichaud anyway, we need to come up with some publishable measure of stability.
hoelzro has never heard of CodeMirror =(
timotimo_ no problem then
hoelzro heh
diakopter heads to twitter
timotimo_ (it's a javascript based code editor used in many, many places)
pmichaud also, I'm struck that "small stable core" and "get p5 folks to switch to p6" are mutually incompatible somehow.
[Coke] pmichaud: I think we have to focus our limited tuits on things we think will have an effect, yes.
hoelzro timotimo_: pygments was step 1 19:58
timotimo_ oh, i know this meme! step 4 is profit, right? :D
hoelzro step 2 is vim-perl
19:58 cognominal left
pmichaud I'm also curious about: 16:21 <kid51> The weird thing about the above discussion is: Parrot has fewer users than Perl 6, but we're not in denial about it. 19:59
my first focus was the "not in denial about it" part... but after thinking about it a bit more, it strikes me that Parrot has to have at least as many users as Rakudo does. :)
diakopter you mean a user of rakudo is a user of parrot..? he means project users
[I think]
pmichaud diakopter: if so, I think there's an implication that he somehow expected Parrot to have more users than just Perl 6, which wouldn't make sense. 20:00
not that any of this makes sense.
moritz
.oO( we've tried it with sense. It was frustrating. )
hoelzro so...there's no way of injecting things from EXPORT into the calling context yet? 20:01
if so, I'll have to just grin and bear it
pmichaud kid51 also repeatedly points to Moose, Catalyst, Dancer, etc. as being examples of p5 evolution within a "stable product", but doesn't recognize that those also were outgrowths of Perl 6.
hoelzro but I really want to write this POSIX module =)
nwc10 Catalyst was?
pmichaud well, perhaps not Catalyst. 20:02
Although I'm sure there are some p6 influencers there.
20:02 census joined
pmichaud my point being simply that p5 evolution hasn't taken place divorced from p6 20:02
hoelzro timotimo_: where else is CodeMirror used? 20:04
anocelot FWIW, TimToady mentioned that the p6 movement added a lot of energy to the p5 work as well. 20:05
20:07 philroc joined 20:08 philroc left
lizmat even though the PONIE project was stopped, it also still caused a lot of cleanup in the Jenga that the internals of Perl 5 are 20:08
so Perl 5 also benefitted a lot from that
20:09 cognominal joined
nwc10 not a *lot*, in absolute terms. But it revealed something specific and useful - that pretty much none of the XS code on CPAN cared about the exact binary layout of the data structures, as long as the macros kept working 20:10
Perl 6 was very useful for Perl 5 20:11
20:13 domidumont left
masak "was"? :P 20:13
diakopter masak: see, that statement could go either way [or both!] on which is over... 20:14
huf still is. it was also cancer because people backported features that dont really work in p5 without _all_ of p6 :)
looking at you, ~~
masak yeah. 20:15
geekosaur why tf did they do that, anyway? 20:16
huf *shrug*
i think p5's switch is cursed
first we had Switch.pm, and what a load of fuckup that was
then ~~ and given/when and ... yeah
hoelzro $*W is Rakudo-specific, right? 20:17
sorear another big issue is that a substantial amount of real-world p5 code relies on prompt DESTROY and a reference-counting VM
yes, $*W is a rakudo detail
looking especially at Scope::Guard, B::Hooks::EndOfScope, namespace::autoclean, EV 20:18
anocelot Update for TimToady friends: He's out of surgery and doing well.
cognominal oops, his prostate? 20:20
pmichaud anocelot: yay!!!
anocelot: thank you for the update, I had been thinking of him all day. 20:21
lizmat thinks healing thoughts for TimToady
sorear is that even public?
diakopter
.oO( "it is now..." ) (yes, it was)
cognominal lizmat++ # so do I
diakopter sorear: like all his other surgeries, afaik 20:24
FROGGS anocelot++ # thanks for the update 20:26
20:26 __rnddim__ is now known as lue
lue hello o/ 20:27
timotimo_ hoelzro: i thought github uses it, but i can't find any indication that that's actually true
FROGGS hi lue
anocelot FROGGS: My guess is *most* of us have been anxious all day, so yeah.
hoelzro timotimo_: github uses pygments 20:28
FROGGS anocelot: well, I know it should happen these days, but I wasnt sure it was today
timotimo_ hoelzro: for the code editor i mean 20:29
hoelzro oh, for that 20:30
20:31 cognominal left 20:32 kurahaupo joined 20:33 cognominal joined
[Coke] anocelot++ 20:34
20:34 Targen left
sorear wonders if anocelot is another of the wall kids 20:37
timotimo_ .o(all in all he's just a- nother kid from the walls) 20:38
masak .oO( anocer of that lot ) 20:39
[Coke] *groan*
20:40 Targen joined
FROGGS .oO( timotimo_s leave Wall's kids alove... ) 20:40
FROGGS skipped that "we don't need no ..." line 20:41
*alone
-.-
anocelot Oy. That was as bad as one of my jokes. :D But to answer your question, I am not. I'm a fellow Perl geek that happens to attend the same church. So I had some of the wall kids in my Sunday School class, if that counts. I teach Perl at the local community college. 20:43
moritz anocelot: did you happen to be on the picture with the '25 years Perl' cake? 20:44
geekosaur ...that's a bit off the Wall...
FROGGS I thought there was diakopter and his son on it...
anocelot moritz: Not that I know of. I know of the cake, but having not acually seen it, I assumed the cake was a lie.
PerlJam FROGGS: aye, it was. 20:45
lue
.oO(must... resist... can't...) So, I guess you're someone who could say "Anocelot about the Walls."
lizmat s/lie/fabrication/ ?
PerlJam anocelot: if you happen to see TimToady or his family before us, relay #perl6's best wishes :)
lizmat has gotten sensitive about calling things lies after a recent discussion on p5p
PerlJam lue: heh 20:46
moritz seconds that (best wishes)
masak thirds that 20:47
FROGGS best wishes from all of us I'd say 20:48
lizmat FROGGS++
anocelot That is certainly how I would phrase it. As opposed to "the following users were online and did not do any well-wishing, so you can think they are terrible: ..." :) 20:49
pmichaud TimToady already knows which of us are terrible anyway. :) 20:50
anocelot pmichaud: Isn't everyone terrible in their own special way?
PerlJam anocelot: Is that a confession? ;> 20:51
FROGGS .oO( There is more than one way to be terrible? ) 20:52
anocelot PerlJam: Sure. I'm a reformed PHP programmer. There but for the grace of Perl go I, or something.
pmichaud I'm not yet reformed. 20:53
anocelot pmichaud: Well, me either, technically, but you won't hear my saying that in public, that's for sure.
FROGGS hehe
anocelot ... wait.
Crap.
PerlJam
.oO( illformed? malformed? )
FROGGS takes notes about the devs who will willingly help with 'use PHP' 20:54
anocelot doesn't know if that's a joke. 20:55
pmichaud anocelot: I didn't hear anything. I did read something on IRC, though... :-P
lizmat has been considering implementing some version of BASIC as an excercise to start implementing Perl 5 in Rakudo
PerlJam anocelot: that's what makes it interesting--no one knows if it's a joke.
lizmat is that bad?
PerlJam lizmat: no.
pmichaud well, since BASIC was one of the core languages used to develop Perl.... it's probably a good start :) 20:56
lizmat wonders about GOTO
anocelot If we're going to consider the 'abuse v5' keyword, I think we can add some PHP capability there, too. Maybe as an Easter egg.
20:56 bruges left
anocelot lizmat: "GOTO is a four-letter word." 20:56
lizmat I knew I was bad :-) 20:57
FROGGS anocelot: 'use PHP' not really meant a joke
pmichaud From Programming Perl (the pink camel).... "BASIC/PLUS: Another ancient language (reminiscent of Indo-European) from which Perl derived exactly one idea. Well, OK, maybe two..." 20:58
20:58 bruges joined
timotimo_ which ideas were those? 20:59
pmichaud "statement modifiers and maybe formats" 21:00
(it's in the footnote at the bottom of the page)
21:00 dmol left
PerlJam pmichaud: you still have your pink camel? 21:00
21:00 myname_ left 21:01 dmol joined
pmichaud PerlJam: that's a book I will never willingly give up. 21:02
(so yes)
PerlJam boggles
sorear lizmat: all cakes are fabrications :) 21:03
pmichaud the glossary alone is special. 21:04
sorear *looks up the pink camel* 21:05
PerlJam So ... does anyone remember the book "Perl 6 and Parrot Essentials"? I wonder if we could get the Perl 6 chapters of that (about 70ish pages I think), and reform them into a "Learning Perl"-style book (or some other early Perl 6 book) 21:06
sorear PerlJam: you want one?
moritz PerlJam: they are in the mu repo somewhere, in PseudoPod format
21:06 skids left
PerlJam sorear: no, I have mine somewhere too, I'm just not sure of the location :) 21:06
sorear HAH
PerlJam moritz: really?
PerlJam looks
sorear (that's exactly how things work around here too) 21:07
(OTOH, I know I don't have a camel of any color, only a llama)
pmichaud I have the P6PE book here also :-)
moritz PerlJam: and the parrot part is in the parrot repo, I think it served as the foundation of the PIR book 21:08
pmichaud right next to "Mastering Regular Expressions", which I never really read and which explains my total ignorance of how regexes work :-P
anocelot FROGGS: If I can't be of use helping with PHP->Perl6, I can at least help test stuff. 21:09
lizmat late dinner& 21:10
masak moritz: ping 21:11
moritz masak: pong(ish)
masak moritz: check this out. gist.github.com/masak/5237570 -- it turned out *much* better than I had hoped. 21:12
FROGGS: you check it out too :)
FROGGS anocelot: well, in case of PHP I guess there is not even a spectest suite :o) 21:13
masak it's un-run, untested code without any of the dependencies implemented.
but it reads really well.
observers++
FROGGS masak++
PerlJam masak: looks ruby-ish.
21:14 spider-mario left
masak yeah, a method-y DSL will do that to code. 21:14
moritz masak++ 21:15
masak some would probably say that it looks JavaScript-ish or jQuery-ish.
PerlJam "whenever.a(Vote).changes('state').to(Finished).do" reminds me of cucumber :)
masak moritz: I'm amazed at how *well* this abstraction works. only the last couple rules even gave me pause.
moritz masak: it's about the length that I'd have expected :-)
lue masak: nitpicky line 60: "the the vote" (but otherwise masak++)
masak lue: thanks.
I realize now that we have to build this thing, if only so we can play it once :) 21:16
but I can't give a time line yet. too many other important things going on.
sorear where do we stand on "use COBOL;" ? 21:17
FROGGS sorear: go ahead :o)
masak heh, I found a bug :) at line 58. 21:18
it's a bit subtle.
21:18 census_ joined
sorear ran into an interface spec last week which looked like "05 FOO-BAR PIC 9(3). 05 BAZ PIC ZZZZ99V99." ; had quite a bit of trouble finding documentation for that syntax 21:18
21:18 census left, census_ left
arnsholt tries to build Rakudo on his NQP branch 21:19
21:19 census joined
geekosaur aaaa, cobol 21:19
masak sorear: looks lovely.
census geekosaur++
masak what a difference 50 years make. 21:20
geekosaur the sad part is I still understood it immediately...
lue Is this referring to that wayward use COBOL comment in S02 in any way?
pmichaud and here I thought the point of COBOL was to have a language that one could read like english. :-P
sorear masak: PIC ZZZZ99V99 is not particularly googleable (or ddgable), especially when the document doesn't tell you in advance that you need to find a cobol manual :(
PerlJam geekosaur: is ZZZZ99V99 leading zeros, and the V is the decimal point?
geekosaur virtual decimal point
PerlJam right 21:21
arnsholt Virtual decimal point?
pmichaud I'm frightened that I remember that.
geekosaur meaning, there isn't an actual decimal point printed, it's implied
21:21 sftp left
PerlJam arnsholt: fixed point arithmetic. 21:21
geekosaur like storing a money value in cents
arnsholt Ah, right
21:21 sftp joined
lue
.oO(With implied decimal points, how can I resist checking out COBOL?)
 21:21
arnsholt Digitale ferdigheter 21:22
Whoops
21:23 pmurias joined
masak sorear: did you know that Hildegard von Bingen was once known as Hildegard von Googlen before Microsoft bought the rights to her? :) 21:24
pmurias FROGGS: should the simple perlito test also end up in roast5? 21:25
21:25 cognominal left
timotimo_ i wonder if i'll get a bit of discussion/feedback for my ADT work if i crosspost it to p6u? 21:28
sorear masak: did not know about either
masak then the pun is less funny, I suppose.
FROGGS pmurias: well, I would keep them seperate? like the simple tests are in v5 directly, and only the original perl5 tests are in t/spec 21:29
hehe
pmurias what's the point of roast5? 21:30
FROGGS I need to put the perl 5 tests somewhere and I need to be able to fudge them 21:32
otherwise I can't track regressions
nor recognize them
lue I'm guessing the only way to modify how Perl 6 handles undefined barewords is through NYI slangs, correct? (Specifically, do an extra step with the bareword before throwing an undefined error) 21:33
masak lue: think so. 21:34
lue: that sounds like a slang-y thing.
21:35 cognominal joined
lue I wonder how painful it would be to try to implement slangs... (my guess: very, esp. with my excellent knowledge of NQP and Rakudo :) ) 21:36
sorear
.oO( no strict 'subs' )
masak lue: by all means, investigate what remains to get slangs at this point. I sure would like a status update of that kind. 21:37
21:37 rindolf left
masak lue: ask pointed, curious questions. get a map of the territory. 21:37
diakopter also, more questions
sorear #1 problem is that the spec has a gaping hole in it
lue Sounds like a nice way to spend the week :) /me can't wait to make this line reality: say "hello' ~~ ☃hello☄; 21:38
sorear the spec lets you define the syntax of slangs. ok
diakopter listens to sorear
sorear there's no specced way to define semantics though!
lue: did you ever get to see C/2011 L4
lue I don't believe I know what that is. 21:39
21:39 kaare_ joined
lue Oh, I didn't see it. (Remember reading about the comet though.) 21:39
sorear it's a ☄ that was very visible a week or two ago. 21:40
sorear did
PerlJam sorear: pics or it didn't happen! ;)
pmichaud I missed it... too many things happening here at sunset for me to see it.
people I know got pics, though :)
masak yeah, 'slangs' need to be grammar+action, not just grammar.
actions*
PerlJam pmichaud: Grady Blount got his on TV (kiiitv.com) 21:41
sorear PerlJam: if I wanted pics I would just have gone online
pmichaud PerlJam: ah, that's the pics I saw :)
lue If only that comet happened when there was plenty of snow (there was some during the latter part of last week, but not much). Then I'd've been able to take that snowman/comet picture!
pmichaud lue: hahahahahahahahahahahah
masak :D 21:42
sorear it was of course very cloudy most of the relevant week
pmichaud wonders about photoshop.
lue I had the idea way back in 2010 (evidenced in a December blog post, but no reason to link to it) 21:43
C/2012 S1 may be the next chance
diakopter sorear: I still want the general termish:<| regex blah { some_pattern } |> (*@matches) { semantics } 21:44
'course, it would need the same context/compile-time as macros 21:45
anocelot If I run across a broken link on tablets.perl6.org, to whom should it be reported...?
tadzik lichtkind, I guess
anocelot Or, alternately, who do I talk to for fixit-yerself rights? 21:46
lichtkind tadzik What what? missed i something?
masak like, four lines of backlog :)
diakopter anocelot: I'll add you
e 21:47
anocelot diakopter: kk
diakopter er
yeah; sec
sorear favorite take on the cloud problem: twitter.com/BadPhysics/status/3127...7021126657
anocelot is expection "anocelot++" or some other addition joke. XD
diakopter what's your github ID 21:48
anocelot no existy
BUT you are on github: github.com/throughnothing/dotfiles...ors#L12304 21:49
lichtkind anocelot: where is it?
lue
.oO(Perhaps first designing a decent syntax for slangs would be in order)
 21:50
diakopter anocelot: apparently someone likes to color your nick bold red 21:51
anocelot tablet 3 toward the end is a link labeled 'go to the twigil chapter..' that doesn't include the "tablet-3-" prefix for variables.html
lichtkind thanks a lot
anocelot diakopter: Oh, right. What is my github name...? Hang on a sec. 21:52
masak lue: not so much syntax needed. we have regexes and grammars.
lue: but we *do*, as sorear pointed out, need a way to hook actions into a slang, not just a grammar.
lue yeah. Probably just need to clarify the actions side of it. 21:53
masak that would be a great start.
anocelot diakopter: (DM'd.) 21:55
21:55 Targen left
diakopter anocelot: see privmsg 21:55
lizmat masak: in your proposal for Nomic, why: $player.votes_made($player.votes_made() + 1); instead of $player.votes_made++ or $player.votes_made += 1; ? 21:56
masak lizmat: excellent question! I was waiting for that one.
lizmat sorry to have kept you waiting
masak :)
lizmat: because the way everything can hook into the changing of values is to have everything in the system be "observables", subscribable getter-setters. 21:57
lizmat: the two alternatives you suggest wouldn't trigger the proper reactions in listeners.
in other words, it's the syntactic/API price one pays for having observables.
lizmat feels like a source for WTF in the future to me 21:58
masak now, having said that, I *guess* everything can be done through Proxy and FETCH/STORE...
geekosaur would comment such thing
masak and that would definitely look nicer.
lizmat: you're right. I'll change it to be easier. 21:59
diakopter anocelot: I need a github id, not an email address
masak I believe I can have my cake and eat it here.
it is, after all, Perl 6. :)
lizmat that would be nice!
lichtkind ancelot there are all sorts of broken links
i currently fix them all
masak lizmat: updated: gist.github.com/masak/5237570 22:02
lizmat masak: are "observables" documented anywhere? 22:03
diakopter anocelot: added
22:03 bluescreen10 left, leprevost joined, lustlife left
lue awful first pass at slangs: gist.github.com/lue/5241230 22:03
masak lizmat: no, but the tutorial at knockoutjs.com/ is very elucidating. 22:04
lizmat looks
masak lue: nope. doesn't work. :) 22:05
good try though.
anocelot lichtkind: Do you mind if I fix this one?
masak lue: even besides the fact that we like keeping grammars and actions as different containers, a token is just a method, so you have a namespace conflict on hand. 22:06
lichtkind i fix all :)
masak lue: examples of modifying the grammar with slangs: derive the main Perl 6 slang. override the method 'identifier' to accept smiley symbols in it. set the main slang to the new derived class. 22:07
example*
22:07 xilo left
anocelot ... so I shouldn't fix this one...? (He asks, nervously.) 22:07
lichtkind i make a push in 10 min 22:08
if its still there please do
diakopter anocelot: you can now design/specify Perl 6, too
anocelot diakopter: I noticed that.
dalek blets: 47fa454 | (Kevin Metcalf)++ | docs/tablet-3-variables.txt:
Update tablet-3-variables.txt

Link fix:
  [are listed their
table](ap-b-lookup.html#special-variables). To understand their secondary sigil [go to the twigil chapter of this tablet](variables.html#twigils).
  *** becomes ***
  [are listed their
table](ap-b-lookup.html#special-variables). To understand their secondary sigil [go to the twigil chapter of this tablet](tablet-3-variables.html#twigils).		 22:10
anocelot OH my. Sorry. >.>
masak accidental push? 22:11
anocelot And I really have to be sure to read the whole back scroll. Not just the "if it's still there please do" line. >.<
lue does augment allow you to replace methods (e.g. augment Str { method new() { } } would replace Str's current new)
anocelot Didn't know the entire update extended description would get dalek'd. 22:12
lichtkind anocelot: dont worry i merge 22:13
masak lue: not that I know. 22:14
lue: feels to me there's a 'method' keyword missing there. but even besides that, I don't think it's accepted syntax. :)
lue what missing 'method' ?
anocelot lichtkind: Thanks. *sees another typo and wanders off*
dalek blets: 1e9b5d9 | (Kevin Metcalf)++ | README.md:
Update README.md

your -> you're		 22:15
22:15 ashleyde1 left
lichtkind just a second i push 22:15
masak lue: oh, sorry, I read your example wrong.
lichtkind but you can see results when server does refresh 22:16
masak lue: I thought you meant to do something like 'augment method new { ... }'
lue: now that I understand your question: yes, 'augment' on a class would replace old methods in the class.
lichtkind anocelot: only every hour html will be compiled so please see if you bug is still in 5 min in git
masak rn: use MONKEY_TYPING; class A { method foo { say "OH HAI" } }; augment class A { method foo { say "OH replaced!" } }; A.new.foo 22:17
p6eval rakudo b972ca: OUTPUT«===SORRY!===␤Package 'A' already has a method 'foo' (did you mean to declare a multi-method?)␤at /tmp/PtxnBumBj0:1␤------> ␤»
..niecza v24-35-g5c06e28: OUTPUT«OH replaced!␤»
lue OK. That makes your "smiley identifier" example work.
anocelot lichtkind: Ah. OK. Will do.
masak lue: no, I meant using subclassing, not augmentation.
lue: overwriting the existing grammar considered antisocial.
lue: because we want the slang to be lexically scoped.
lue But, that's the only way slangs are used in the spec: perlcabal.org/syn/S02.html#Slangs (specifically the " by lexically scoped declaration of a new sublanguage:" line before the augment slang example) 22:19
dalek : 8a71fa1 | (Tobias Leich)++ | / (4 files):
added roast-alike tests
: a14d3f9 | (Tobias Leich)++ | .gitignore:
added gitignore
FROGGS is there a way to tell t/harness to prepend something to every test file?
lichtkind anocelot: that was not nice i did merge and you pushed next change 22:20
now i have to merge yet again
masak lue: ah. you're right. 22:21
lichtkind sorry its in differen file
lue so... slang HappyMAIN is MAIN { } ? Then { $~MAIN = HappyMAIN; } ?
dalek blets: 317fa2d | (Herbert Breunung)++ | docs/ (2 files):
fixed links in 3
blets: 9837181 | (Herbert Breunung)++ | /:
merged with my head
blets: 39d1361 | (Herbert Breunung)++ | README.md:
Merge branch 'master' of github.com:perl6/tablets
lichtkind its all fine
masak lue: seems to me 'augment' in that case would be a handy shortcut for 'subclass and replace in this scope'.
FROGGS [Coke]: ping
masak lue: whereas the 'augment' for classes augments once and for all, and you can't go back.
lue Changing the definition of 'augment' for slangs though almost feels DWIMmy, but also indescribably... weird? 22:22
lichtkind anocelot: as you see in github.com/perl6/tablets/commit/31...bc616ce4fa you dont need full url to link to same page 22:23
lue (updated gist: gist.github.com/lue/5241230)
masak lue++ # thinking about how to get actions in there
lue: if you crack that nut, you will have left an indelible mark on the evolution of Perl 6. :) 22:24
lue I almost want to say slang is a shortcut for grammar [Appropriate Grammar's Name], but then what would the shortcut for actions be?
timotimo_ swagger
anocelot lichtkind: When I clicked the link on the tablets.perl5.org I got a 404 error, so I assumed it needed the full URL.
masak lue: well, you'd tend to want to define them together. 22:25
timotimo_ "actions" may not be longer than "swagger", but it amuses me to suggest it, and it's not totally wrong
anocelot lichtkind: For the .dm file it's OK? Sorry; I'm probably not as familiar with Git as I should be.
lichtkind i dont know which link it was and i have to leave soon
masak lue: and just as the grammar (in the 'augment' case) is a subclass of a grammar, you'd want the actions class to be a subclass of an existing actions class.
timotimo_ a random question: "augment class" is a lexical or global change? 22:26
lichtkind there are no .dm files to edit as far as i know
masak timotimo_: global.
timotimo_: because classes are global.
lichtkind oh readme
haha
anocelot lichtkind: md. Sorry. 22:27
22:27 xilo joined
lichtkind tsall good 22:27
just leave me message have really other stuff todo now
22:27 ashleydev joined, kaare_ left
anocelot OK. Thanks for your help! 22:27
lue I feel the actions side of slangs would be easier if we had a separate actions construct 22:28
timotimo_ masak: the 'lexical equivalent' of "augment class" is just deriving from it, aye?
lue timotimo_: except currently augmenting a class changes it everywhere until the End of Time™. 22:29
while augment slang is using a special definition of "augment".
masak timotimo_: yes, if (1) you make it a 'my class', and (2) you give it the same name, so that it lexically shadows the outer one. 22:30
lue
.oO(can action methods accept anything other than $/ as a parameter?)
masak r: class A { method foo { say "outer!" } }; { my class A { method foo { say "inner!" } }; A.new.foo }; A.new.foo
p6eval rakudo b972ca: OUTPUT«inner!␤outer!␤»
timotimo_ lue: not that i know, but the actions classes make use of $* and friends a lot
masak r: class A { method foo { say "outer!" } }; my $x; { my class A { method foo { say "inner!" } }; $x = A.new }; say $x.WHAT; $x.foo 22:31
p6eval rakudo b972ca: OUTPUT«(A)␤inner!␤»
timotimo_ hm, or maybe there is a way
lue would like changing the actions of a language separated from normal class augmentation, like slang for grammars 22:34
22:38 xorp left
lue would also like to know what the original author thought a slang encompassed 22:38
22:39 xorp joined
dalek : f64b510 | (Tobias Leich)++ | Makefile:
pass --add_use_stmt=perl5 to fudge script
 22:39
22:42 wingfold joined
FROGGS current status of rakudo v5.pm: it passes one of the 485 perl 5 test files 22:43
pmurias: --------^
lue FROGGS++ 22:44
huf you're planning to reimplement all of the weird quirks of p5? 22:45
FROGGS now I "just" need to fudge the failing tests
huf: well, I need to do as most as possible to get cpan modules working
huf wow.
FROGGS I'm not sure what exactly that means atm 22:46
huf which one? the "wow" or what you said? :)
FROGGS what I said :o)
huf ah. i dont think anyone knows :) 22:47
FROGGS because it can't support XS, and maybe many other things
huf yes, XS is out, things that rely on p5's gc are out (or at least difficult)
and things that use weeeeird features of p5 are probably not worth it ;)
lue Don't forget (from S01): "Some platforms may restrict this to a subset of Perl 5 when it is not expedient to include a full Perl 5 interpreter." 22:48
huf :)
FROGGS :o) 22:49
huf this should be interesting, we'll find out how small that subset is, to run a large number of useful modules
FROGGS ya, I hope so :o) 22:51
but now my bed is the next thing on my todo
gnight!
22:51 skids joined 22:53 jeffreykegler joined
lue Some initial slang action ideas: gist.github.com/lue/5241230 22:54
masak lue: well, you've put your finger on something interesting here. 22:56
lue: 'augment slang' means lexiaclly replace some grammar in the current scope. (or dynamically replace, from the perspective of the parser.) 22:57
but 'augment class', which one would use for an actions class, would have a global, irreversible effect.
lue I wonder if augment grammar is closer to class or to slang... 22:58
masak on the one hand, it does kinda feel like double the work to have to augment two things just to build a new slang. on the other hand, a slang feels like just the grammar, not grammar+actions.
so there's kind of a tension there.
lue Like I said earlier, I'd like to know what the original author (likely TimToady) thought a slang was.
timotimo_ is it even possible to sensibly expand only the grammar, but keep "api compatibility" to the previous grammar's actions? 23:00
23:00 Khisanth joined
lue
.oO(maybe slang should contain a special type of method that takes two curly braces: slang MAIN { feature token-or-something {regex}{actions};}; )
 23:00
timotimo_ oh yikes, that seems weeeeeeiird 23:01
cognominal is it possible to hide a rule in a slang or change an operator precedence? 23:02
masak timotimo_: re api compatibility, that's a really good question too. essentially we're into the usual problems with classes and inheritance.
cognominal that's really two independent questions
lue timotimo_: changing Perl 6 from within is weeeeeeiird too :)
timotimo_ not as weird as introducing such a strange concept as two curly blocks following with nothing in between 23:03
cognominal bed time. Will read the log.
timotimo_ i mean we have that for s{...}{...} already, but that's ... different
lue good ♞, cognominal o/
23:03 PacoAir left
cognominal thx, lue, but I prefer go to chess. :) 23:04
lue but, there's no kNight pun in Go :)
cognominal mask, is there a full implementation of macro forthcoming? :) 23:05
lue, yes I know. :)
lue updated gist with the two braces option: gist.github.com/lue/5241230 23:06
masak cognominal: yes. 23:08
'night, #perl6
lue good ♞, masak o/ 23:09
23:09 cognominal left
timotimo_ lue: oh, the eskimo kiss operator is back :P 23:10
23:10 cognominal joined
lue with new meaning :) 23:10
cognominal masak: how? 23:11
bonsoir les ♗ de Perl 6, réellement cette fois :) 23:12
lue .u ♗ 23:13
yoleaux U+2657 WHITE CHESS BISHOP [So] (♗)
cognominal fou en Français :)
… et non évêque 23:14
lue :) [nous ne sommes pas les fous!]
lue is starting to like the double brace idea because it's -Ofun to him. 23:19
23:27 jeffreykegler left 23:36 cognominal__ joined 23:37 cognominal left 23:40 FROGGS left, FROGGS joined
timotimo_ lue: my 0.2 cents is that it's nicer to have a keyword before each of the braces. it's like the difference between positional args and keyword args. the latter seems more readable to me. 23:52
also, consider that the }{ may be separated by lots and lots of lines from the beginning and end of the whole construct 23:53
lue yeah. I was thinking of maybe a different brace for the second set.
timotimo_: I just typed }{ because I prefer it. Of course a bunch of newlines inbetween the } and { is fine 23:54
timotimo_ i dunno. it may feel disconnected and look like a loose block if there's two newlines for example
lue different brace style might be the way to go, with some enforced connectedness perhaps (feature newtoken /grammar/{actions} or feature newtoken {grammar}«actions» or ...) 23:55

Powered by Cro and the Raku® Programming Language.

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

Thank you!