»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by masak on 12 May 2015.
00:07 jercos joined, hahainternet joined, garu joined, Guest72901 joined, aindilis joined, mantovani joined, gugod joined, xxx joined, moritz joined, pink_mist joined, oka_ joined, sjn joined, tm joined, sitaram joined, maddingue joined, bakedb joined, shmibs joined, tinita joined, huf joined, cibs joined, MilkmanDan joined, stmuk joined, Sgeo__ joined, Timbus joined, rhr joined, labster joined, pRiVi_ joined, btyler joined, mrf joined, freeze joined, Grrrr joined, ggoebel joined, [particle]1 joined, advwp joined, cognominal joined, TimToady joined, grondilu joined, ilbelkyr joined, agentzh joined, |Tux| joined, burnersk joined, Jonis joined, kfo joined, SHODAN joined, ofzyh__ joined, kid51 joined, Juerd joined, dustinm` joined, avar joined, ribasushi joined, jordman_ joined, dpk joined, yoleaux joined, Hor|zon joined, softmoth joined, crux joined, literal joined, confuseAcat joined, wtw joined, Humbedooh joined, sjohnson joined, verne.freenode.net sets mode: +v yoleaux, apotheon joined, mprelude joined, Zoffix joined, gonz_ joined, sQuEE` joined, AW3i joined, felher joined, bitmap joined, simcop2387 joined, salva joined, lestrrat joined, dg joined 00:08 dwarring joined, xvcds joined, avar left, solarbunny joined, pnu joined 00:09 hohoho joined, ssflying joined, Util joined, noganex joined, [Sno] joined, Exodist joined, avar joined 00:11 Kogurr joined 00:13 Psyche^ joined, xinming joined, PerlJam joined, lea joined, ambs joined, obra joined, xnrand joined, lizmat joined, rjbs joined, flussence joined, [ptc] joined, jsimonet joined, frew joined, mtj_ joined, jdv79 joined, dalek joined, boegel|quassel joined, diakopter joined, ilbot3 joined, oha joined, synbot6 joined, DarthGandalf joined, preyalone joined, erxeto_ joined, cosimo joined, isacloud joined, skarn joined, hoelzro joined, mullagainn joined, timbunce joined, verne.freenode.net sets mode: +v dalek 00:14 leedo joined, zoosha joined, tony-o joined, jevin joined, murre joined, pyrimidi_ joined, Celelibi joined, colomon joined, jantore joined, f3ew joined, intrigue joined, tadzik joined, oahong joined, avalenn joined, geekosaur joined, rudi_s_ joined, BinGOs joined, cxreg2 joined, chenryn joined, jonadab joined, bhm joined, roguelazer joined, mattp__ joined, stux|RC-only joined, broquaint joined, [Coke] joined, ab5tract_ joined, renormalist joined, esh joined, ggherdov joined, daxim joined, Gothmog_ joined, danstone1 joined, retupmoca joined, gtodd joined, DrForr joined, domm joined, go|dfish joined, doublec joined, sivoais joined, nowan joined, KotH joined, b2gills joined, samb1 joined, robinsmidsrod joined, slobo joined, yuppie joined, sunnavy joined, _sri joined, n_i_c_k_ joined, arnsholt joined, apejens joined, smash joined, rvchangue joined, bartolin joined, eiro joined, pierrot joined, raiph joined, sufrostico joined, lucs joined, bpmedley joined, iv joined, olinkl joined, jferrero joined, takadonet joined, jrm joined, moznion joined, gfldex joined, ponbiki joined, charsbar joined, diego_k joined, chansen_ joined, mephinet joined, timotimo joined, silug joined, BooK joined, xdbr joined, Ulti joined, cfloare joined, nine joined 00:16 WizJin joined, Fleurety joined 00:17 telex joined, woodruffw joined, sergot_ joined, CurtisOvidPoe joined, spacebat joined, krakan joined, sftp joined, Upasaka_ joined, ShimmerFairy joined, vytas joined, pochi_ joined, mrsolo joined, japhb joined, quuxman joined, ben joined, clkao joined, Bucciarati joined, Khisanth joined, mathw joined, zostay joined, integral joined, jnthn joined, JimmyZ joined, El_Che joined, ilmari joined, skaji joined, mls joined, raydiak joined, burnersk left 00:18 PsuzeannePsy joined, mst joined, awwaiid joined, inokenty joined, emdashcomma joined, Praise joined, carlin joined, Possum joined, petercommand joined, nebuchadnezzar joined, xiaomiao joined, avar left, avar joined 00:19 ingy joined, revdiablo joined, breinbaa1 joined, baest_ joined, khw joined, FireFly joined, FROGGS joined, TEttinger joined, atta joined, Hotkeys joined, jojotus joined, captain-adequate joined, lsm-desktop joined, zemmihates joined, kipd joined, SmokeMachine_ joined, erdic joined, Fluor joined, xenu joined, avuserow1 joined, Debolaz joined, [Tux] joined, ugexe joined, riatre joined, PotatoGim joined, nightfrog joined, Spot__ joined, dylanwh_ joined, camelia joined, ashleyde1 joined, psch joined, Gardner joined, masak joined, hobbs joined, kst joined, verne.freenode.net sets mode: +v camelia 00:21 WizJin left, WizJin joined 00:22 jkva joined 00:23 sivoais_ joined
dalek kudo/nom: 039069e | hoelzro++ | src/Perl6/Compiler.nqp:
Eagerly evaluate CORE::.keys for REPL completions
 00:23
kudo/nom: d1c42ea | hoelzro++ | src/Perl6/Compiler.nqp:
Treat failure to set up completions differently from failure to load Linenoise

For the former, we should deliver a little bit of information to help
  developers fix the issue. For the latter, it's usually just that
Linenoise isn't installed, so it's not a big deal
kudo/nom: a32c30f | hoelzro++ | src/Perl6/Compiler.nqp:
Refactor generating completion list into helper method
00:23 burnersk joined 00:24 mrsolo left 00:25 dpk left, avar left 00:26 jkva left, avar joined, dpk joined 00:27 sivoais left 00:28 Util left, Util joined 00:29 confuseAcat left, confuseAcat joined, aborazmeh joined 00:30 jonadab left 00:31 aborazmeh left, aborazmeh joined, pierrot is now known as Guest89352 00:35 jonadab joined 00:36 avar left, avar joined, avar left, avar joined 00:42 avar left, uruwi joined, avar joined, avar left, avar joined 00:45 n0tjack joined 00:49 n0tjack left
uruwi I noticed that holding a key down gives a delay before it fires repeatedly. Is the method to remove that delay specific to the terminal emulator being used, or can I change it with a command line setting? 00:50
TimToady do you really want random letters repeating while you type normally? 00:51
uruwi Normally I'd say no
But here I'll have to say yes
pink_mist pretty sure that depends on either your OS or your terminal 00:54
00:56 llfourn joined, cognominal left
uruwi Probably should be using a graphical interface instead by now, considering terminals can't properly handle simultaneous keypresses 00:57
I like the feel of playing stuff on a terminal I guess. 00:59
01:01 rabato joined, SuzieQueue joined, llfourn left
rabato hi, rakudo. Is this ordered? .say for 'abcd'.comb>>.succ; 01:02
uruwi m: .say for 'abcd'.comb>>.succ 01:03
camelia rakudo-moar 271e84: OUTPUT«b␤c␤d␤e␤»
uruwi Seems so
rabato 'abcd'.comb>>.succ.say 01:04
but it not ordered
01:05 PsuzeannePsy left, preyalone left
ugexe because say is a side effect 01:08
uruwi m: 'abcd'.comb>>.succ.say 01:09
camelia rakudo-moar 271e84: OUTPUT«(b c d e)␤»
rabato ugexe: so! .say for 'abcd'.comb>>.succ; is ordered?
01:10 softmoth left
rabato uruwi: I think the output is ordered, but ">>" behaver is not ordered 01:13
ugexe m: my @x = 'abcd'.comb>>.succ; say @x
camelia rakudo-moar 271e84: OUTPUT«[b c d e]␤»
ugexe in your example, .say gets called out-of-order, but the result from *that* (True) will come back ordered
timotimo yeah, the order of the results is preserved, but things get executed not-in-order
to be fair, the say in >>.succ.say will get executed "in-order", as the say isn't hypered 01:14
rabato timotimo: maybe something changed, i read before. thay say it is not ordered. 01:16
timotimo there's hyper sequences that'd do that, but i don't think those get created with >>.
rabato timotimo: why say is not hypered? 01:17
ugexe m: 'abcd'.comb>>.succ>>.say 01:18
camelia rakudo-moar 271e84: OUTPUT«b␤c␤d␤e␤»
ugexe i would think that would be out of order
timotimo m: "1234567".comb>>.say 01:20
camelia rakudo-moar 271e84: OUTPUT«1␤2␤3␤4␤5␤6␤7␤»
rabato ugexe: i think i will be ordered.
timotimo huh, we don't do that any more?
01:20 Guest89352 is now known as pierrot 01:21 pierrot is now known as Guest90228
rabato timotimo: give me an answer? please? ".say for 'abcd'.comb>>.succ;" 01:22
ugexe star: 'abcd'.comb>>.succ>>.say # looks like it disappeared a little while ago
camelia star-m 2015.09: OUTPUT«b␤c␤d␤e␤»
01:23 Guest90228 is now known as pierrot
timotimo i did not know that! 01:23
i know nothing!
01:23 pierrot left, pierrot joined
rabato timotimo: anyway, thanks. 01:23
ugexe rabato: that will say them in order, even if >> was unordering them
rabato ugexe: thanks! 01:24
01:26 sufrostico left 01:29 sivoais_ is now known as sivoais, sivoais left, sivoais joined
timotimo sorry rabato 01:32
rabato timotimo: it's ok. 01:33
01:34 PsuzeannePsy joined 01:37 SuzieQueue left 01:39 sufrostico joined 01:44 adu joined 01:46 Suzeanne joined 01:49 PsuzeannePsy left 01:50 pyrimidi_ left 01:51 pyrimidine joined 01:53 tokuhiro_ joined, sufrostico left 01:54 preyalone joined 01:56 rabato left 01:57 tokuhiro_ left 02:00 rurban joined 02:12 PotatoGim left 02:13 kid51 left 02:17 noganex_ joined 02:19 noganex left 02:22 jkva joined 02:26 jkva left 02:33 SuzieQueue joined 02:35 chenryn_ joined 02:37 Suzeanne left 02:40 chenryn_ left, chenryn_ joined
hohoho m: require ::($_) for <NativeCall> 02:44
camelia rakudo-moar 271e84: OUTPUT«===SORRY!===␤Cannot invoke this object␤»
hohoho hmm
m: for <NativeCall> { require ::($_) }
camelia ( no output )
hohoho what's difference?
i seem normal 'for' stmt and postfix 'for' have same meanings 02:45
timotimo m: { require ::($_) for <NativeCall> }
camelia rakudo-moar 271e84: OUTPUT«===SORRY!===␤Cannot invoke this object␤»
timotimo it's not a bug about being inside { } or not at least 02:46
i don't think you'd need to runtime-require NativeCall, as it comes with rakudo nowadays
geekosaur m: { require ::($_) } for <NativeCall>
camelia ( no output )
timotimo geekosaur: are you sure that gets executed?
geekosaur no, actually
timotimo m: { say "hello" } for <NativeCall>
camelia rakudo-moar 271e84: OUTPUT«hello␤»
timotimo ah
geekosaur well. 02:47
hohoho NativeCall is just a example. 02:49
geekosaur m: $_.^name.say for <NativeCall> 02:54
camelia rakudo-moar 271e84: OUTPUT«Str␤»
geekosaur m: { $_.^name.say } for <NativeCall>
camelia rakudo-moar 271e84: OUTPUT«Str␤»
02:56 n0tjack joined 02:58 llfourn joined 03:00 skids joined 03:01 n0tjack left 03:03 llfourn left 03:05 PotatoGim joined 03:11 PsuzeannePsy joined 03:15 SuzieQueue left 03:17 somoto joined
somoto could you me tell how 03:17
a) create an emtpy hash
b) check if a hash is emtpy
c) check if an argument is a method is null (or rather, passed to it at all)?
c) check if an argument passed in a method is null (or rather, passed to it at all)? 03:18
03:18 BenGoldberg joined
timotimo m: my %foo; say %foo.perl 03:18
camelia rakudo-moar 271e84: OUTPUT«{}␤»
timotimo is that acceptable for a) ?
TimToady or {} as a literal
timotimo m: say {}.WHAT
camelia rakudo-moar 271e84: OUTPUT«(Hash)␤» 03:19
timotimo m: say {}.elems
camelia rakudo-moar 271e84: OUTPUT«0␤»
timotimo m: say {food => "good"}.elems
camelia rakudo-moar 271e84: OUTPUT«1␤»
TimToady m: say so {}
camelia rakudo-moar 271e84: OUTPUT«False␤»
TimToady m: say so { :any<thing> }
camelia rakudo-moar 271e84: OUTPUT«True␤»
03:20 rurban left
somoto and "c" please? 03:20
timotimo i'd go with $that_argument.defined, but it's also possible to pass an undefined value for that argument and it'll trigger that check, too 03:21
m: sub test($foo?) { say $foo.defined }; test(123); test(); test(Int);
camelia rakudo-moar 271e84: OUTPUT«True␤False␤False␤»
TimToady m: sub foo ($a = "No value passed") { say $a }; foo("value"); foo() 03:22
camelia rakudo-moar 271e84: OUTPUT«value␤No value passed␤»
03:22 Ben_Goldberg joined
somoto > say $fdsfd.defined; 03:22
===SORRY!=== Error while compiling <unknown file>
Variable '$fdsfd' is not declared
> say $fdsfd.defined;
===SORRY!=== Error while compiling <unknown file>
Variable '$fdsfd' is not declared
> say $fdsfd.defined;
===SORRY!=== Error while compiling <unknown file>
Variable '$fdsfd' is not declared
> say $fdsfd.defined;
===SORRY!=== Error while compiling <unknown file>
Variable '$fdsfd' is not declared
> say $fdsfd.defined;
===SORRY!=== Error while compiling <unknown file>
is there any way to safely check if it's defined or not?
TimToady there's no way to check that it's *declared*, which is different from defined 03:23
skids Well, other than to use EVAL.
TimToady if it's not declared, it's a compile error
somoto Ok. I just want to check if an optional argument is passed to my method inside the method, what's the best way to do that? 03:24
skids Use a default like TimToady showed above.
03:24 BenGoldberg left
somoto thanks 03:25
03:26 AlexDaniel joined
AlexDaniel “I think using \r\n is non-sense in 2015. We can use \n instead of \r\n.” – I wish there was a way to go for it without making everyone suffer 03:27
on the other hand, lots of software on windows actually understand \n quite nicely
.oO( another potentially good idea is throw away support for some operating systems altogether, bhaha )
 03:28
skids There will always be the task of processing archived data created with old saoftware, that'll probably never go away. 03:29
03:29 somoto left
uruwi Yep 03:38
Screw \r\n
Screw notepad
03:50 Gardner left 03:51 Gardner joined, captain-adequate left, captain-adequate joined 03:54 tokuhiro_ joined 03:59 tokuhiro_ left, n0tjack joined 04:00 llfourn joined 04:03 n0tjack left 04:05 SuzieQueue joined 04:07 uruwi left 04:08 PsuzeannePsy left 04:13 tomoko44 joined
tomoko44 I'm trying to check if arguments are emtpy (or passed at all) or not: 04:13
method method1($var1?, $var2?) {
my %my-hash;
if $var1 {
say "if var1 is True";
%my-hash{'var1'} = $var1;
}
if $var2 {
say "if var2 is True";
%my-hash{'var2'} = $var2;
}
}
method1(var1 = "fdsomfdsfds");
But the code in the blocks "if $var1 { ... }" and "if $var2 { ... }" never executes. Why?
formatted:
I'm trying to check if arguments are emtpy (or passed at all) or not:
method method1($var1?, $var2?) { 04:14
my %my-hash;
if $var1 {
say "if var1 is True";
%my-hash{'var1'} = $var1;
}
if $var2 {
say "if var2 is True";
%my-hash{'var2'} = $var2;
}
}
method1(var1 = "fdsomfdsfds");
But the code in the blocks "if $var1 { ... }" and "if $var2 { ... }" never executes. Why?
is there any better way to accomplish the same thing?
skids tomoko44: for future reference, you can make a github gist and feed it to the evalbot. 04:15
tomoko44 ok
04:16 AlexDaniel left
TimToady first, change 'method' to 'sub', if you're going to call it as a subroutine, not as a method 04:17
seocond, you don't need var1 = there, just pass the string, since you declared a positional argument
skids m: class A { method foo ($v1?, $v2) { if $v1 { say "if var1 is true" }; if $v2 { say "if var2 is true" } }; A.new.foo(1,2);
camelia rakudo-moar 271e84: OUTPUT«5===SORRY!5=== Error while compiling /tmp/4qKoNP5WnF␤Cannot put required parameter $v2 after optional parameters␤at /tmp/4qKoNP5WnF:1␤------> 3class A { method foo ($v1?, $v27⏏5) { if $v1 { say "if var1 is true" }; if␤ expecting any of…»
TimToady if you wanted named parameters, then declare them :$var1, :$var2
skids m: class A { method foo ($v1?, $v2?) { if $v1 { say "if var1 is true" }; if $v2 { say "if var2 is true" } }; A.new.foo(1,2); 04:18
camelia rakudo-moar 271e84: OUTPUT«5===SORRY!5=== Error while compiling /tmp/VCiXLi8k24␤Missing block␤at /tmp/VCiXLi8k24:1␤------> 3y "if var2 is true" } }; A.new.foo(1,2);7⏏5<EOL>␤»
TimToady and pass var1 => 'string'
(not = char, this isn't python)
tomoko44 yeah, I meant named parameters, that is :$var1, :$var2 04:19
ok, thx. 04:20
04:20 tomoko44 left
skids m: class A { method foo ($v1?, $v2?) { if $v1 { say "if var1 is true" }; if $v2 { say "if var2 is true" } } }; A.new.foo(1,2); 04:20
camelia rakudo-moar 271e84: OUTPUT«if var1 is true␤if var2 is true␤»
skids m: class A { method foo (:$v1, :$v2) { if $v1 { say "if var1 is true" }; if $v2 { say "if var2 is true" } } }; A.new.foo(:v1,:v2(3)); 04:22
camelia rakudo-moar 271e84: OUTPUT«if var1 is true␤if var2 is true␤»
04:28 mayuresh joined
mayuresh hello :) 04:29
is anyone awake? 04:30
04:30 PsuzeannePsy joined
mayuresh i guess not! :) 04:31
04:31 Ben_Goldberg left
mayuresh see you all after 12 hours, that's when i'd be headed off to bed though. :) 04:31
04:32 mayuresh left
skids o/ 04:32
oh.
04:33 SuzieQueue left 04:36 WizJin left 04:42 chenryn_ left, WizJin joined, konobi joined
konobi howdy! are there any examples of folks using grammars to do DSL style stuff? 04:43
04:46 hohoho left 04:47 khw left
llfourn konobi: this is an example of DSL as a slang: github.com/tony-o/perl6-slang-sql 04:52
04:55 Suzeanne joined 04:56 ShimmerFairy left 04:58 PsuzeannePsy left 05:05 chenryn_ joined 05:09 ShimmerFairy joined 05:14 timotimo left 05:15 adu left 05:16 n0tjack joined 05:18 skids left 05:21 n0tjack left 05:24 jkva joined
konobi llfourn: "slang" ? 05:28
05:28 jkva left
raiph konobi: S99:slang 05:32
synbot6 Link: design.perl6.org/S99.html#slang
05:33 raiph left
konobi ta 05:35
llfourn konobi: A slang is like a grammar but the difference is that you are modifying the existing Perl 6 grammar to incoroporate your DSL into the Perl 6 language itself 05:36
konobi: In terms of straight Grammar DSLs the only things I can think of are various templating languages
konobi: github.com/softmoth/p6-Template-Mustache/ 05:37
konobi ah, kk... i was thinking more in terms of something like Ansible... but with sensible parsing, not just YAML loading 05:39
llfourn I actually happen to be doing something like that in Perl 6 as well but have not yet put it on GH 05:40
but yes -- you could do that with a normal grammar -- and use the "actions" to translate it to whatever structure you need 05:41
05:45 SuzieQueue joined 05:46 hohoho joined 05:48 Suzeanne left 05:50 PsuzeannePsy joined 05:51 hohoho left 05:54 SuzieQueue left 05:56 tokuhiro_ joined 06:00 tokuhiro_ left
konobi llfourn: yeah, getting my head around it now =0) 06:04
06:08 diana_olhovik joined 06:10 hohoho joined, hohoho left, hohoho joined 06:13 Woodi joined
[Tux] test 50000 39.166 39.048 06:16
test-t 50000 39.600 39.483
06:20 baest_ is now known as baest 06:21 bjz joined 06:24 FROGGS left 06:26 Alina-malina joined, bjz_ joined, bjz left, SuzieQueue joined 06:27 firstdayonthejob joined 06:29 hartenfels joined, hartenfels is now known as askmeaboutloom, askmeaboutloom left 06:30 PsuzeannePsy left, |Tux| left, bjz_ left 06:31 |Tux| joined 06:32 Ven joined 06:33 hartenfels joined 06:38 |Tux| left 06:41 |Tux| joined 06:44 jkva joined 06:45 |Tux| left 06:46 vike joined, bjz joined, |Tux| joined 06:51 bjz_ joined 06:52 bjz left, FROGGS joined 06:53 |Tux| left 06:56 |Tux| joined 06:59 RabidGravy joined 07:04 bjz_ left, lea left 07:05 bjz joined 07:06 Alina-malina left 07:10 abraxxa joined 07:13 dwarring left 07:15 Ven left, boegel|quassel is now known as boegel, Alina-malina joined 07:16 jkva_ joined, jkva left, lea joined 07:18 ilbelkyr is now known as u, boegel left 07:19 boegel joined 07:21 firstdayonthejob left
RabidGravy morning! 07:24
07:24 ely-se joined 07:28 lea left
dalek kudo/nom: 2324810 | TimToady++ | src/core/metaops.pm:
allow ».[1;1] to pass single array arg

Fixes RT #126370		 07:36
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=126370
dalek ast: f8c0844 | TimToady++ | S03-metaops/hyper.t:
hyper multidim now works (RT #126370)
 07:37
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=126370
07:37 KDr2 joined 07:38 aborazmeh left 07:39 Kogurr left, Ven joined 07:42 lea joined
konobi RabidGravy: yo 07:43
RabidGravy hey!
you good?
konobi aye, not too shabby
RabidGravy fabulous! 07:44
07:44 Ven left
konobi getting back into the swing of job-hunting/contract-seeking 07:44
07:44 n0tjack joined
RabidGravy :) 07:44
konobi and attempting to re-learn patience in awaiting clients 07:45
^_^
07:46 tokuhiro_ joined
konobi RabidGravy: how bout you mate? 07:47
RabidGravy All fabulous, having a big slack/personal hack time at the moment - may have to start looking for a new gig at some point ;-) 07:48
07:49 n0tjack left
RabidGravy tokuhiro_, are you the tokuhirom who submitted github.com/sergot/http-useragent/issues/92 ? 07:49
07:50 tokuhiro_ left
RabidGravy oh, I'll just carry on writing the reply then 07:50
07:50 mrsolo joined, zakharyas joined 08:00 u is now known as ilbelkyr
RabidGravy anyone here with an IPV6 connection to the internet care to chime in on github.com/sergot/http-useragent/issues/92 ? 08:00
08:07 Ven joined 08:15 Coldkeys joined, Coldkeys left, darutoko joined, Ven left 08:18 Ven joined 08:22 Ven left 08:24 Ven joined 08:26 bjz left, bjz joined 08:29 chenryn_ left 08:31 espadrine joined, Ven_ joined, Ven left
moritz m: say Int:D ~~ Int:D 08:35
camelia rakudo-moar 271e84: OUTPUT«False␤»
08:40 chenryn_ joined, n0tjack joined 08:45 n0tjack left 08:46 Alina-malina left 08:48 xfix joined 08:51 ][Sno][ joined, tokuhiro_ joined 08:52 [Sno] left, ][Sno][ is now known as [Sno], espadrine left 08:54 margeas joined 08:56 tokuhiro_ left, mprelude_ joined 08:59 mprelude left 09:00 FireFly left 09:01 FireFly joined
Ven_ are "is nodal" and "is raw" formally documented somewhere? 09:01
[ptc] is it possible to submit smoke/test reports to testers.p6c.org?
RabidGravy right, off out - play nicely.
FROGGS [ptc]: sure 09:05
[ptc] FROGGS: tried to find how on the website, but couldn't find it. How do I submit reports?
FROGGS [ptc]: PANDA_SUBMIT_TESTREPORTS=1 panda smoke - or - PANDA_SUBMIT_TESTREPORTS=1 panda install Foo
[ptc] ah, ok 09:06
09:06 RabidGravy left
[ptc] FROGGS: is that mentioned on the testers site? 09:06
FROGGS: ... and I just happened to miss it?
Ven_ .tell TimToady irclog.perlgeek.de/perl6/2015-10-15#i_11381191 putting "Foo_" or "Foo-" would mean I'm reinventing namespacing for no reason. I want `::` to be about packages. the name is inside the package, the value is dynamically scoped 09:07
yoleaux Ven_: I'll pass your message to TimToady.
09:07 KDr2 left
[ptc] FROGGS: is it possible to submit a test report for a module without installing it as well? 09:08
FROGGS [ptc]: I fear it is not mentioned :S
[ptc]: not atm
[ptc] FROGGS: that's ok, that gives me the opportunity to submit a PR for testers.p6c.org :-)
FROGGS \o/ 09:10
[ptc]++
[ptc]: a welcome page that explains things would be awesome
[ptc] FROGGS: maybe it's best for me to open a ticket; then at least it's written down somewhere... 09:12
09:12 mprelude_ left, andreoss joined
FROGGS aye 09:13
09:16 jkva_ left, duncan_dmg joined 09:18 telex left 09:19 c0rt042 joined 09:20 Ven_ left, telex joined, ofzyh__ left 09:21 duncan_dmg left, g4 joined 09:22 Ven joined
jnthn morning, #perl6 09:24
09:24 duncan_dmg joined
Ven o/, jnthn 09:24
vytas how is nqp interactive shell different from perl6's ? they both use HLL/Compiler.nqp, don't they ?
Ven jnthn: "Looking further into call vs. coercion and coercion API as part of RT #114026; post a proposal for discussion" you didn't add a link to that :P 09:26
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=114026
09:26 brrt joined
vytas I am looking at commit 74875ba93cde72aed295a071ab1ec6a392154f6f, and I don't think it is working as per commit message - my guess is due to \n\n 09:26
konobi what ever happened to the idea of nqp on llvm? 09:27
jnthn Ven: gist.github.com/anonymous/8efdab238ead35b1d4fa 09:29
Ven konobi: nobody did it? :P
brrt konobi: well volunteered :-P
09:29 Psyche^_ joined
jnthn Ven: Linked it from the post now too :) 09:30
Ven jnthn++
09:31 azawawi joined
azawawi morning 09:31
brrt morning azawawi 09:32
azawawi github.com/azawawi/perl6-selenium-...es/ex01.p6 # more automated website ui testing is now working :)
09:32 espadrine joined 09:33 Psyche^ left
azawawi any update on panda supporting module versions? 09:35
e.g. panda install Selenium::WebDriver@0.15 09:36
konobi Ven: there was talk of it quite a few years back... just wondering if someone took a look again, now that llvm is a bit better organized
Ven konobi: do you still have it? 09:37
konobi and vmkit working for jvm and clr
some of the llvm tooling could make for some interesting ways to analyse the code 09:39
brrt konobi: i'm going speculate here, and suggest that the things lLVM is good at is only in a small respect overlapping with the things that are difficult in perl6 09:42
09:42 jkva joined
brrt i'm all for more backends, more power to whoever wants to do it 09:42
but i honestly would expect better results from pmurias++'s efforts of nqp-js than on llvm 09:43
vytas is it ok to log bugs for perl6 repl in nqp github issues page? 09:44
09:44 Ven left
brrt vytas: i think we use RT here, not sure exactly how it works 09:44
nobody will mind you putting bugs into the github issues page, it's just not the primary tracker :-) 09:45
ShimmerFairy if it's a rakudobug, you really shouldn't use the issues page, since rakudo uses RT. (Maybe the perl6 organization should disable it, if that's an option) 09:46
oh, it is already :P 09:47
vytas ShimmerFairy, it's enabled 09:48
according to github.com/rakudo/rakudo/ bugs related to nqp or moarvm should go through github issue tracker, rakudo/perl6 via [email@hidden.address]
09:48 ely-se left
ShimmerFairy vytas: not for rakudo/rakudo I mean 09:48
09:49 edehont joined 09:53 virtualsue joined 09:54 ely-se joined 09:55 bjz left 09:58 azawawi left
vytas If I go to rakudo.org/rt/open-all ( linked from rakudo.org/tickets/ ) I get possible cross-site request forgery 09:58
09:59 bjz joined 10:02 rindolf joined
vytas e-mailed to perlbug-admin at perl.org 10:02
10:02 rurban joined, ely-se left
brrt that may not come to the right place 10:03
seems like that may be for perl5 mostly?
vytas rt.perl.org is used for both
jnthn Indeed
vytas would bugs like RT #112634 be welcomed to fix in core? I mean my personal opinion is that Linenoise should be implemented in core too.. 10:06
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=112634
10:07 breinbaa1 left 10:08 breinbaas joined, edehont left
jnthn vytas: I'm pretty sure some folks did spend a good bit of time on RT #112634; I'm not sure what came of it. 10:09
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=112634
10:09 rindolf left
jnthn vytas: We actually used to have linenoise integrated at the VM level, and had to maintain a separate thing for JVM too, and then it didn't provide good access to most of the functionality for doing anything non-trivial, so it was better pushed out to a Perl 6 module. 10:10
10:12 Ven joined
Ven brrt: clearly, we should use llvm to then compile it to js! :P 10:12
brrt lol
why not
that's what emscriptem does isn't it?
ShimmerFairy jnthn: I'm personally of the opinion that rakudo requiring externally-hosted modules for some of its features is a poor idea. I'd like if a Linenoise-ish module, and a basic module for perl6-debug, would be provided like Pod::To::Text 10:13
vytas I am with ShimmerFairy, having advanced REPL in nqp makes little sense, but it must be installed with perl6 10:14
FROGGS ShimmerFairy: then use Rakudo Star 10:15
brrt i'm not. perl5 doesn't have a functional repl and we deal with that just fine
FROGGS that what distributions are for
or is Linenoise not included in Star?
10:15 Alina-malina joined
brrt (shockingly, my fedora installation apparantly doens't have Test::More :-O) 10:16
jnthn FROGGS: Not sure, but that's where it should go if it ain't already.
brrt Test::More is a p5 standard module, isn't it?
ShimmerFairy FROGGS: disagree. If the rakudo source code has a dependency on a module it doesn't provide, then that's a mistake. When you go to get the rakudo compiler, you should not be getting an incomplete compiler.
jnthn ShimmerFairy: You seem to have a strangely broad definition of compiler. 10:17
ShimmerFairy (besides, I always use the latest from the repositories, so 'just use R*' is not an acceptable solution)
jnthn: well, if the compiler needs it (e.g. the default UI for perl6-debug), that sounds to me like a case of something that should be a part of the compiler :)
brrt hmm 10:18
that's a reasonable point i guess
10:19 margeas left
FROGGS ShimmerFairy: rakudobrew build moar && rakudobrew build panda (or how it is called exactly) is no solution? really? 10:19
err, install Task::Star*
(one can see that I'm not using it)
10:20 brrt left
ShimmerFairy FROGGS: I don't use rakudobrew. I install straight from the repos (not even using the --gen options), and I've been a bit pissed about the fact that this doesn't give me a complete compiler, when wanting to use `perl6-debug` or the most basic of REPL line-editing functions. 10:20
vytas Task::Star fails to build at the moment, last time i tried it - it was not there. and no i don't like that ( even if my vote does not count )
ShimmerFairy And I will not change where I get my rakudo from simply because the repository refuses to be complete :) 10:21
FROGGS still, a compiler is a compiler is a compiler 10:22
fancy repl stuff really belongs to module space
also to allow alternatives
vytas in my opinion - python got interactive shell very right 10:23
and they got iPython
FROGGS yes, though Pythons vision is that there shall only be one (obvious) way to do things 10:24
which is not quite perlish
jnthn ShimmerFairy: Well, once we get the slang stuff solidified for 6.d, there'll be a stable API to do what perl6-debug does with the AST and it too can be booted back out to the ecosystem, the way I originally wanted things anyway. :) 10:25
konobi node's repl works pretty decently
10:25 bjz left, apotheon left
ShimmerFairy FROGGS: if you insist that a default Linenoise and such shouldn't be in the repo, then what are NativeCall, Pod::To::Text, Test, or lib doing in there? Why is a Linenoise-ish module different from those? 10:25
FROGGS so we already had linenoise included in the compiler and we recognized this as a failure... just dont let us do it again
10:26 margeas joined
FROGGS ShimmerFairy: that's quite simple... lib and Test are used to run the spectests, and NativeCall is tied to the VM version 10:26
10:26 bjz joined
ShimmerFairy Could there really not even be a basic REPL::Basics , which just makes sure you can press the left and right arrows keys and have them do the right thing? 10:27
10:27 xfix left
FROGGS if we can do this ultra portably, sure 10:27
which is what the problem was
vytas ShimmerFairy++ and fix bugs like #121117
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=121117
jnthn On Windows, without anything extra installed, the arrow keys (and up/down) already work perfectly well without needing modules, fwiw. :)
vytas so as i understand issue is with Linenoise not that it would not be welcomed? 10:28
10:28 Ven left
FROGGS vytas: this is not about fancy REPL stuff, this is about how the REPL (shipped with "the compiler") works 10:28
vytas: so even with Linenoise, you get this repl bug 10:29
ShimmerFairy FROGGS: I suppose my problem is that rakudo's REPL is laughably ridiculous unless you magically know that your compiler needs a couple more batteries it didn't come with (honestly, you might as well not have it and leave that to something like perl6 repl.p6 , the way it is)
vytas I am talking about basic REPL, fancy would be fancy :)
10:30 andreoss left
FROGGS ShimmerFairy: why do you think a fancy REPL belongs to the very core compiler package? 10:30
10:30 apotheon joined
FROGGS vytas: we ship a very basic repl 10:30
(with rakudo)
10:31 dalecooper joined
vytas Ok i only started playing with perl6 yesterday, so maybe i am missing something. but the one that comes up after i type in "perl6" is not functioning. 10:31
prompt comes as "> " i click enter it changes to "* "
FROGGS vytas: not functioning in what way?
vytas sounds like there is more than one prompt. - sorry for a bad joke 10:32
ShimmerFairy FROGGS: not overly-fancy, but the fact that perl6 -e 'loop { say EVAL(prompt("> ")) }' almost perfectly replicates the basic REPL is ridiculous.
FROGGS I dunno what the star is about but when you enter a statement it interprets and runs it... 10:33
jnthn FROGGS: I think the * was what the multi-line input was using?
vytas i see commits like "allow \ for multiple lines" - I try it - and it doesn't work
lizmat .botasnack
.botsnack # not awake yet
yoleaux :D
synbot6 om nom nom
FROGGS jnthn: yes, though when I do: "<enter> say 42" it does nothing 10:34
jnthn FROGGS: Yeah. :/
FROGGS ShimmerFairy: so what do you want to have extra which would be portable? 10:35
konobi tried with a semi-colon at the end?
FROGGS no luck there
vytas: the backslash does not seem to be handled... it will already evaluate the current line 10:36
ShimmerFairy FROGGS: I can see why you wouldn't want a REPL that gives you every bell and whistle, but I don't think you can avoid the fact that other compilers with a REPL are far more user friendly.
konobi huh... libuv in moarvm... interesting 10:37
vytas FROGGS, list is simple for me: "my @a = 1..3;" - should not print as i didn't ask it to. "{" should allow me to enter on next line. arrow back should work etc. :)
ShimmerFairy FROGGS: I think the REPL could use an introductory message, perhaps a nicer/more informative prompt than '>' (dunno what info it might provide), and as for features, at the very least the left and right keys should let me in-place edit, instead of print ^[[D and ^[[C
vytas python prints version, and that keyword "help" is available for more help 10:38
FROGGS ShimmerFairy: is see it this way: users will install either a package (rakudo+linenoise+other basic stuff) or use rakudobrew... installing the compiler only from the repo is like building gcc on your own... you know what you are doing and you know that you will have extra work 10:39
10:39 TEttinger left, dalecooper left
FROGGS ShimmerFairy: about the left/right cursor movements... is that portable to cmd.exe and powershell without needing to depend on external libraries? 10:40
jnthn FROGGS: YOu get it *for free* in cmd.exe!
FROGGS \o/
jnthn: so everybody should use it :P
jnthn wonders what the OSX situation is
konobi well, the debugger is just another repl really
jnthn Well yes, it's a bit surprising to me that the much maligned cmd.exe manages to provide a much better experience out of the box :P 10:41
FROGGS ShimmerFairy: please open an RT ticket about the core features the rakudo repl should provide... including a help mechanism and a greeting msg or so
ShimmerFairy FROGGS: ok, I'll ponder what useful basic features there should be and open a ticket. :) 10:42
konobi tput should work in most places 10:43
10:43 ely-se joined
ShimmerFairy FROGGS: btw, another concern of mine is that "image" thing: it just doesn't look good that rakudo's REPL isn't much better than a looping EVAL'd prompt. And despite how many people will use rakudobrew and such, there will be people who go more along my method and find themselves with a crappy REPL :) 10:44
FROGGS ShimmerFairy: aye, and then well will talk about what of these features will be implementable without loading a dll/so/dylib
ShimmerFairy FROGGS: agreed, my mind is gearing towards things that wouldn't be too stressful for nqp to implement itself. 10:45
konobi there's an npm library called "blessed" that does native terminal stuff directly using the "tput" methodology... no external dependencies needed. 10:46
vytas *from user point of view* i don't care where it comes from - i want "apt-get install perl6" to include working repl. :)
FROGGS konobi: that's unixish thinking though
vytas: exactly
konobi iirc it works on windows too 10:47
FROGGS konobi: there is no tput on my windows box :o)
konobi it's a method 10:50
10:50 margeas left
vytas FROGGS, at the moment packages.debian.org/source/sid/rakudo does not depend on Task::Star, should it ? 10:51
10:52 chenryn_ left 10:53 tokuhiro_ joined
moritz uhm, no 10:53
konobi yup... on windows you can do the same thing, just need to have a copy of the windows-ansi terminfo available.
moritz the other way 'round, maybe
vytas moritz, silly question, but same logic - one button to have repl not more :) 10:54
moritz vytas: well, Task::Star needs rakudo, so rakudo can't need Task::Star 10:55
vytas: the typical solution is to recommend the installation of a meta package, like 'perl6-all' or so that depends on all of them
10:57 tokuhiro_ left 10:58 jkva left 11:05 karol_ joined
jdv79 will probably never understand why a REPL is an important feature 11:14
moritz: can we up the max files limit on hack?
elasticsearch gets very hungry it seems 11:15
moritz jdv79: I already did 11:24
11:24 WizJin left
jdv79 oh, thanks 11:24
hmm, /proc says 100k but ulimit still says 2k and denies up'ing. is that right? 11:28
ilmari check /etc/security/limits.conf? 11:30
11:30 Ven joined 11:31 duncan_dmg left 11:32 kid51 joined
jdv79 moritz: gist.github.com/anonymous/950bf286dcc617e4d616 11:33
11:33 duncan_dmg joined
moritz jdv79: did you log in anew? 11:33
jdv79 yes, fresh shell
moritz jdv79: there's no line with 'nofile' in /etc/security/limits.conf 11:34
jdv79 does that one liner work for you?
moritz jdv79: nope 11:35
11:36 WizJin joined
moritz jdv79: now it does 11:36
seems the absence of a limit doesn't mean it's unlimited :(
jdv79 took me a bit to fix it on my box here 11:37
11:39 jkva joined
vytas perl6 --doc=NonExisting shows shell rather than warning/error 11:42
11:44 Ven left 11:45 Ven joined 11:47 Ven left, Ven joined 11:49 Ven left, Ven joined 11:55 hohoho left, Ven left, kid51 left 11:56 chenryn_ joined 11:58 virtualsue left 12:01 Ven joined 12:04 mayuresh joined 12:08 mayuresh left 12:09 Ven left, ely-se left 12:12 ely-se joined 12:14 rindolf joined
moritz vytas: please report bugs at github.com/perl6/doc/issues/ 12:15
dalek kudo-star-daily: a66e8d9 | coke++ | log/ (8 files):
today (automated commit)
 12:16
c: 99228f0 | (Lloyd Fournier)++ | doc/Type/Routine.pod:
typo in routine signature
 12:20
12:22 apotheon left 12:23 bjz left
grondilu Error while compiling, type X::Syntax::ConditionalOperator::PrecedenceTooLoose 12:24
operator: =>
at line 29755, near "?? $contro"
at gen/moar/stage2/NQPHLL.nqp:519
12:27 chenryn_ left 12:28 apotheon joined, bjz joined 12:29 Ven joined, sufrostico joined
vytas moritz, ok thanks 12:31
12:32 pb_ joined 12:33 pb_ left 12:34 cognominal joined 12:37 FROGGS left 12:38 Ven left, chenryn left, ely-se left 12:39 hohoho joined 12:44 Ven joined 12:45 Ven left 12:47 abraxxa left, ely-se joined 12:49 apotheon left 12:54 tokuhiro_ joined 12:56 apotheon joined 12:59 tokuhiro_ left 13:00 ely-se left 13:02 Ven joined 13:04 adu joined 13:05 ely-se joined 13:09 mullagainn left, adu left 13:10 PsuzeannePsy joined, Alina-malina left, lichtkind joined 13:11 Ven left 13:12 rarara_ joined
rarara_ m: constant \c :=12; c=11; say c; 13:12
camelia rakudo-moar 271e84: OUTPUT«Cannot modify an immutable Int␤ in block <unit> at /tmp/QAlx9VUVhJ:1␤␤»
rarara_ m: my \c :=12; c=11; say c;
camelia rakudo-moar 271e84: OUTPUT«Cannot modify an immutable Int␤ in block <unit> at /tmp/_j0HXk8FSx:1␤␤»
rarara_ are those exactly the same?
jnthn No
constant evaluates the RHS at compile time 13:13
vytas got randomly segfault pastebin.com/raw.php?i=hvCqJqx6
jnthn m: for ^5 { constant \c = do { say 'here'; 5 }; say c; }
camelia rakudo-moar 271e84: OUTPUT«here␤5␤5␤5␤5␤5␤»
jnthn m: for ^5 { my \c = do { say 'here'; 5 }; say c; }
camelia rakudo-moar 271e84: OUTPUT«here␤5␤here␤5␤here␤5␤here␤5␤here␤5␤»
13:13 SuzieQueue left
jnthn rarara_: ^^ there's the difference 13:14
13:15 SuzieQueue joined
jnthn vytas: Reproducable? 13:15
rarara_ OK :)
vytas jnthn, trying now, it was very random because i clicked enter on yesterday's opened perl6 repl with empty prompt 13:16
jnthn vytas: Did you rebuild your Perl 6 since then?
vytas agh yes 13:17
jnthn Ah
vytas stupid me sorry
jnthn np...interesting failure mode
13:17 apotheon left 13:19 PsuzeannePsy left 13:22 apotheon joined 13:25 cognominal left 13:26 cognominal joined 13:29 FROGGS joined
FROGGS m: my $foo = "foo foo"; say $foo ~~ s:2x{fo+}=q{bar}; say $foo 13:33
camelia rakudo-moar 271e84: OUTPUT«｢foo｣␤bar bar␤»
13:37 bjz_ joined
FROGGS $_ = "foo foo"; say s:2x{fo+} = q{bar}; 13:38
m: $_ = "foo foo"; say s:2x{fo+} = q{bar};
13:38 bjz left
camelia rakudo-moar 271e84: OUTPUT«｢foo｣␤» 13:38
13:39 ely-se left
Woodi hallo today :) 13:39
FROGGS hi Woodi 13:40
Woodi m: say say "11" ~| "00"
camelia rakudo-moar 271e84: OUTPUT«11␤True␤»
Woodi also: perldoc.perl.org/perlop.html#Bitwis...-Operators && Scheme/C#/PHP/... have such thing 13:41
FROGGS m: say "11" +| "00" 13:44
camelia rakudo-moar 271e84: OUTPUT«11␤»
FROGGS Woodi: why would we need an ~| op here? 13:45
ShimmerFairy Woodi: however, Perl 5 (and the other languages) operate on what we call a buffer (e.g. Blob, Buf, the utf8 type, etc.), and not so much on what we call a string
FROGGS aye 13:46
ShimmerFairy "The behavior of these operators is problematic (and subject to change) if either or both of the strings are encoded in UTF-8" gives it away in the linked perldoc :)
FROGGS m: say "foo foo".match(m:2x/fo+/) 13:49
camelia rakudo-moar 271e84: OUTPUT«Method 'match' not found for invocant of class 'Any'␤ in block <unit> at /tmp/grVoL_tlzf:1␤␤»
FROGGS eww
m: say "foo foo".match(/fo+/, :2x) 13:50
camelia rakudo-moar 271e84: OUTPUT«(｢foo｣ ｢foo｣)␤»
jnthn FROGGS: That'd need to be rx: I think
FROGGS ahh
m: say "foo foo".match(rx:2x/fo+/)
camelia rakudo-moar 271e84: OUTPUT«5===SORRY!5=== Error while compiling /tmp/gVZj_oiDXl␤Adverb x not allowed on rx␤at /tmp/gVZj_oiDXl:1␤------> 3say "foo foo".match(rx:2x/fo+/7⏏5)␤»
jnthn Heh :)
FROGGS :o)
jnthn OK, it hsas to be what you wrote
But m is "do it right now"
FROGGS *nod*
13:51 retupmoca left 13:52 Ven joined, retupmoca joined 13:55 tokuhiro_ joined
vytas m: say [1,[2]].flat 13:56
camelia rakudo-moar 271e84: OUTPUT«(1 [2])␤»
vytas m: say ~[1,[2]] 13:57
camelia rakudo-moar 271e84: OUTPUT«1 2␤»
vytas does flat work only on top level ?
jnthn It works until it encounters a Scalar container, and an Array is a collection of Scalar containers. 13:58
But another way, and array is already flat by definition.
*an
*Put
gah, Friday typing
13:59 rindolf left, pmurias joined 14:00 Ven left, tokuhiro_ left
[Coke] I think you mean typoing. 14:00
14:01 c0rt042 left 14:03 Ven joined 14:04 margeas joined
Ulti "/usr/share/dict/words".IO.words.classify(*.lc.comb.sort.join).sort({$^b.value.elems <=> $^a.value.elems})[0].say is a fun one liner and a little simpler than the equivalent from 2010 on RosettaCode rosettacode.org/wiki/Anagrams#Perl_6 :3 14:05
14:06 Upasaka_ is now known as Upasaka, jkva left
Ven Ulti: also, we can remove the \ from there! 14:07
Ven never understood the diff. between classify and categorize 14:08
14:09 softmoth joined
Ven "+@($_)" <- is the @() really useful here? 14:10
m: my $a = $[1, 2, 3]; say +$a
camelia rakudo-moar 271e84: OUTPUT«3␤»
[Coke] Ulti: if you're calling the same function on each sort item, you only need to specify it once. (I just tripped over this last week) 14:11
CurtisOvidPoe Where can I find documentation for things like Str:D (I don’t know what the :D is and I’d rather read all about that than just have someone tell me) :)
Though if you tell me, I promise not to be too upset. 14:12
14:12 FROGGS_ joined
Ven CurtisOvidPoe: it's a smiley. :D 14:13
CurtisOvidPoe Happy args are happy. But they’re also strange to me :)
Ven CurtisOvidPoe: perl6advent.wordpress.com/2013/12/...pe-object/
CurtisOvidPoe Cheers!
14:13 RabidGravy joined
vytas how do i flatten something like @a = 1..9, a..z; ? 14:15
TimToady CurtisOvidPoe: S12:Abstract_vs_Concrete_types
yoleaux 09:07Z <Ven_> TimToady: irclog.perlgeek.de/perl6/2015-10-15#i_11381191 putting "Foo_" or "Foo-" would mean I'm reinventing namespacing for no reason. I want `::` to be about packages. the name is inside the package, the value is dynamically scoped
synbot6 Link: design.perl6.org/S12.html#Abstract_...rete_types
Woodi my $t = "This is "; my $s = " SPARTA!"; my $p = " Perl6"; say $t ~| $p;
Ven vytas: flat
Woodi m: my $t = "This is "; my $s = " SPARTA!"; my $p = " Perl6"; say $t ~| $p;
camelia rakudo-moar 271e84: OUTPUT«this is Perl6␤»
Ven m: my @a = flat 1..9, 'a'..'z'; say @a.perl; 14:16
camelia rakudo-moar 271e84: OUTPUT«[1, 2, 3, 4, 5, 6, 7, 8, 9, "a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "x", "y", "z"]␤»
14:16 FROGGS left, karol_ left
vytas thanks 14:17
my @a = 1..9, 'a'..'z'; my @b = flat @a; say @b.perl; 14:19
m: my @a = 1..9, 'a'..'z'; my @b = flat @a; say @b.perl;
camelia rakudo-moar 271e84: OUTPUT«[1..9, "a".."z"]␤»
grondilu am I the only one who thinks there are two many quotes in 'a'..'z'? Couldn't we have somthing like say <a..z>? 14:21
Ven please no
grondilu m: say <a..z>.perl # not sure what it is right now 14:22
camelia rakudo-moar 271e84: OUTPUT«"a..z"␤»
grondilu oh, just a string. ok.
14:22 [Sno] left
ilmari <...> is just '...'.split(/\s+/) 14:23
14:23 [Sno] joined
Ulti [Coke]: would that get me the list sorted from highest to lowest though? 14:23
pink_mist <a .. z>?
Ulti or is there a :reverse adverb or something
grondilu m: say <a .. z>.perl 14:24
camelia rakudo-moar 271e84: OUTPUT«("a", "..", "z")␤»
pink_mist d'oh
jnthn m: say <a>..<z>
camelia rakudo-moar 271e84: OUTPUT«"a".."z"␤»
jnthn No quotes! ;)
ilmari m: say <a b>..<c d> 14:25
camelia rakudo-moar 271e84: OUTPUT«2..2␤»
FROGGS_ m: say <a b>...<c d>
camelia rakudo-moar 271e84: OUTPUT«(a b c d)␤»
FROGGS_ m: say <a c>...<x z>
camelia rakudo-moar 271e84: OUTPUT«(a c d e f g h i j k l m n o p q r s t u v w x z)␤»
FROGGS_ m: say <a c>...'z'
camelia rakudo-moar 271e84: OUTPUT«(a c d e f g h i j k l m n o p q r s t u v w x y z)␤»
TimToady Ulti: just use .sort({-*.value.elems) 14:26
er, s/\{//
m: say <1>..<9> 14:28
camelia rakudo-moar 271e84: OUTPUT«IntStr.new(1, "1")..IntStr.new(9, "9")␤»
TimToady hmm...
m: say list <1>..<9>
camelia rakudo-moar 271e84: OUTPUT«(1 2 3 4 5 6 7 8 9)␤»
Ulti TimToady aha thanks
also -*.value might be enough since it will default to the length in numeric context right?
TimToady probably 14:29
14:29 khw joined
Ven m: say [...] <a b>; # grondilu? :P 14:30
camelia rakudo-moar 271e84: OUTPUT«(a b)␤»
Ven aw.
m: say [...] <a z>; # errr >.>
camelia rakudo-moar 271e84: OUTPUT«(a b c d e f g h i j k l m n o p q r s t u v w x y z)␤»
grondilu m: say [..] |<a z>
camelia rakudo-moar 271e84: OUTPUT«5===SORRY!5=== Error while compiling /tmp/wVy7nAJEfB␤Cannot reduce with .. because structural infix operators are diffy and not chaining␤at /tmp/wVy7nAJEfB:1␤------> 3say [7⏏5..] |<a z>␤ expecting any of:␤ argument list␤ …»
grondilu oh yeah 14:31
14:32 ely-se joined
grondilu m: say &[..](|<a z>) 14:32
camelia rakudo-moar 271e84: OUTPUT«"a".."z"␤»
Ven m: say [..]<a z>
camelia rakudo-moar 271e84: OUTPUT«5===SORRY!5=== Error while compiling /tmp/8iSbjP2Bsb␤Cannot reduce with .. because structural infix operators are diffy and not chaining␤at /tmp/8iSbjP2Bsb:1␤------> 3say [7⏏5..]<a z>␤ expecting any of:␤ argument list␤ …»
Ven m: say [..](<a z>)
camelia rakudo-moar 271e84: OUTPUT«5===SORRY!5=== Error while compiling /tmp/J47p9Ms0f2␤Cannot reduce with .. because structural infix operators are diffy and not chaining␤at /tmp/J47p9Ms0f2:1␤------> 3say [7⏏5..](<a z>)␤ expecting any of:␤ argument list␤ …»
grondilu m: say <a z>.&[..]
camelia rakudo-moar 271e84: OUTPUT«Too few positionals passed; expected 2 arguments but got 1␤ in block <unit> at /tmp/H2tEa3a4YK:1␤␤»
nine .tell psch I don't know what reposession is exactly, but what you describe sounds very much like it could be the cause for my strange bug in the bool_enum branch. Attribute and Bool both have strange objectids and both have a BEGIN .^compose 14:33
yoleaux nine: I'll pass your message to psch.
grondilu m: say &[..] |<a z>
camelia rakudo-moar 271e84: OUTPUT«any(sub infix:<..> ($min, $max) { #`(Sub+{<anon|80798352>}+{Precedence}|44790688) ... }, (a z))␤»
Ulti probably yes :) "/usr/share/dict/words".IO.words.classify(*.lc.comb.sort.join).sort(-*.value)[0].say perhaps a bit too terse but kind of logical if you follow left to right
nine jnthn: do you think we're onto something here regarding my message to psch++?
14:34 s_kilk joined
jnthn nine: Maybe, though type objects aren't really meant to be subject to being repossessed 'cus they have no state 14:35
14:35 jkva joined
nine What exactly does reposession mean anyway? 14:35
moritz an object is defined in the setting 14:38
so it's in the setting's serialization context
now a precompiled module changes that object
so it must be in the precompiled module's serialization context
so it must somehow declare that it is now responsible for that object, and the setting should yield control over it 14:39
that's called repossession
jnthn Right. Also I wrote some stuff up on it in the nqp internals course, but moritz++'s example nails it pretty well
[Coke] Ulti's one liner is wicked slow here. :| 14:40
Ulti yup
jnthn The more interesting case for Bool is that it's created in BOOTSTRAP, but it's meta-object and STable are changed while compiling CORE.setting, so I'd expect those two to be repossessed.
[Coke] Ulti: not that it matters much given that our dictionary is mostly ascii, but probably want .fc instead of .lc 14:41
Ulti well since a few days ago I do :P
jnthn++
TimToady Ven: what purpose is served by putting a name into a package if the value cannot be accessed through the package? and it cannot, unless by dynamic scoping you mean "temporizing" rather than what Perl 6 means by dynamic scoping 14:42
14:42 Alina-malina joined
Ven TimToady: the value can (and has to) be accessed through the value 14:43
14:43 zakharyas1 joined
psch jnthn: so scwb{dis,en}able around Attribute.^compose would be the appropriate solution? 14:43
yoleaux 14:33Z <nine> psch: I don't know what reposession is exactly, but what you describe sounds very much like it could be the cause for my strange bug in the bool_enum branch. Attribute and Bool both have strange objectids and both have a BEGIN .^compose
Ven "the value is dynamic, the container isn't" – or something in sp6k
TimToady: I'm not sure what temporizing means, maybe that's what I mean.
psch jnthn: in that context still can't help but wonder why moar doesn't exhibit the jvm SC problem...
ely-se Ven: you mean "through the name"
jnthn psch: Surely that'd mean we lose the changes?
Ven ely-se: yes, I do. thanks
jnthn psch: And Attribute comes out busted still?
TimToady temporizing means you're actually using global containers, which is not thread-safe 14:44
Ven ely-se: maybe you can describe it to TimToady++ better than I do ;-)
psch jnthn: i don't know if we lose the changes on moar, but the SC stays BOOTSTRAP
*on jvm
ely-se dynamic variables that require declarations and those declarations must be either qualified or lexically in scope
jnthn psch: The SC of what exactly?
ely-se but yeah, in P6 I think you can just use regular lexically-scope variables, and use temp 14:45
Ven TimToady: scheme/racket/clojure-style `(parameterize ([param value]) (call...))` doesn't seem to require a global container to me?
jnthn psch: If you're seeing type objects themselves getting repossessed, that may well be a bug
psch jnthn: Attribute. the problem on jvm is that the classHandle for Attribute gets repossessed to CORE.setting
ely-se although preferably thread-local
jnthn OK. I don't think type objects should ever get repossessed. Because...they don't have any state..
ely-se is the modification done by temp thread-local?
psch jnthn: so the problem is that composition somehow repossesses even though it shouldn't..? 14:46
TimToady ely-se: only if the lexical is declared inside the thread
jnthn psch: Sounds like. I can try to find out if we ever repossess a type object on Moar I guess 14:47
ely-se :(
psch jnthn: that could be a hint, thanks 14:48
TimToady temp is not our preferred dynamic scoping mechanism
Ven my is :P
ugexe curious, what does `trace` do that turns the $x from a Hash into a Block? 14:49
TimToady indeed, but my has nothing to do with packages
ely-se It's trivial to implement Clojure-like dynamic variables though
ugexe m: my $x = { :ok(1) }; say $x.WHAT
camelia rakudo-moar 271e84: OUTPUT«(Hash)␤»
jnthn psch: You may want to try and work out what causes that to happen on JVM too
ugexe m: use trace; my $x = { :ok(1) }; say $x.WHAT
camelia rakudo-moar 271e84: OUTPUT«2 (/tmp/BJv0oLmAfY:1)␤my $x = { :ok(1) }␤7 (/tmp/BJv0oLmAfY:1)␤say $x.WHAT␤(Block)␤»
14:49 rindolf joined
ely-se all you need is thread-local variables and try/finally 14:49
Ven TimToady: indeed, it'd be around `our` and `temp` 14:50
TimToady and those are globals
Ven m: module A { our $*a = 5; }; say $A::*a;
camelia rakudo-moar 271e84: OUTPUT«5===SORRY!5=== Error while compiling /tmp/BUbKwFPbN5␤Variable '$A' is not declared␤at /tmp/BUbKwFPbN5:1␤------> 3module A { our $*a = 5; }; say 7⏏5$A::*a;␤»
psch jnthn: yeah, i think i have an idea or two for what i could look for
Ulti [Coke]: mattoates.co.uk/files/perl6/anagram_profile.html looke liks classify-list is taking up about the same time as the function its calling, assuming I read the call graph right 14:52
also the profiler output is way nicer than the last time I looked at it! 14:53
TimToady Ven: we are not going to let people confuse temp variable notation with dynamic variable notation in Perl 6 14:55
if you want temp vars, you don't need the *
and I believe that our lexically-based dynamic variables address the need for thread-local storage adequately without multiplying entities that will lead to user confusion 14:59
We're feebly trying to keep Perl 6 a language that mere mortals can use... 15:00
15:03 n0tjack joined
n0tjack is there a Twitter module for p6? 15:04
ely-se How about immortals like me? 15:06
TimToady you can write modules :)
ely-se mortules 15:07
n0tjack TimToady: I "could" yodel, too. Just don't feel like learning how ;)
15:07 g4 left
n0tjack I figure my next little self-assignment to learn p6 would be a OEIS twitter bot. You @tweet it a series of integers, and it uses the OEIS API to suggest potential functions to generate that series. 15:08
15:09 weihan joined
n0tjack To date, all my projects have been functional, which is where I'm comfortable. I figured I'd try something stateful, with more moving parts. 15:09
15:10 diana_olhovik left
TimToady I don't see where state comes into this problem 15:10
n0tjack connection to twitter
TimToady I guess
15:10 zakharyas1 left
n0tjack and I'd figured I'd put in a little memoization/caching 15:10
I don't want to hammer the OEIS if I can help it 15:11
TimToady I'd just load OEIS into a trie or some such 15:12
mrf Is there a characted class for 'Is a Control char'. Or better is a 'Non Whitespace Control char'?
TimToady m: say uniprop(3) 15:13
camelia rakudo-moar 271e84: OUTPUT«Cc␤»
mrf is learning grammers by generating the RFC-2822 spec
jnthn I think the <control> rule in a regex will doit
geekosaur m: say uniprop(160)
camelia rakudo-moar 271e84: OUTPUT«Zs␤»
mrf TimToady: is that not the physicall CTRL key? 15:14
TimToady wha? 15:15
TimToady is not even sure that rises to the level of category error...
pink_mist keyboard scancodes don't have unicode properties 0_o
Ven TimToady: I don't think mortals will use many dynamic variables. And maybe they'd be grateful they're not getting their dynamic values overriden by some other module
mrf is clearly being misleading here. 15:16
ugexe token obs-NO-WS-CTL { <[\c[1]..\c[8]\c[11,12]\c[14]..\c[31]\c[127]]> }
mrf I am looking for something that will match all ASCII control characters excluding \s matchs
ugexe you can use this as a reference mrf: github.com/ugexe/Perl6-Grammar--HT...FC5322.pm6
mrf ugexe: That looks good
hoelzro o/ #perl6 15:18
mrf ugexe++ Thank you that looks perfect 15:19
jnthn missed why <control> doesn't cut it... :)
mrf ugexe: Also indicates that serveral of the shared rules/tokens between MIME headers and HTTP could form some sort of parent class.
jnthn: Can I do <-[\s]> & <control>? 15:20
ugexe mrf: yep. they share tokens
mrf s/class/grammer
jnthn mrf: yes
Well
<control-[\s]>
tony-o m: use trace; my $x = { :ok(1); }; $x<ok>.say;
camelia rakudo-moar 271e84: OUTPUT«2 (/tmp/cvO2oQ62_u:1)␤my $x = { :ok(1); }␤7 (/tmp/cvO2oQ62_u:1)␤$x<ok>.say␤Type Block does not support associative indexing.␤ in block <unit> at /tmp/cvO2oQ62_u:1␤␤Actually thrown at:␤ in block <unit> at /tmp/cvO2oQ62_u:1␤␤»
tony-o m: my $x = { :ok(1); }; $x<ok>.say;
camelia rakudo-moar 271e84: OUTPUT«1␤»
15:20 muraiki joined
ugexe mrf: if you break your tokens into 'roles' then you can share them between each other 15:21
mrf jnthn: is that literally 'The Control names group without \s'? Its not a syntax I have seen (All the perl6 regex is very new) 15:22
ugexe grammar HTTP::Grammar does (RFC1, RFC2) { }
jnthn mrf: Yes
ugexe im not sure how well the perl6 tokens for things like \n, \s, etc map up with the RFC definitions though 15:23
mrf jnthn: thats cool.
15:24 ely-se left
konobi well, there's the rfc standard for "Message" which is the same for http, email, nntp, etc. 15:24
mrf ugexe: for most of the 2822 spec it uses Folded whitespace which seems to be \s*
ugexe: OOI why do all the named groups have a preceeding period. e.g <.FWS> ? 15:26
ugexe: ^ In Grammer::HTTP::RFC5322
ugexe because capturing those slow it down immensely, and those captures are not used
mrf ugexe: so <> is a named version of ()? 15:27
ugexe m: say "\n" ~~ /\s/
camelia rakudo-moar 271e84: OUTPUT«｢␤｣␤»
mrf and <.> is like []
ugexe thats why you cant just use the default perl tokens for everything 15:28
for the most part 15:29
i dont understand the comparison to [] though
mrf [] is a non capturing group no? 15:30
ugexe: IIRC WSP is \h and \t only and FWS includes \v, \n and \r so becomes \s 15:31
konobi that'd be (?:)
15:32 hohoho left
ugexe [] is a group yes. <.token-name> is the same as <token-name>, except there will be no token-name => SomeValue in the match object 15:32
mrf ugexe: I am not sure how that is different from 'non-capturing named group' but I may be missing things 15:33
konobi: I was under the impression that (?:) had been replaced by []
ugexe if you wanted to repaste your entire grouping in between [ and ], it would be no different
jnthn Well, not quite 15:34
[...] is just a group
psch jnthn: i didn't see any repossession of SMOs with isconcrete == 0 (fwiu, that's equivalent to "is a TypeObject")
jnthn If you capture inside of the group it's still a capture
psch: Yes, isconcrete == 0 is exactly equivalent
psch jnthn: but considering type objects never get repossessed, isn't a comparison that checks the SC when looking for an attribute on an object bound to eventually fail? 15:35
jnthn: i mean, that's the bandaid, right. "ignore SC when looking for the hintMap"
jnthn I'm not sure why we'd look at SC there; on Moar we just look at the type object to match, I *think* 15:36
psch github.com/perl6/nqp/blob/master/s...e.java#L17 # this line used to not have the ".st" on both sides of == 15:37
15:37 timo joined
psch i'm just not sure that's a sufficient fix, which is why i keep revisiting it... :) 15:37
15:37 timo is now known as Guest16417 15:38 llfourn left
jnthn It's odd that it'd need that 15:38
15:38 Guest16417 is now known as timotimo
jnthn Comparing the two type objects should just work out 15:38
So yeah, I agree something deeper is going wrong
psch maybe we need an .equals..?
jnthn No, they really should be exactly the same objects 15:39
Reference equality should be enough
psch well, the LHS is an Attribute object from CORE, the RHS is the Attribute type object from BOOTSTRAP
and the classHandle is both cases references the SC
*in
jnthn Yeah, that's what shouldn't be happening. Hmm. 15:40
psch the classHandle shouldn't know the SC?
jnthn No 15:41
The Attribute type ocject from cORE should be the one from BOOTSTRAP 15:42
There shouldn't be two of them
I don't see how that's happening
nine That's exactly what I see with Bool
jnthn But curiously not with Attribute on Moar, it seems. 15:43
nine As far as I remember that was my first question for this bug: if type objects created in BOOTSTRAP are copied somewhere
15:43 rindolf left
jnthn nine: Well, apparently this one is being, but "should it be"...well, no. 15:44
15:44 blackcat joined, blackcat is now known as loren
jnthn github.com/MoarVM/MoarVM/blob/mast...aque.c#L24 gets by on pointer equiv, not STable equiv, just to verify 15:44
15:44 Ven left 15:46 zakharyas left
loren I write a simple ftp client module, Net::Ftp will be a valid name? github.com/araraloren/Net-Ftp 15:51
avar Net::FTP would be better :) 15:52
15:53 s_kilk left
loren Hm, thank u, avar 15:54
15:55 uruwi joined, s_kilk joined 15:57 llfourn joined, tokuhiro_ joined
jdv79 looks like the tests for a module of mine are faster since maybe 6 months ago 15:59
now only 1 order of mag behind p5 instead of 2 i think they were 16:00
profile still dominated with listy stuff though:(
jnthn jdv79: In exclusive time or inclusive? 16:01
16:01 tokuhiro_ left
jnthn It'll often rank high in inclusive 16:01
jdv79 exclusive
though to a lesser degree than before iirc 16:02
jnthn Yeah
There's still some hotspots, though they tend to be more isolated and smaller pieces of code than before at least
16:07 Ven joined
dalek kudo/nom: 9a78309 | ugexe++ | src/core/Supply.pm:
Fix precedence for Supply status

Fixes JVM build
  travis-ci.org/rakudo/rakudo/jobs/85681162#L330		 16:08
kudo/nom: c60e687 | lizmat++ | src/core/Supply.pm:
Merge pull request #553 from ugexe/patch-5

Fix precedence for Supply status
16:09 mayuresh joined
mayuresh hello :) 16:09
dalek kudo/p6for_op: ba14084 | (Stefan Seifert)++ | src/Perl6/Actions.nqp:
Work in progress: p6for desugar op

Move some <for> logic into a desugar op for easier handling by the optimizer. Label handling still broken in this commit.		 16:10
mayuresh i just realized that perl6 is such a beast! 16:11
i might as well base my 'pxn' project around the ideas of smalltalk execution environment.
ugexe a beast with wings
mayuresh pxn == Perl6 eXecution eNvironment. :) 16:12
ugexe, yeah, a beast with wings, quite like the mythical pegasus 16:13
;)
16:14 ^elyse^ joined
pmurias mayuresh: what is pxn supposed to be? 16:15
16:15 rindolf joined
[Coke] pmurias: IIRC, a new backend that doesn't separate the compiler from the vm. 16:16
mayuresh pmurias, a combined compiler + virtual machine.
pmurias combined in what way? 16:17
mayuresh one should be able to "run" perl6 code the way people presently run perl5 code.
no need to first compile it, and then separately execute it on a 'vm'. 16:18
flussence so basically PHP?
[Coke] mayuresh: you know you don't have to do that with rakudo,r ight?
mayuresh flussence, yeah somewhat! :)
[Coke] you can run it without generating moarvm bytecode files.
mayuresh [coke], i had no idea about that. 16:19
[Coke] pmurias: I just saw your blog post about nqp-js! that's great news!
mayuresh: ./perl6 -e '3.say' - that doesn't generate a moarvm bytecode file on disk.
mayuresh nqp-js! does that mean p6 will be compiled for node.js?
[coke], what about an entire program written in a file?
[Coke] same. 16:20
mayuresh cool.
pmurias mayuresh: commpiled to js, to be run either on node.js or in the browser
[Coke] you only get moarvm bytecode output if you ask for it with --target
mayuresh pmurias, you worked on that! cool man.
[coke], then why go for all that 'vm' stuff in the first place?
pmurias [Coke]: node nqp-bootstrapped.js runs 8 nqp tests, working on bug fixes and missing small things to pass the other ones 16:21
[Coke] pmurias: still need to config with moarvm & js in the backend?
mayuresh: it's still running in the vm.
mayuresh which 'vm'? if not moar, then there's a 'vm' inside of rakudo? 16:22
pmurias [Coke]: yes
[Coke] mayuresh: no, moar.
mayuresh dang, now i am confused.
[Coke] mayuresh: it's running in the moar vm. there's just no external moarvm bytecode files generated.
mayuresh so if my system has just rakudo, and no moarvm, will i be able to "run" p6 code?
[Coke] no 16:23
flussence unless it's nqp-jvm
mayuresh :)
pmurias [Coke]: nqp-js running on top of moarvm can compile nqp-js, nqp-js running on top node of can't compile itself yet
[Coke] pmurias: ok.
mayuresh [coke], does that mean, p6 "has" to rely on those two separate components? a compiler (rakudo) and a 'vm' (moarvm)?
don't you think it would be better to have a slimmer and more focused all-in-one execution system for p6? 16:24
jnthn mayuresh: Problem not *has* to, but it's the practical way to implement it.
mayuresh i have just been intrigued by process managers embedded in interpreters. 16:25
jnthn mayuresh: Something somewhere needs to do the gutsy stuff (garbage collection, I/O, abstracting away OS differences in various other places).
[Coke] mayuresh: I might slightly prefer it, but what jnthn said. it took us a long time to get -here-. it would have taken even longer to get there without going here first.
mayuresh ture.
true
i agree
[Coke] so, I wish you well on your journey, but I'm staying focused on rakudo.
jnthn mayuresh: At the same time, you really need to write most of a Perl 6 compiler in Perl 6
16:25 ^elyse^ left
mayuresh and i don't look down on your work at all. 16:25
[coke] please do, and thanks for the wishes.
jnthn mayuresh: For example, you'll really struggle to implement Perl 6 if you don't parse it using a Perl 6 grammar. 16:26
mayuresh jnthn, okay.
i think it would be better if i do work on the 'pxn', i really want to learn about how to do what i intend to do. :) 16:27
[Coke] pmurias: sadly, the build died. presumably my old not-quite-the-right-node issues. :|
jnthn Learning about compilers/interpreters/language impl is always worth it :)
pmurias [Coke]: please paste the error log
mayuresh [coke] why not use io.js? 16:28
jnthn, it is very intriguing, the way those systems work. :)
konobi io.js doesn't exist anymore =0P
jnthn mayuresh: Very. It's a fun area. :-)
timotimo i think they merged back together with node.js
konobi yup
pmurias didn't node.js merge back into io.js? 16:29
konobi io.js -> node.js
s_kilk io.js essentially became node.js 4.0
[Coke] pmurias: : gist.github.com/coke/1b4ce850129de526dd58
16:30 s_kilk left
mayuresh s_kilk, coolness. :) 16:30
pmurias mayuresh: re better to have it all-in-one, not really
konobi though node is now v4.2.1
mayuresh pmurias, may i know why? to understand why i am thinking weirdly. 16:31
flussence even perl5 has separate parser and opcode vm stages 16:32
[Coke] mayuresh: what problem are you trying to solve? 16:33
pmurias mayuresh: moarvm and rakudo are kept seperate mostly to keep order in the repos and to allow different backends
mayuresh: having the code just tangled together doesn't really give us anything
[Coke]: you are building things from inside the nqp-js directory
[Coke]: it seems the modules built correctly but touch node_modules/npm_installed failed 16:34
mayuresh [coke], not solving a specific problem, especially when none exists in a big way.
[coke], just learning, and going in a different direction than what exists at the moment.
jnthn mayuresh: There's probably lots of interesting answers, but for me it's an exercise in complexity management - that is, keeping different things apart, so they can be understood in relative isolation - because having to understand it all at once is too much for anyone.
konobi pmurias: oh... nqp-js is atop node? 16:35
16:35 jkva left
mayuresh jnthn, okay, is that high level of complexity there because p6 is such a beast of a language? 16:35
jnthn, btw, if i understood it right, live object systems like smalltalk, java, dart, all do need a 'vm' like approach, right? 16:36
jnthn, is that why p6 too is going with a 'vm'?
[Coke] is node_modules supposed to be a directory under nqp build dir? 16:37
pmurias yes
flussence any non-trivial language (basically, anything more useful than VB/PHP) uses a vm, whether or not it makes it visible
mayuresh by live objects, i imply the opposite of dead objects, like in c++ and ada.
[Coke] didn't get built.
jnthn mayuresh: It's a factor to some degree, but if you look at most language implementations that are used in the wild, they're all pretty involved in their own ways, and the folks building them have found typically found ways to cope with that by breaking it into manageable parts. 16:38
[Coke] I just did a config, then 'make js-all' - maybe a mis-ordered dep? i see it in there under one of the build targets.
pmurias did stuff get installed into /Users/williamcoleda/node_modules/nqp-runtime/? 16:39
[Coke] yes
16:39 weihan left, ^elyse^ joined
pmurias konobi: the browsers are the real target I'm interested in, I'm just using node.js because it always to easily run stuff on the command line 16:40
mayuresh jnthn, true, but, i am so impressed with smalltalk, it's a zang crazily well designed system of live objects.
the only thing that's better would be "self".
jnthn Yes, Smalltalk is neat :-)
We borrowed ideas from there and from CLOS for the Perl 6 MOP.
pmurias [Coke]: seems strange
jnthn needs to go and make dinner :-)
bbl o/
Ven
.oO( we added only a few operators )
mayuresh jnthn, you guys went all the way of 'clos'!! that's rather neat. 16:41
pmurias [Coke]: where did you get your node.js from?
mayuresh jnthn, have a good meal. :)
16:43 weihan joined
konobi pmurias: mkay... well, if you have node/npm specific questions, lemme know 16:43
mayuresh just a bit off-topic; are any of you guys on twitter too? 16:45
16:45 espadrine left
dalek p: 0c4bb9a | (Pawel Murias)++ | src/vm/js/ (4 files):
[js] Make nqp::unshift work on captures.

Pass tests 06 and 08.		 16:46
p: 6fe7900 | (Pawel Murias)++ | tools/build/ (2 files):
[js] fix a build issue when there is a node_module in a ancestor directory.
16:46 ab6tract joined
ab6tract o/ #perl6 16:47
pmurias [Coke]: it seemed there was a build issue if you had a node_modules already in your homedirectory
[Coke]: thanks for finding the bug
mayuresh hi ab6tract. :)
16:47 llfourn left
ab6tract hi mayuresh :) 16:48
16:48 weihan left
ab6tract can someone in the know tell me what this doc should look like in light of multi-dem ASSIGN-POS? doc.perl6.org/routine/ASSIGN-POS 16:48
16:48 weihan joined
konobi pmurias: to avoid npm from doing silly things like using node_modules in a homedir, make sure that there's a package.json in the folder above node_modules 16:48
ab6tract is it even supported outside of core?
sorry to keep harping on this, but it's a blocker on the async Terminal::Print :) 16:49
[Coke] pmurias: macports 16:50
psch m: say Any.^can('ASSIGN-POS')[0].candidates[*-1].signature 16:51
camelia rakudo-moar 271e84: OUTPUT«($: **@indices, *%_)␤»
psch interestingly that doesn't have a defined invocant in CORE...
ab6tract ooof... what does $: mean in that signature? 16:52
psch ab6tract: it's an anonymous invocant
ab6tract psch: m: class A { has @!a; method ASSIGN-POS(**@b) { my $v = @b.pop; @!a[ @b[0],@b[1] ] = $v }; method AT-POS(**@b) { @!a[ @b[0], @b[1] ] } }; my $t = A.new; $t[6;7] = "fooo"; dd $t[6;7]
erm 16:53
m: m: class A { has @!a; method ASSIGN-POS(**@b) { my $v = @b.pop; @!a[ @b[0],@b[1] ] = $v }; method AT-POS(**@b) { @!a[ @b[0], @b[1] ] } }; my $t = A.new; $t[6;7] = "fooo"; dd $t[6;7]
camelia rakudo-moar 271e84: OUTPUT«("fooo", Any)␤»
ab6tract psch: ^^
m: m: class A { has @!a; method ASSIGN-POS(**@b) { my $v = @b.pop; $v.WHAT.say @!a[ @b[0],@b[1] ] = $v }; method AT-POS(**@b) { @!a[ @b[0], @b[1] ] } }; my $t = A.new; $t[6;7] = "fooo"; dd $t[6;7]
camelia rakudo-moar 271e84: OUTPUT«5===SORRY!5=== Error while compiling /tmp/FV6_5Rbx67␤Two terms in a row␤at /tmp/FV6_5Rbx67:1␤------> 3-POS(**@b) { my $v = @b.pop; $v.WHAT.say7⏏5 @!a[ @b[0],@b[1] ] = $v }; method AT-PO␤ expecting any of:␤ infix␤ infi…»
ab6tract m: m: class A { has @!a; method ASSIGN-POS(**@b) { my $v = @b.pop; $v.WHAT.say; @!a[ @b[0],@b[1] ] = $v }; method AT-POS(**@b) { @!a[ @b[0], @b[1] ] } }; my $t = A.new; $t[6;7] = "fooo"; dd $t[6;7]
camelia rakudo-moar 271e84: OUTPUT«("fooo", Any)␤»
psch ab6tract: CORE adds the last element seperately 16:54
or assigns, rather
ab6tract psch: ^^ that's following what I found in src/core/Array.pm 16:55
konobi pmurias: seems like that build structure is kinda wonky for npm
psch ab6tract: oh, i was looking at Any
dalek pan style="color: #395be5">perl6-examples: 4b3170e | moritz++ | categories/games/blackjack.p6:
Fix blackjack.p6

closes #35		 16:56
ab6tract ' self.AT-POS(@indices) = value' ... curious
16:56 weihan left
ab6tract psch: a bit surprised not to be finding this stuff in Positional 16:57
I'll look into Any.pm
nine ab6tract: array_slice.pm
ab6tract thanks nine! nice to read you btw 16:58
mayuresh alrighty, time for me to hit bed.
pmurias konobi: what would be the sane way of doing the build?
mayuresh thanks for the healthy discussion people. 16:59
pmurias konobi: or is just having package.json enough?
mayuresh it has been a pleasure chatting with you all. :)
TimToady sleep well
mayuresh mr. wall!!
cool.
konobi pmurias: lemme see if I can build this locally
mayuresh sure, yes, will. :)
TimToady you should try to avoid excitement at bedtime :) 17:00
mayuresh :D
it's unavoidable. ;)
bye. :)
17:00 mayuresh left
nine ab6tract: likewise :) 17:00
ab6tract nine: so am I to read this as: the answer is BIND-POS ? ;P 17:01
17:02 Ven left
dalek kudo/nom: 82d4eac | lizmat++ | src/core/ (2 files):
Make X::Adverb::Slice more general as X::Adverb
 17:02
pan style="color: #395be5">perl6-examples: 62bfdc9 | moritz++ | categories/games/tictactoe.p6:
GLR-Fixes for tictactoe.p6, closes #34
konobi pmurias: any nqp-js specific args to Configure.pl?
dalek ast: a367d98 | lizmat++ | S (3 files):
Handle X::Adverb::Slice -> X::Adverb conversion
 17:03
pmurias --backends=moar,js
konobi ta
ab6tract pmurias: you are awesome :)
pmurias konobi: ta? 17:04
konobi pmurias: "thanks" 17:05
pmurias ok :)
dalek pan style="color: #395be5">perl6-examples: 6bbe840 | moritz++ | categories/games/wizard.pl:
Avoid a warning in wizard.pl

no fix the the actual problems yet
moritz uhm, did wizard.pl ever work? 17:06
nine ab6tract: sorry, network is laggy as hell. Am not really sure what you're trying to do, but array_slice.pm contains an important part of the code dealing with...well array slices ;) 17:07
konobi pmurias: the "sleep" module probably wants to be updated to use "bindings" 17:09
pmurias it doesn't build? 17:10
ab6tract nine: I'm trying to write a multi-dim ASSIGN-POS for a class. I can't just use a 'handles' trait because I want to use the ASSIGN-POS sub to send them into a supply
konobi it does... but you'll never get a debug build that way
dalek pan style="color: #395be5">perl6-examples: 5296582 | moritz++ | categories/games/wizard.pl:
Partial fix for wizward.pl

this cannot have worked for a looong time. There were type constraints that had no chance of ever working, so it must be from a type before we enforced attribute type checks.		 17:11
nine ab6tract: I guess you'll need to implement a couple other methods like EXISTS-POS
17:11 margeas left
konobi pmurias: so basicially "node_modules" is the build output for the js backend? 17:11
ab6tract nine: ok, I've got AT-POS already
pmurias yes
konobi: the sleep module is just something I found on npm 17:12
dalek kudo/p6for_op: ee8b0cb | (Stefan Seifert)++ | src/Perl6/Actions.nqp:
Work in progress: p6for desugar op

Move some <for> logic into a desugar op for easier handling by the optimizer. S04-statements/for.rakudo.moar fails Need to do the same for statement modifier <for>, too.		 17:13
ab6tract nine: from a class composition standpoint, though, the absolute most ideal situation would be to be able to define something like: method ASSIGN-POS($x,$y,$val)
konobi pmurias: mkay... i'd suggest an nqp-js directory with node_modules inside it... then you can add package.json in there, specify required dependencies (including elsewhere on the filesystem) 17:14
ab6tract if I only want to provide two-dimensional interaction with my object
nine ab6tract: sure. As long as we keep the hot paths fast :) Array interaction is kind of important in Perl ;) 17:15
Have to leave now... cya
konobi pmurias: (even running `npm init nqp-js` should give a reasonable layout) 17:16
ab6tract nine: still getting the same ol' error. "Cannot assign to a readonly variable or a value". AT-POS works fine though 17:17
pmurias konobi: that nqp-js directory should then be used as the build output? 17:18
konobi pmurias: yup... and I'd have a lib directory and put the generated JS files in there 17:19
ab6tract nine: even stranger, ASSIGN-POS works fine during subemethod BUILD
nine: gist.github.com/ab5tract/bb20ff79dddf965febe4 17:20
in case it piques your interest
pmurias konobi: is npm expected to rebuild all my dependencies everytime I run npm install src/vm/js/nqp-runtime? 17:21
ab6tract ... doh
konobi not all of them
ab6tract nope, nevermind. Not a 'doh' after all. Storying a container in the @!grid during BUILD doesn't change anything 17:23
konobi pmurias: if you declare nqp-runtime as a local fs dependency... then `npm install .` will rebuild all
17:23 n0tjack left
pmurias konobi: it rebuilds things such as sleep 17:24
17:24 duncan_dmg left 17:25 uruwi left
konobi pmurias: yeah, sorta the nature of the npm beast 17:29
17:31 n0tjack joined, pfhork joined
pfhork is there a similar perl5 channel? 17:31
konobi pmurias: though you could use npm link to avoid that
pmurias pfhork: there is a #perl channel for Perl 5 questions 17:34
konobi but that's on irc.perl.org
[Coke] konobi: there's a #Perl on freenode, also. 17:35
konobi oh right
pfhork thanks; I didn't know my way around the irc commands and listing channels and such (links to ref/docs accepted). ttyl 17:36
pmurias konobi: #perl on freenode is the one where you want to ask questions on 17:37
dalek rl6-roast-data: 43e9e89 | coke++ | / (9 files):
today (automated commit)
17:38 Kogurr joined
konobi pmurias: yeah... just depends if you're looking for p5porters style channel 17:38
[Coke] pmurias: one failure in nqp-js on the last run:
17:38 n0tjack left
dalek kudo/nom: 80112e1 | lizmat++ | src/core/Any-iterable-methods.pm:
Introduce :k/v/kv/p adverbs for List.grep

  :k is equivalent to current grep-index
   in fact, grep-index just now calls grep(:k)
  :kv generates a list for use in for @a.grep(:kv) -> \k, \v {}
  :p generates a list for use in for @a.grep(:p) -> \pair {}
  :v is just a synonym for no adverb at all
The :k, :kv and :p adverbs can also be negated: it will just act as if no adverb was specified (handy for variable :k($force-key) case)
A negated :v will cause a failure, as well as any combination of :k, :kv,
  :p or :v, or any unrecognized adverb.
grep-index is now also between 1.5 and 2x faster.		 17:38
[Coke] github.com/coke/perl6-roast-data/b...y.out#L143
17:38 n0tjack joined, AlexDaniel joined
[Coke] rakudo-jvm horribly borked. 17:38
konobi tap would be a devDependency 17:39
[Coke] rakudo-moar and moar-jit failing differently on require tests.
konobi (from npm that is)
[Coke] still getting random failures on S17 tests. :(
pmurias [Coke]: thanks, I forgot to add a dependency 17:40
konobi yay for tap
lizmat [Coke]: in parallel testing, or also just by themselves ? 17:42
[Coke] the daily runs don't do parallel. 17:43
even in ||, our S17 tests shouldn't be failing.
(like, it's worse if they only fail then!)
oh, wait, moar jit is working, moar-nonjit is dead.
doing a local build to see if I can duplicate the issue. 17:44
17:44 ^elyse^ left, llfourn joined 17:45 hohoho joined, spider-mario joined 17:47 kmel joined 17:49 llfourn left
dalek kudo/nom: 8a8bf63 | lizmat++ | src/core/Any-iterable-methods.pm:
Make sub grep() and sub grep-index() also work
 17:49
17:49 hohoho left
dalek ecs: eb42709 | TimToady++ | S (11 files):
remove mentions of Parcel/parcel
 17:50
17:51 uruwi joined
[Coke] looks like it died on Error while constructing error object:Could not locate compile-time value for symbol X::Syntax::ConditionalOperator::PrecedenceTooLoose 17:58
but I cannot duplicate it locally right now.
17:58 tokuhiro_ joined 18:03 tokuhiro_ left 18:05 ab6tract left
[Coke] down to 61 test files in roast that aren't part of spec or stress test. 18:06
is t/spec/S05-grammar/std.t reasonable? 18:07
kmel hello 18:08
[Coke] kmel: hio 18:09
18:09 vendethiel joined
kmel hi [Coke] 18:10
I was wondering what editors do you use for Perl 6?
PerlJam kmel: vi(m) 18:11
boggard notepad 18:12
18:12 pfhork left
lizmat m:dd <a b c>.grep(Str,:k) 18:12
m: dd <a b c>.grep(Str,:k)
camelia rakudo-moar 82d4ea: OUTPUT«("a", "b", "c").Seq␤»
[Coke] m: END: { say "end"}; say "mid" ; BEGIN: { say "beg"; exit} 18:13
camelia rakudo-moar 82d4ea: OUTPUT«end␤mid␤beg␤»
uruwi Atom
lizmat PerlJam: should soon say 1,2,3
kmel vim!! I am jealous. i never learned how to use it.
[Coke] ^^ should that print mid?
PerlJam lizmat++
[Coke] m: END { say "end"}; say "mid" ; BEGIN { say "beg"; exit}
camelia rakudo-moar 82d4ea: OUTPUT«beg␤»
kmel uruwi : i am using atom too
[Coke] er, I meant that, because there's a test where the END doesn't seem to run.
uruwi Prefer it to Notepad++ now
[Coke] t/spec/S04-phasers/exit-in-begin.t 18:14
uruwi Only thing keeping me from using it all the time is Danmakufu :P
lizmat [Coke]: so what was the difference ?
kmel Danmakufu?
lizmat END: as a label ?
[Coke] no, that was a typo on my part.
trying to golf it now.
kmel i use n++ when i am on windows 18:15
uruwi kmel dmf.shrinemaiden.org/
[Coke] m: END {say 2}; BEGIN {exit}; 18:16
camelia ( no output )
[Coke] ^^ there, that's the golf.
test assumes exit in BEGIN still allows END's to run.
lizmat [Coke]: well, not running would make sense if the exit was in the *first* BEGIN block seen 18:17
18:17 ab6tract joined 18:18 n0tjack left
lizmat otherwise, other BEGIN blocks may have set things up that are expected to be handled by END blocks also already set up 18:18
[Coke] m: END {say 2}; BEGIN {""}; BEGIN {exit};
camelia ( no output )
[Coke] I wouldn't expect the order of begin blocks to impact if any end blocks were run
lizmat well, BEGIN blocks *are* executed in order seen, no ? 18:19
I mean, they're executed immediately
PerlJam aye
[Coke] Sure, but why should that impact END blocks?
uruwi But why doesn't the same thing happen when you put the colons in?
[Coke] either an exit stops processing immediately, or END blocks are still run. 18:20
lizmat uruwi: because then they're just labels
uruwi Ah
[Coke] uruwi: yah, that was a typo on my part, apologies.
lizmat exit is supposed to run ENC blocks, afaik
uruwi Suggest not allowing BEGIN and END as label names.
[Coke] ok. Will ticket it, fudge the test file, and add it to the spectest run. 18:21
18:21 n0tjack joined
uruwi Also does anyone else think tau should be a built-in constant? 18:21
[Coke] nope.
uruwi I know there's a module for it, but I don't feel like saying "use Math::Tau" just to use it
[Coke] we've already had that discussion. and it's dead simple to add it in your code.
PerlJam uruwi: constant tau = 2 * pi; # done :) 18:22
uruwi :P
[Coke] ... I am surprised to find that someone put a module for that out there. 18:23
n0tjack m: constant τ = 2*pi; say τ; 18:24
camelia rakudo-moar 8a8bf6: OUTPUT«6.28318530717959␤»
n0tjack m: constant τ = 2*π; say τ;
camelia rakudo-moar 8a8bf6: OUTPUT«6.28318530717959␤»
18:24 loren left
n0tjack heh, of course π is predefined 18:24
[Coke] uruwi: irclog.perlgeek.de/perl6/2014-04-29#i_8653594 18:25
TimToady m: constant τ = 2 * π; say τ;
camelia rakudo-moar 8a8bf6: OUTPUT«6.28318530717959␤»
[Coke] m: say e**(i*π);
camelia rakudo-moar 8a8bf6: OUTPUT«-1+1.22464679914735e-16i␤»
TimToady oh, you did that already :)
18:26 Hor|zon left
uruwi m: import Math::Tau; say tau; 18:26
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/OeP5t4D3tk␤Could not find module Math::Tau to import symbols from␤at /tmp/OeP5t4D3tk:1␤------> 3import Math::Tau7⏏5; say tau;␤»
uruwi m: use Math::Tau; say tau; 18:27
camelia rakudo-moar 8a8bf6: OUTPUT«===SORRY!===␤Could not find Math::Tau in any of:␤ file#/home/camelia/.perl6/2015.09-393-g8a8bf63/lib␤ inst#/home/camelia/.perl6/2015.09-393-g8a8bf63␤ file#/home/camelia/rakudo-inst-2/share/perl6/lib␤ file#/home/camelia/rakudo-inst-2/share/per…»
[Coke] uruwi: m has no modules to speak of. star: has some, but not that one.
TimToady m: say e**π\i;
camelia rakudo-moar 8a8bf6: OUTPUT«-1+1.22464679914735e-16i␤»
TimToady star: use Math::Tau; say tau;
camelia star-m 2015.09: OUTPUT«===SORRY!===␤Could not find Math::Tau in any of:␤ file#/home/camelia/.perl6/2015.09/lib␤ inst#/home/camelia/.perl6/2015.09␤ file#/home/camelia/star-2015.09/share/perl6/lib␤ file#/home/camelia/star-2015.09/share/perl6/vendor/lib␤ file#/home/cam…»
uruwi Since when did Perl6 have a backslash operator?
PerlJam uruwi: you can probably guess that I would not mind it if tau were built-in, but I also have no strong feeling on the matter. 18:28
[Coke] I was going to show my kid how to find an approximate fraction for pi using perl6 last night, but she ran away.
TimToady it's not an operator, it's just a degenerate unspace
[Coke] m: \s\a\y 3; #should fail
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/zxQJL9ptee␤Malformed replacement part; couldn't find final \␤at /tmp/zxQJL9ptee:1␤------> 3\s\a\y 3; #should fail7⏏5<EOL>␤ expecting any of:␤ infix stopper␤»
uruwi m: say e**πi
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/lFz7BgNVs_␤Undeclared routine:␤ πi used at line 1␤␤»
TimToady m: say e**(π)i;
camelia rakudo-moar 8a8bf6: OUTPUT«-1+1.22464679914735e-16i␤»
uruwi m: say 2 π
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/QFHNSz2zf9␤Two terms in a row␤at /tmp/QFHNSz2zf9:1␤------> 3say 27⏏5 π␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement …»
TimToady parens work too, but are twice as long
uruwi m: say 2(π) 18:29
camelia rakudo-moar 8a8bf6: OUTPUT«Cannot find method 'CALL-ME'␤ in block <unit> at /tmp/WBGsTNnC1u:1␤␤»
uruwi m: say (2)π
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/hxW79JqLrW␤Two terms in a row␤at /tmp/hxW79JqLrW:1␤------> 3say (2)7⏏5π␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement…»
TimToady it's i that's the operator here
uruwi Aaaah
18:29 rurban left
TimToady it's just you don't need the \ when you say 42i 18:29
18:31 sufrostico left
TimToady m: sub postfix:<π>($n) { $n * π }; say 2π 18:33
camelia rakudo-moar 8a8bf6: OUTPUT«6.28318530717959␤»
TimToady so one could do the same trick with π
m: sub postfix:<π>($n) { $n * π }; say i\π R** e 18:34
camelia rakudo-moar 8a8bf6: OUTPUT«-1+1.22464679914735e-16i␤»
TimToady but that way lies madness :) 18:35
18:35 Hor|zon joined
TimToady or at least irrationality :) 18:35
PerlJam Math::Tau seems fairly ridiculous as far as modules go, but perhaps a More::Math module that had a bunch of useful mathematical constants and operators and such would be nice. 18:36
ab6tract interesting.. gist does not support directories, and AFAICT use lib does not support '.' with 'Foo::Bar::Nest' in a single file called Foo-Bar-Nest.pm6 18:37
18:38 uruwi left
kmel anyone tried head numbers in pod? 18:38
i can't make it work
i tried design.perl6.org/S26.html#Numbered_headings but no luck 18:39
PerlJam doubts that any such configuration works yet
vendethiel explained TimToady++'s RosettaCode/Mad Libs 18:41
ab6tract ah ha! use lib './'
:)
vendethiel three times today. that was fun :P.
18:42 uruwi joined, edehont joined
n0tjack in J, the easiest way to get a harmonic series is - reduce (1,2,3,4) , because that translates to 1-2-3-4 in J, which executes RTL, thus 1-(2-(3-(4))). I can't seem to find a succinct way to do the same in P6. None of [-], [-] reverse, [R-], or [R-] reverse does what I want. 18:43
I even tried map *+-*, (1,2,3,4), no joy
sorry, not harmonic, I forget the term for a+b-c+d-e+f 18:44
psch that's harmonic iff a > b > c > d > e > f 18:45
(with abs() around every letter)
uruwi m: [-] (1, 2, 3, 4)
camelia ( no output )
n0tjack yeah, but I think there's a specific term for alternating sum and difference, but it escapes me
uruwi m: say [-] (1, 2, 3, 4)
camelia rakudo-moar 8a8bf6: OUTPUT«-8␤»
n0tjack that should be -2 18:46
uruwi m: say [R-] (1, 2, 3, 4)
camelia rakudo-moar 8a8bf6: OUTPUT«-2␤»
uruwi m: say [-] reverse (1, 2, 3, 4)
camelia rakudo-moar 8a8bf6: OUTPUT«-2␤»
uruwi What's the problem?
n0tjack huh, I really thought I tried [R-]
PerlJam n0tjack: um ... I think it's just called an alternating series.
ab6tract self contained multi dim ASSIGN-POS question: gist.github.com/ab5tract/bb20ff79dddf965febe4
AlexDaniel what does that R mean? 18:47
uruwi Reverse I suppose
PerlJam AlexDaniel: "reverse the operands"
[Coke] reverse.
AlexDaniel hm, are there any other features like R?
uruwi ab6tract, is that example from Terminal::Print? 18:48
apparently my first instinct in typing Terminal::Print is to spell Terminal with 2 L's
n0tjack m: say [R-] 1 2 3 4 5 6;
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/lN_PwATU75␤Two terms in a row␤at /tmp/lN_PwATU75:1␤------> 3say [R-] 17⏏5 2 3 4 5 6;␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ …»
PerlJam AlexDaniel: they're called meta-ops and there are several
n0tjack m: say [R-] (1,2, 3, 4, 5, 6);
camelia rakudo-moar 8a8bf6: OUTPUT«-9␤»
psch m: my @s = -1/1, 1/1, -(1/(1 + (*.denominator))) ... *; say @s[^10]
camelia rakudo-moar 8a8bf6: OUTPUT«(-1 1 -0.5 -0.333333 -0.25 -0.2 -0.166667 -0.142857 -0.125 -0.111111)␤»
n0tjack m: say (1 - 2 + 3 - 4 + 5 - 6);
camelia rakudo-moar 8a8bf6: OUTPUT«-3␤»
psch m: my @s = -1/1, 1/1, -(1/(1 + (*.denominator))) ... *; .Rat.say for @s[^10] 18:49
camelia rakudo-moar 8a8bf6: OUTPUT«-1␤1␤-0.5␤-0.333333␤-0.25␤-0.2␤-0.166667␤-0.142857␤-0.125␤-0.111111␤»
psch m: my @s = -1/1, 1/1, -(1/(1 + (*.denominator))) ... *; .perl.say for @s[^10] # still not pretty :/ 18:50
camelia rakudo-moar 8a8bf6: OUTPUT«-1.0␤1.0␤-0.5␤<-1/3>␤-0.25␤-0.2␤<-1/6>␤<-1/7>␤-0.125␤<-1/9>␤»
PerlJam AlexDaniel: See S03:4127
synbot6 Link: design.perl6.org/S03.html#line_4127
psch eh, and sign-confusion
n0tjack uruwi: ^^ that's the problem. I need the rightmost op to always be - . With [R-] and even-parity series, it's + .
m: say [R-] (1,2, 3, 4, 5, 6,0); 18:51
camelia rakudo-moar 8a8bf6: OUTPUT«-21␤»
uruwi Simply create a list of signs
AlexDaniel PerlJam:
oops
n0tjack uruwi: at that point, I'd be better off just writing a loop 18:52
uruwi: I was hoping for somethign succinct, like [-] or [R-] or even [R-] reverse, etc.
AlexDaniel PerlJam: now I see that I'm actually familiar with the rest, it's just that I've never seen R
n0tjack AlexDaniel: There's others, like Z. For example, (100,10,1) Z* (1,2,3)
AlexDaniel n0tjack: yeah :) 18:53
n0tjack oops, sorry, see you said that already :)
PerlJam AlexDaniel: weird ... we're always doing stuff with X, Z, and R. S is usually the one that surprises people
18:53 Hor|zon left
AlexDaniel PerlJam: Well, I'm not sure if there are many times you might want to use S in user code 18:54
I might be wrong though
PerlJam: yet R seems to be very useful
ab6tract uruwi: it's from my async branch, ues 18:55
*yes
uruwi Terminal::Print just deals with printing, doesn't it? 18:56
AlexDaniel and of course, I just could not resist to do this
ab6tract if this works out right, the only functional difference in the API will be that you use [ ; ] instead of [ ][ ]
AlexDaniel m: say [RRRRRRRRRRRRRRRRRRRRRRR-] [1, 2, 3, 4];
camelia rakudo-moar 8a8bf6: OUTPUT«-2␤»
AlexDaniel rrr!
ab6tract it should be *much* faster
especially the creation of new grids
uruwi And the new syntax for grid access is much cleaner too!
But what about fullwidth characters? 18:57
Wait, you're just dealing with character-by-character output.
18:57 ^elyse^ joined
n0tjack m: infix:</>(1,*)>>(1,2,3); # How do I do reciprocal? 18:58
camelia rakudo-moar 8a8bf6: OUTPUT«Cannot call METAOP_HYPER_POSTFIX(Num, Int, Int, Int, Int); none of these signatures match:␤ (\obj, \op)␤ (\obj, @args, \op)␤ (\obj, \args, \op)␤ in block <unit> at /tmp/NkKwdBIc4W:1␤␤»
18:59 apotheon left
uruwi m: say 1/>>(1, 2, 3); 18:59
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/sdx1CueAwu␤Missing infix inside HYPER␤at /tmp/sdx1CueAwu:1␤------> 3say 1/>>7⏏5(1, 2, 3);␤ expecting any of:␤ infix␤ infix stopper␤»
uruwi m: say 1>>/>>(1, 2, 3); 19:00
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/qrms93_00O␤Malformed postfix␤at /tmp/qrms93_00O:1␤------> 3say 1>>7⏏5/>>(1, 2, 3);␤ expecting any of:␤ postfix␤»
uruwi m: say 1 >>/>> (1, 2, 3);
camelia rakudo-moar 8a8bf6: OUTPUT«Lists on either side of non-dwimmy hyperop of infix:</> are not of the same length␤left: 1 elements, right: 3 elements␤ in block <unit> at /tmp/g1Xw82_NnM:1␤␤»
uruwi m: say 1 <</<< (1, 2, 3);
camelia rakudo-moar 8a8bf6: OUTPUT«(1 0.5 0.333333)␤»
psch m: sub postfix:['⁻'] { 1 / $^a }; say 5⁻
camelia rakudo-moar 8a8bf6: OUTPUT«0.2␤»
uruwi ta-da
19:00 ab6tract left
PerlJam you can remember which of >> or << to use by always putting the fat end to the listy thing and the pointy end to the single-valued thing. 19:01
19:01 skids joined
AlexDaniel m: [SR-] 19:02
camelia rakudo-moar 8a8bf6: OUTPUT«===SORRY!===␤Unknown QAST node type NQPMu␤»
uruwi Or spear one side to make it DWIM.
19:02 PsuzeannePsy joined
FROGGS_ AlexDaniel: that one is rakudobuggable 19:02
AlexDaniel FROGGS_: as LTA, right? 19:03
FROGGS_ it leeks internal stuff without telling what's wrong, yes
psch m: say 5 S- 5
camelia rakudo-moar 8a8bf6: OUTPUT«Cannot find method 'CALL-ME'␤ in block <unit> at /tmp/0v5aaysYOc:1␤␤»
PerlJam I don't think S is fully implemented either
psch is meta S even implemented?
... :)
PerlJam m: [S-] 19:04
camelia rakudo-moar 8a8bf6: OUTPUT«Method 'count' not found for invocant of class 'Int'␤ in block <unit> at /tmp/iOpHdjInm7:1␤␤»
n0tjack uruwi: ah, never thought of <</<<
PerlJam: That's a helpful tip. Thank you. 19:05
19:05 SuzieQueue left 19:06 apotheon joined, skids left 19:07 skids joined
AlexDaniel oh great, just misspelled "leaks" 19:07
psch: well, if not, then perhaps it should show NYI warning? 19:08
psch AlexDaniel: i'm not saying having an RT is a bad idea 19:09
AlexDaniel psch: :)
19:17 diana_olhovik_ joined
kmel excluding rendering it to html using pod::to::html. How can one view the pod files? 19:17
19:18 mattprelude joined
[Coke] theoretically, perl6/doc's p6doc (though it's dying for me here trying to use it) 19:19
perl6 --doc path/to/file.pm works
kmel i can't install p6doc 19:21
[Coke] thanks perl6 --doc works
TimToady the easiest way to remember which way to put the >> is to use X instead 19:22
then you can drop the parens too 19:23
m: say 1 X/ 1,2,3
camelia rakudo-moar 8a8bf6: OUTPUT«(1 0.5 0.333333)␤»
19:23 ab6tract joined, diana_olhovik_ left
TimToady but for some reason people never think of the degenerate X case with a scalar on one side 19:24
19:24 apotheon left
vendethiel m: say 1 Z/ 1, 2, 3 19:25
camelia rakudo-moar 8a8bf6: OUTPUT«(1)␤»
vendethiel hah, yeah
ab6tract uruwi: we can support wide character's as well. not sure what the best way will be but it shouldn't be too hard to wire up
TimToady ab6tract: when I was playing with forest fire, I found [$i;$j] is considerable slower than [$i][$j] so far
presumably it'll get optimized though
19:26 edehont left
uruwi ab6tract is there any reason you have to use [$i;$j]? 19:26
ab6tract re: "just about printing", it's designed to allow the construction of higher level 'text ui' libraries. so it should also be able to handle input in the way you want as well
[Coke] TimToady: I use the degenerate X all the time when trying to solve ken ken puzzles! :)
ab6tract TimToady: interesting.. in my case I would construct Column objects which would construct Cell objects.. it was all a bit over designed 19:27
creating a new grid object could take a second or more
uruwi: well, we could still create columns as individual classes, I guess
it's also that I want to document the multi dim ASSIGN-POS as well :) 19:28
and to shake it down for robustness in ecosystem code
19:28 jkva joined
ab6tract uruwi: the [ $x][$y] form existed because grid had *-POS to columns which had *-POS to cells. do you prefer the [][] syntax? I guess we could keep it without too much trouble whatsoever 19:30
19:30 weihan joined
uruwi Well, TimToady says that the [;] syntax is slower. 19:30
ab6tract to that end, are you finding the library useful in your rogue quest?
uruwi I haven't tried it yet
ab6tract fair enough :)
it will definitely take care of the 'keys echo characters' case 19:31
19:31 rindolf left
FROGGS_ jnthn++ # blog 19:31
19:32 jkva left 19:33 skids left
ab6tract but there is currently no API for registering input key stokes. that is a relatively complex story in it's own right, but I think it could be added pretty easily with supplies. 19:34
uruwi I think I solved the problem github.com/bluebear94/messing-arou...er/kbd.pl6 19:35
n0tjack m: say 1,-2,3,-4 ... 10; # :( 19:36
camelia rakudo-moar 8a8bf6: OUTPUT«===SORRY!===␤Unable to deduce arithmetic or geometric sequence from -2,3,-4 (or did you really mean '..'?)␤»
19:36 ^elyse^ left
[Coke] m: (1..10).map({state $a = -1; $a *= -1; $a*$_}).say 19:40
camelia rakudo-moar 8a8bf6: OUTPUT«(1 -2 3 -4 5 -6 7 -8 9 -10)␤»
[Coke] seems like we should be able to golf that to a $ instead of a state $a 19:41
n0tjack m: map {$_ * (-1) ** (+$_%2)}, (1..10);
camelia ( no output )
n0tjack m: say map {$_ * (-1) ** (+$_%2)}, (1..10);
camelia rakudo-moar 8a8bf6: OUTPUT«(-1 2 -3 4 -5 6 -7 8 -9 10)␤»
psch %% 19:42
m: say map {$_ * (-1) ** (+$_%%2)}, (1..10);
camelia rakudo-moar 8a8bf6: OUTPUT«(1 -2 3 -4 5 -6 7 -8 9 -10)␤»
n0tjack I didn't need that +
[Coke] %% works, but only on the original sequence.
psch well, or not, depending on whether you want -1 or not
n0tjack I think of $_%2 as a Bool, but obviously it isn't
psch: For mod 2, %% is overkill 19:43
[Coke] m: (1..10).map({state$a=-1;$a*=-1;$_*$a}).say # look at all the whitespace I can skip.
camelia rakudo-moar 8a8bf6: OUTPUT«(1 -2 3 -4 5 -6 7 -8 9 -10)␤»
n0tjack psch: It would be helpful for mod 3+
psch shrugs 19:44
n0tjack m: say map{$_*(-1)**(+$_%2)},(1..10);
19:44 llfourn joined
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/amSq_F4cwy␤Calling map() will never work with any of these multi signatures:␤ (&code, + is raw)␤at /tmp/amSq_F4cwy:1␤------> 3say 7⏏5map{$_*(-1)**(+$_%2)},(1..10);␤» 19:44
n0tjack heh
TimToady m: say (1,-1,1) Z* 1..*
camelia rakudo-moar 8a8bf6: OUTPUT«(1 -2 3)␤»
TimToady m: say (1,-1,1...* Z* 1..*
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/nnvkUzdYA4␤Operators '...' and 'Z*' are non-associative and require parentheses␤at /tmp/nnvkUzdYA4:1␤------> 3say (1,-1,1...* Z7⏏5* 1..*␤ expecting any of:␤ infix␤ infix stopp…»
TimToady m: say (1,-1,1...*) Z* 1..*
camelia rakudo-moar 8a8bf6: OUTPUT«(...)␤»
TimToady m: .say for (1,-1,1...*) Z* 1..* 19:45
n0tjack don't you want 1 xx * ?
camelia rakudo-moar 8a8bf6: OUTPUT«(timeout)1␤-2␤3␤-4␤5␤-6␤7␤-8␤9␤-10␤11␤-12␤13␤-14␤15␤-16␤17␤-18␤19␤-20␤21␤-22␤23␤-24␤25␤-26␤27␤-28␤29␤-30␤31␤-32␤33␤-34␤35␤-36␤37␤-38␤39␤-40␤41␤-42␤43␤-44␤45␤-46␤47…»
TimToady Z* 1 xx * is a no-op 19:46
n0tjack I would have thought (blah) Z* 1 xx *; would turn the whatever into "as many elems as (blah)"
TimToady it does, but multiplying everything by 1 does nothing 19:47
19:47 skids joined
TimToady m: .say for |(1,-1) xx * Z* 1..20 19:48
camelia rakudo-moar 8a8bf6: OUTPUT«1␤-2␤3␤-4␤5␤-6␤7␤-8␤9␤-10␤11␤-12␤13␤-14␤15␤-16␤17␤-18␤19␤-20␤»
n0tjack oh, I see, 1..* isn't 1,1,1,1,1, it's 1,2,3,4
don't know why I thought otherwise
19:48 llfourn left
n0tjack m: say 4 * [+] map *+-*, (1 X/ (1,3,5,7 ... 25e3)); 19:50
camelia rakudo-moar 8a8bf6: OUTPUT«3.14151265358992␤»
19:51 Hor|zon joined
n0tjack m: say 4 * [+] (1,-1,1 ... *) Z* (1 X/ (1,3,5,7 ... 25e3)); 19:51
camelia rakudo-moar 8a8bf6: OUTPUT«3.1415126535899␤»
n0tjack Whatever code is cute.
TimToady that one doesn't converge very fast...
n0tjack no, but it's short 19:52
TimToady testing against a floater is kind of a steep price to pay for saving one char over 25000 19:53
19:54 zakharyas joined
n0tjack where is the test? 19:54
TimToady 1,3,5,7 ... 25e3 19:55
ab6tract uruwi: nice!
n0tjack there's a test in there?
FROGGS_ the endpoint
TimToady how do you suppose it stops?
ab6tract I think you should wrap that in a module and ship it :)
19:55 telex left, Hor|zon left
uruwi ab6stact I'm mainly waiting for Terminal::Print to be documented, though 19:55
n0tjack TimToady: fair point, hadn't occured to me.
TimToady: Though I suppose if you impute the function, you could try to invert it. 19:56
TimToady there's another function hiding in there as well, since it intuits a *+2 for you
19:56 telex joined, Peter_R joined
ab6tract fair enough. 19:57
TimToady n0tjack: this is the first time that's come up, so I think the demand is too low to make it a DWI
*M
19:57 zakharyas left
n0tjack I wouldn't implement function inversion just to make ... more efficient. 19:57
But it's really useful in functional programming for broader purposes
ab6tract how to pass 'self' to the constructor of another another object? 19:58
19:59 zakharyas joined, kmel left 20:00 tokuhiro_ joined
ab6tract uruwi: your work could easily go into another module entirely.. Terminal::Input, perhaps? 20:01
TimToady nap & 20:02
uruwi I'm kind of reluctant since it works only if stty is installed
20:02 dbron joined, dbron left
uruwi Also a lot of things are hardcoded in the demo 20:02
20:03 n0tjack left, n0tjack_ joined, ^elyse^ joined, margeas joined
n0tjack_ how are identity functions declared? 20:03
for example, p6 knows +'s identity is 0, and *'s is 1. 20:04
can I declare my own identity value for a function?
grondilu m: say 4 * [+] (1,-1,1 ... *) Z/ (1,3,5,7 ... 25e3);
camelia rakudo-moar 8a8bf6: OUTPUT«3.1415126535899␤»
n0tjack_ grondilu: even better!
20:04 tokuhiro_ left
ab6tract \o/ got it! 20:04
grondilu m: say 4 * [+] (1,-1,1 ... *) Z/ (1,3 ... 25e3);
camelia rakudo-moar 8a8bf6: OUTPUT«3.1415126535899␤» 20:05
grondilu I'm a bit surprised this stops, considering 25e3 is even
psch n0tjack_: for infix:<+> there's "multi sub infix:<+>($x = 0) { $x.Numeric }" as one candidate
grondilu m: say (1,3 ... 25e3).elems
camelia rakudo-moar 8a8bf6: OUTPUT«12500␤»
grondilu m: say (1,3 ... 25e3)[*-5] 20:06
camelia rakudo-moar 8a8bf6: OUTPUT«24991␤»
psch m: say 1, 2, 4 ... 9
camelia rakudo-moar 8a8bf6: OUTPUT«(1 2 4 8)␤»
grondilu m: say (1,3 ... 25e3)[*-5..*]
camelia rakudo-moar 8a8bf6: OUTPUT«(24991 24993 24995 24997 24999)␤»
grondilu m: say (1,3...10)
camelia rakudo-moar 8a8bf6: OUTPUT«(1 3 5 7 9)␤»
n0tjack_ psch: I see, so reduction doesn't apply any fancy logic, each infix can just return the non-identity argument 20:07
FROGGS_ aye
n0tjack_ psch: the reason I was asking was I thought there might be a "has identity" or something trait; I was hoping to extent that idea to "has inverse"
ab6tract uruwi: what do you use stty for? 20:08
uruwi to disable stdin buffering, so keystrokes can go through without pressing Enter 20:09
20:09 kmel joined, necrophcodr joined
n0tjack_ what I would like to do is say something like multi sub:infix<log> {$x, $y} has inverse(&infix<exp>) {...} . 20:09
20:10 FROGGS_ left
necrophcodr is rakudo p5 build.pm broken, or is something missing with the 2015.09 rakudo star distribution? 20:10
20:10 n0tjack_ is now known as n0tjack
psch n0tjack: not sure how you'd use that 20:11
[Coke] necrophcodr: as far as I know the 2015.09 star distro was fine.
necrophcodr after installing rakudo star 2015.09 and running `panda install v5` it only gets to "Generating build order", and then doesn't go any further. i've had it running for hours.
psch n0tjack: as in, you'd need some syntax to invert a function first
[Coke] don't use v5.
20:11 FROGGS joined
[Coke] use Inline::Perl5 20:11
necrophcodr [Coke]: how does that help running v5 scripts?
[Coke] well, I should phrase that: "are you sure you want v5 and not Inline::Perl5?" 20:12
20:12 darutoko left
n0tjack psch: I'd want to use it to write a meta-op, call it "under", something like sub under(&foo, *@) { &foo.inverse foo @* }; 20:12
necrophcodr i am attempting to use perl6 as a perl5 drop in replacement (i am well aware that it isn't)
lizmat necrophcodr: it is my understanding that v5 is a project in hibernation
n0tjack what that would let me do is express things like (silly example) + under log, which is of course *
[Coke] I don't see any commits on v5 that aren't cleanup since maybe january. 20:13
So, it could just be that v5's build is borked. Pretty sure the panda that shipped with the last star was ok.
necrophcodr i haven't had any bad experience with the star 2015.09 distribution either, i just wanted to make sure. 20:14
lizmat: that could be it. thanks for the help, it is very valuable!
[Coke] I just tried to run the 00 sanity test in v5, and it dies. 20:15
20:15 ab6tract left
[Coke] ah. they used src, not lib. 20:15
m: say EnumMap
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/9gOKbQcUrZ␤Undeclared name:␤ EnumMap used at line 1␤␤»
necrophcodr i'll have to stick to using only perl5 for now then. i hope that in the future a proper v5 will be out there, in a couple of years. would make my maintainence a lot easier.
[Coke] ^^ v5 still has that.
there is no current plan for making v5 work, at least on this channel, that I'm aware of. 20:16
I think everyone would love it if it -did- work, mind.
necrophcodr yeah, moving entire codebases isn't always feasible.
n0tjack necrophcodr: I honestly think of p5 and p6 as separate languages. 20:17
[Coke] necrophcodr: right. the plan isn't to have people migrate from 5 to 6.
it was a plan, probably over a decade ago, but that's not the current plan.
psch m: my &infix:<+> = CORE::('&infix:<+>') but role { method inverse { &infix:<-> } }; sub under(&f where &f.inverse.defined, *@a) { &f.inverse( f( @a ) ) }; 20:18
camelia ( no output )
[Coke] I opened github.com/rakudo-p5/v5/issues/12 to mention the EnumMap issue.
n0tjack psch: !!!!
psch n0tjack: not sure if that does anything useful though...
n0tjack even that it can be approached is awesome
necrophcodr [Coke]: i think it may still happen though, through time. although there are certain valuable uses even today, for perl5. with that, i mean mostly the amazing compatability with older perl5 installations. 20:19
not implying that modern perl5 isn't in itself useful. it is indeed.
20:19 Suzeanne joined
psch n0tjack: it's probably kinda expensive with the role mixin, and i think the parameters of under might be wrong-ish too 20:19
20:20 lucastheisen joined
psch n0tjack: like, if it should work as "+ under log @a" (or, without macros more like "&infix:<+> under &log, @a") you clearly need two callable parameters 20:20
n0tjack psch: To really support it, it'd have to be implemented in the core. Inversion has its merits, but I'm not going to try to argue it should be a near-term priority.
20:20 ^elyse^ left 20:21 lucastheisen left
psch n0tjack: i don't know, i think it could work well enough in a module, but probably wants macro support to give closer-to-meta-op syntax 20:21
20:23 PsuzeannePsy left
n0tjack psch: yeah, the more I play with the lang, the more excited I get for macros 20:23
20:23 apotheon joined 20:26 colomon left
lizmat m: sub a($a) {}; a IterationEnd #hmmmm 20:26
20:26 colomon joined
camelia rakudo-moar 8a8bf6: OUTPUT«X::TypeCheck::Binding exception produced no message␤ in sub a at /tmp/wW0MmuULW7:1␤ in block <unit> at /tmp/wW0MmuULW7:1␤␤» 20:26
lizmat hmmm... this hangs for me... 20:27
locally
20:28 colomon left 20:29 colomon joined, necrophcodr left
Hotkeys is there a reason doc.perl6.org/language/regexes doesn't talk about lazy matching (eg. /.+/ vs /.+?/) 20:30
or does it and I just don't see it 20:31
grondilu not sure "lazy matching" is the correct expression
Hotkeys I don't know the right term
grondilu rather "eager matching" as opposed to "minimal matching" 20:32
· To force the preceding atom to do frugal backtracking (also sometimes known as "eager matching" or "minimal matching"), append a ":?" or "?" to the atom. If the preceding token is a quantifier, the ":" may be omitted, so "*?" works just as in Perl 5.
Hotkeys ah
grondilu ^in S05
Hotkeys okay thanks 20:33
I've seen it talked about as greedy and lazy before
n0tjack Hotkeys: I have too.
grondilu r u sure you don't confuse with list evaluation?
20:34 ab6tract joined
FROGGS it is usually called greedy and non-greedy matching 20:34
20:34 Peter_R left, zakharyas left
FROGGS lazy is about (potentially infinite) lists 20:34
n0tjack just tried [max] 1..*; and got exactly what he deserved
FROGGS :P
Hotkeys lol
grondilu m: say [+] 1 .. 10**100; 20:35
Hotkeys do you need the square brackets there? (assuming a not-infinite list)
m: say max 1 .. 10
camelia rakudo-moar 8a8bf6: OUTPUT«(timeout)»
rakudo-moar 8a8bf6: OUTPUT«10␤»
psch m: say + 1..10
camelia rakudo-moar 8a8bf6: OUTPUT«1..10␤»
psch m: say [+] 1..10 20:36
camelia rakudo-moar 8a8bf6: OUTPUT«55␤»
Hotkeys you do for +
n0tjack you do for any dyad
s/dyad/infix/
Hotkeys right
bc [] does reduction
psch m: say &infix:<max>
camelia rakudo-moar 8a8bf6: OUTPUT«sub infix:<max> (| is raw) { #`(Sub+{<anon|68770224>}+{Precedence}|62511920) ... }␤»
Hotkeys but max doesn't need reduction
psch m: say &max
camelia rakudo-moar 8a8bf6: OUTPUT«sub max (+ is raw, :&by = { ... }) { #`(Sub|72590472) ... }␤»
psch Hotkeys: max is just also a sub that takes a list
grondilu m: sub prefix:<[+]>(Range $r) { $r.min*($r.max+1)/2 }; say [+] ^10;
camelia rakudo-moar 8a8bf6: OUTPUT«0␤»
Hotkeys ah 20:37
does it secretly do [max] behind the scenes
n0tjack I find it weird we have &infix:<max> and &prefix:<max> but no &infix:<log> etc.
psch m: say &prefix:<max> 20:38
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/9eihiU5CHd␤Undeclared routine:␤ &prefix:<max> used at line 1. Did you mean '&prefix:<~>', '&prefix:<temp>', '&prefix:<->', '&prefix:<so>', '&prefix:<+>'?␤␤»
FROGGS m: say &max
camelia rakudo-moar 8a8bf6: OUTPUT«sub max (+ is raw, :&by = { ... }) { #`(Sub|66163848) ... }␤»
n0tjack m: say max 10, 100; say 10 max 100; say log 100, 10; say 10 log 100;
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/cVyOaKX5PN␤Two terms in a row␤at /tmp/cVyOaKX5PN:1␤------> 3 say 10 max 100; say log 100, 10; say 107⏏5 log 100;␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ …»
n0tjack well, you get what I mean.
FROGGS m: say &infix:<max> 20:39
camelia rakudo-moar 8a8bf6: OUTPUT«sub infix:<max> (| is raw) { #`(Sub+{<anon|68770224>}+{Precedence}|53111600) ... }␤»
FROGGS so you're missing the infix log
n0tjack yeah
FROGGS is that how you use log often?
n0tjack in general, all fundamental 2-adic math primitives should be offered as prefix and infix
I write most of my functional code linearly 20:40
FROGGS hmmm, I've got no opinion on &infix:<log> 20:41
n0tjack so I'd want to say [max] 2 log 3,7,22,108 to find the bit-widest number
(as a contrived example)
20:42 kfo left
n0tjack the problem with a bunch of prefix function is you (end (up, with, (lispitis(of (the, code))))))))))))))))))) 20:42
or you have to break simple things into multiple lines. 20:43
FROGGS true
n0tjack I suppose there could be some kinda hyper, say F, which turned $a F log $b into log $a,$b 20:44
all the birds, with one stone
psch m: say [max] 2 [&log] 3, 7, 22, 108 20:46
camelia rakudo-moar 8a8bf6: OUTPUT«108␤»
psch :P
n0tjack m: say 2 [&log] 0b1, 0b10, 0b11, 0b100, 0b111, 0b1000 20:47
camelia rakudo-moar 8a8bf6: OUTPUT«Attempt to divide 0.693147180559945 by zero using /␤ in block <unit> at /tmp/Wc_6YRYmBv:1␤␤Actually thrown at:␤ in block <unit> at /tmp/Wc_6YRYmBv:1␤␤»
n0tjack m: say 2 [&log] (0b1, 0b10, 0b11, 0b100, 0b111, 0b1000)
camelia rakudo-moar 8a8bf6: OUTPUT«0.386852807234542␤»
n0tjack yeah, no
Hotkeys m: [max] map *.log(2), 3,7,22,108 20:48
camelia ( no output )
Hotkeys wow
worked for me in my shell
n0tjack say "say"
Hotkeys oh right
m: say [max] map *.log(2), 3,7,22,108
camelia rakudo-moar 8a8bf6: OUTPUT«6.75488750216347␤»
n0tjack yeah, whatever-code is sick 20:49
Hotkeys tru
psch m: say log 2, 0b1
camelia rakudo-moar 8a8bf6: OUTPUT«Attempt to divide 0.693147180559945 by zero using /␤ in block <unit> at /tmp/k0UAGQiTpO:1␤␤Actually thrown at:␤ in block <unit> at /tmp/k0UAGQiTpO:1␤␤»
n0tjack m: say log(2,0b1); 20:50
camelia rakudo-moar 8a8bf6: OUTPUT«Attempt to divide 0.693147180559945 by zero using /␤ in block <unit> at /tmp/Qmw1ACetbB:1␤␤Actually thrown at:␤ in block <unit> at /tmp/Qmw1ACetbB:1␤␤»
n0tjack m: say 0b1;
camelia rakudo-moar 8a8bf6: OUTPUT«1␤»
psch m: say log 2, 1
camelia rakudo-moar 8a8bf6: OUTPUT«Attempt to divide 0.693147180559945 by zero using /␤ in block <unit> at /tmp/RGM64g4f9f:1␤␤Actually thrown at:␤ in block <unit> at /tmp/RGM64g4f9f:1␤␤»
Hotkeys second number is base
n0tjack wait, wait
m: say log 1,2;
camelia rakudo-moar 8a8bf6: OUTPUT«0␤»
n0tjack m: say log 100, 10;
camelia rakudo-moar 8a8bf6: OUTPUT«2␤»
n0tjack the radix comes 2nd 20:51
psch m: say max (0b1, 0b10, 0b101, 0b1100) >>[&log]>> 2
camelia rakudo-moar 8a8bf6: OUTPUT«3.58496250072116␤»
psch shrugs
not that i really understand what that snippet was about
n0tjack also, log *,1 should return Inf
20:51 sufrostico joined
psch m: say log Inf, 1 20:51
camelia rakudo-moar 8a8bf6: OUTPUT«Attempt to divide Inf by zero using /␤ in block <unit> at /tmp/Nqer4p1xW7:1␤␤Actually thrown at:␤ in block <unit> at /tmp/Nqer4p1xW7:1␤␤»
psch m: say log *, 1
camelia rakudo-moar 8a8bf6: OUTPUT«Cannot call log(Whatever, Int); none of these signatures match:␤ (Numeric $x)␤ (Numeric $x, Numeric $base)␤ (Cool $x)␤ (Cool $x, Cool $base)␤ (num $x)␤ in block <unit> at /tmp/gjWLfSxl3u:1␤␤»
n0tjack or Nan
20:52 margeas left
n0tjack it shouldn't just croak 20:52
20:52 Hor|zon joined
psch oh, the * wasn't a Whatever but a glob 20:52
(or maybe a whatever, not a Whatever..?)
Hotkeys it should be a divide by zero error
just like
m: say 3/0
camelia rakudo-moar 8a8bf6: OUTPUT«Attempt to divide 3 by zero using div␤ in block <unit> at /tmp/lI0MtYM7I1:1␤␤Actually thrown at:␤ in block <unit> at /tmp/lI0MtYM7I1:1␤␤»
Hotkeys or NaN
NaN works
n0tjack I think 3/0 should be NaN too 20:53
or at least a pragma to switch "loud failures" vs "NaN failures"
the ugly part about NaN is it's a data virus. Once it gets in, it spreads.
20:53 margeas joined
FROGGS .oO( $data.grep: NaN ) 20:54
n0tjack m: grep {$_ !== NaN}, (1,NaN,2,NaN,3,4); 20:55
camelia ( no output )
n0tjack m: say grep {$_ !== NaN}, (1,NaN,2,NaN,3,4);
camelia rakudo-moar 8a8bf6: OUTPUT«(1 NaN 2 NaN 3 4)␤»
FROGGS ahh, nvm, NaN smartmatches itself :o)
n0tjack NaN == NaN :)
20:56 Hor|zon left
vendethiel it bugs me that it bugs me that NaN equals NaN... :P 20:56
FROGGS m: say NaN == NaN 20:57
camelia rakudo-moar 8a8bf6: OUTPUT«False␤»
psch m: say NaN === NaN
camelia rakudo-moar 8a8bf6: OUTPUT«True␤»
n0tjack FROGGS: I meant !==
it's an important feature, like the smell they add to propane
FROGGS *g* 20:58
n0tjack you *want* to be forced to remember and deal with NaNs
psch m: say NaN eq NaN
camelia rakudo-moar 8a8bf6: OUTPUT«True␤»
psch m: say grep {$_ ne NaN}, (1,NaN,2,NaN,3,4);
camelia rakudo-moar 8a8bf6: OUTPUT«(1 2 3 4)␤»
psch ship it
Hotkeys lol
n0tjack if they worked "reasonably" sometimes (e.g. ==), then you'd forget about them, until they failed catastrophically
21:00 sufrosti1o joined
FROGGS m: say Inf/Inf x 10, <Batman> 21:01
camelia rakudo-moar 8a8bf6: OUTPUT«NaNNaNNaNNaNNaNNaNNaNNaNNaNNaNBatman␤»
FROGGS gnight
n0tjack hahaha.
lizmat gnight FROGGS
n0tjack g'night FROGGS
[Coke] ww2.kqed.org/mindshift/2013/10/08/c...g-to-code/ 21:03
Hotkeys m: say sqrt(-1)
camelia rakudo-moar 8a8bf6: OUTPUT«NaN␤»
Hotkeys m: say sqrt(-1.Complex)
camelia rakudo-moar 8a8bf6: OUTPUT«6.12323399573677e-17-1i␤»
Hotkeys that's a fun number
n0tjack m: e**pi*i; 21:04
camelia rakudo-moar 8a8bf6: OUTPUT«WARNINGS:␤Useless use of "*" in expression "**pi*i" in sink context (line 1)␤»
n0tjack m: say e** pi*i ;
camelia rakudo-moar 8a8bf6: OUTPUT«0+23.1406926327793i␤»
Hotkeys should it be 0+1i ?
or am i bad at the notation
[Coke] n0tjack: m: say e**(pi*i)
n0tjack Hotkeys: it is 0+1i, notice the e-17
Hotkeys I see that
[Coke] m: say e**(pi*i)
camelia rakudo-moar 8a8bf6: OUTPUT«-1+1.22464679914735e-16i␤»
Hotkeys I thought perl 6 was magical at rounding
[Coke] precedence. 21:05
21:05 ollej joined
[Coke] we use rationals until we can't. the use of pi makes it numeric, not rational. 21:05
n0tjack [Coke]: Yeah, specifically the non-precedence of ** always gets me
Hotkeys ah
n0tjack Is there some way to ask for the numerator/denominator expression of a Rat?
Hotkeys m: say .3.perl 21:06
camelia rakudo-moar 8a8bf6: OUTPUT«0.3␤»
n0tjack m: Rat.new(22,7); # I wanna see 22/7.
camelia ( no output )
Hotkeys er
[Coke] m: say pi.Rat(1e-10).nudes;
camelia rakudo-moar 8a8bf6: OUTPUT«Method 'nudes' not found for invocant of class 'Rat'␤ in block <unit> at /tmp/8O2U0GHPGQ:1␤␤»
[Coke] m: say pi.Rat(1e-10).nude;
n0tjack m: say Rat.new(22,7); # I wanna see 22/7.
camelia rakudo-moar 8a8bf6: OUTPUT«(312689 99532)␤»
rakudo-moar 8a8bf6: OUTPUT«3.142857␤»
Hotkeys right
that
n0tjack hahah nude
[Coke] *nu*merator*de*nominator
n0tjack of course.
psch that's why the s is funny :)
n0tjack m: say Pi.Rat(1e-10).scantily-clad; 21:07
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/KlslqObmUh␤Undeclared name:␤ Pi used at line 1␤␤»
Hotkeys lol
n0tjack boy, no middle ground.
psch m: (1/2).perl.say; (22/7).perl.say # this confuses me, tbh 21:09
camelia rakudo-moar 8a8bf6: OUTPUT«0.5␤<22/7>␤»
[Coke] accidentally killed his full copy of perl6-doc html that he built as a reference and has to rebuild the whole thing . zzzz.
psch why does 1/2 turn into decimal notation but 22/7 doesn't? 21:10
n0tjack 0.5 is pretty clean
[Coke] psch: yah, that's not going to be fun to explain.
n0tjack 22/7 is 3.14286
uruwi 0.5 terminates but 22/7 doesn't
[Coke] m: (1/4).perl.say
camelia rakudo-moar 8a8bf6: OUTPUT«0.25␤»
uruwi m: say pi.Rat(1e-10).scantily-clad;
[Coke] m: (1234/5678).perl.say
camelia rakudo-moar 8a8bf6: OUTPUT«Method 'scantily-clad' not found for invocant of class 'Rat'␤ in block <unit> at /tmp/sxbhv4P7dL:1␤␤»
rakudo-moar 8a8bf6: OUTPUT«<617/2839>␤»
n0tjack hahaha 21:11
kmel m: say (4/2).perl.say;
camelia rakudo-moar 8a8bf6: OUTPUT«2.0␤True␤»
n0tjack I think we need an excuse to add a .mole function, so we could have Rat.nude.mole # naked mole rats
m: (1/3).perl.say; 21:12
camelia rakudo-moar 8a8bf6: OUTPUT«<1/3>␤»
Hotkeys m: my @test = 1/2, 1/4, 1/8 ... 0; say [+] @test;
camelia rakudo-moar 8a8bf6: OUTPUT«1␤»
n0tjack now that's cute.
Hotkeys it's actually a finite series
v sad
[Coke] m: my @test = 1/2, 1/4, 1/8 ... 0; my $a = [+] @test; say $a.WHAT; 21:13
camelia rakudo-moar 8a8bf6: OUTPUT«(Num)␤»
psch m: say (1..* Z/ 2..15)>>.perl
camelia rakudo-moar 8a8bf6: OUTPUT«(0.5 <2/3> 0.75 0.8 <5/6> <6/7> 0.875 <8/9> 0.9 <10/11> <11/12> <12/13> <13/14> <14/15>)␤»
psch vOv
not quite what i was going for
Hotkeys lol
m: say (1, 1 ...* Z/ 2..15)>>.perl 21:14
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/v2_zonOvB1␤Operators '...' and 'Z/' are non-associative and require parentheses␤at /tmp/v2_zonOvB1:1␤------> 3say (1, 1 ...* Z7⏏5/ 2..15)>>.perl␤ expecting any of:␤ infix␤ inf…»
Hotkeys m: say ((1, 1 ... *) Z/ 2..15)>>.perl
camelia rakudo-moar 8a8bf6: OUTPUT«(0.5 <1/3> 0.25 0.2 <1/6> <1/7> 0.125 <1/9> 0.1 <1/11> <1/12> <1/13> <1/14> <1/15>)␤»
psch m: say (1 X/ 1..15)>>.perl
camelia rakudo-moar 8a8bf6: OUTPUT«(1.0 0.5 <1/3> 0.25 0.2 <1/6> <1/7> 0.125 <1/9> 0.1 <1/11> <1/12> <1/13> <1/14> <1/15>)␤»
psch m: say (1 X/ 1..25)>>.perl 21:15
camelia rakudo-moar 8a8bf6: OUTPUT«(1.0 0.5 <1/3> 0.25 0.2 <1/6> <1/7> 0.125 <1/9> 0.1 <1/11> <1/12> <1/13> <1/14> <1/15> 0.0625 <1/17> <1/18> <1/19> 0.05 <1/21> <1/22> <1/23> <1/24> 0.04)␤»
Hotkeys I guess that works
psch m: say 1/11
camelia rakudo-moar 8a8bf6: OUTPUT«0.090909␤»
psch is it "less than 5 digits after the radix?"
looks like that's the rule, to me at least :P
although 1/11 should probably keep going..?
m: say (1/32).perl 21:17
camelia rakudo-moar 8a8bf6: OUTPUT«0.03125␤»
Hotkeys looks like it stops at 6
psch m: say (1/64).perl
camelia rakudo-moar 8a8bf6: OUTPUT«0.015625␤»
psch m: say (1/128).perl
camelia rakudo-moar 8a8bf6: OUTPUT«0.0078125␤»
psch m: say (1/256).perl
camelia rakudo-moar 8a8bf6: OUTPUT«0.00390625␤»
psch m: say (1/512).perl
camelia rakudo-moar 8a8bf6: OUTPUT«0.001953125␤»
psch m: say (1/1024).perl
camelia rakudo-moar 8a8bf6: OUTPUT«0.0009765625␤»
psch welll...
m: say (1/1022).perl 21:18
camelia rakudo-moar 8a8bf6: OUTPUT«<1/1022>␤»
Hotkeys magic
m: say (1 Z/ (2, 4, 8 ... 100000000)) 21:19
camelia rakudo-moar 8a8bf6: OUTPUT«(0.5)␤»
Hotkeys ...
m: say (1 X/ (2, 4, 8 ... 100000000))
camelia rakudo-moar 8a8bf6: OUTPUT«(0.5 0.25 0.125 0.0625 0.03125 0.015625 0.007813 0.003906 0.001953 0.000977 0.000488 0.000244 0.000122 0.000061 0.000031 0.000015 0.0000076 0.0000038 0.0000019 0.00000095 0.00000048 0.00000024 0.00000012 0.000000060 0.000000030 0.000000015)␤»
psch m: say 1 X/ 5, 10, 15 ... 1000
camelia rakudo-moar 8a8bf6: OUTPUT«5===SORRY!5=== Error while compiling /tmp/H6RDtiFjcw␤Operators 'X/' and '...' are non-associative and require parentheses␤at /tmp/H6RDtiFjcw:1␤------> 3say 1 X/ 5, 10, 157⏏5 ... 1000␤ expecting any of:␤ infix␤ infix s…»
psch m: say 1 X/ (5, 10, 15 ... 1000)
camelia rakudo-moar 8a8bf6: OUTPUT«(0.2 0.1 0.066667 0.05 0.04 0.033333 0.028571 0.025 0.022222 0.02 0.018182 0.016667 0.015385 0.014286 0.013333 0.0125 0.011765 0.011111 0.010526 0.01 0.009524 0.009091 0.008696 0.008333 0.008 0.007692 0.007407 0.007143 0.006897 0.006667 0.006452 0.00625 0.…»
Hotkeys m: say ((1 X/ (2, 4, 8 ... 100000000)).perl)
camelia rakudo-moar 8a8bf6: OUTPUT«(0.5, 0.25, 0.125, 0.0625, 0.03125, 0.015625, 0.0078125, 0.00390625, 0.001953125, 0.0009765625, 0.00048828125, 0.000244140625, 0.0001220703125, 0.00006103515625, 0.000030517578125, 0.0000152587890625, 0.00000762939453125, 0.000003814697265625, 0.0000019073…»
Hotkeys that's neat
psch 2 and 5 are what we internally reduce by
for Rat.perl 21:20
lizmat m: class A { multi method a(Mu $t,:$last!,*%a) { say "last" }; multi method a(Callable:D $t) { say "call" } }; A.a( * %% 2, :last ) # expect to this to say "last"
camelia rakudo-moar 8a8bf6: OUTPUT«call␤»
lizmat am I missing sometihng ???
psch m: say (* %% 2) ~~ Callable 21:21
camelia rakudo-moar 8a8bf6: OUTPUT«True␤»
lizmat I guess the Callable:D is narrower than the Mu, so there's no tie-breaking involved ?
psch lizmat: i think so, yeah
additionally :last fits into the Callable candidate
and i think there's a bug about required nameds not being narrower than implicit nameds 21:22
lizmat yeah, they're done in order
the first one that tie-breaks, gets it
psch ah, the bug i had in mind was about subs 21:23
so probably not quite applicable
21:23 ab6tract left
n0tjack g'night all 21:24
21:24 n0tjack left 21:25 skids left 21:26 vendethiel left
kmel anyone having problems with pod? 21:28
numbered headings are not working: design.perl6.org/S26.html#Numbered_headings 21:30
lizmat kmel: could you provide a gist that describes the problem ? 21:31
kmel lizmat : yup just give me 2 sec 21:33
lizmat what link should i send you? 21:39
sorry i am new to github :) 21:40
just made an account
RabidGravy the gist url
kmel gist.github.com/kmel88/fb1c48b71b7f634e6323
this one ^ ?
RabidGravy yep 21:41
m: gist.github.com/kmel88/fb1c48b71b7f634e6323
camelia ( no output )
21:41 muraiki left
lizmat I guess it's a case of NYI ? 21:41
RabidGravy well it oaeses :)
psch hm, i don't S06:1083 21:42
synbot6 Link: design.perl6.org/S06.html#line_1083
psch +get
kmel what is NYI = Not yet implemented ?
lizmat S99:NYI
PerlJam S99:NYI
synbot6 Link: design.perl6.org/S99.html#NYI
PerlJam heh
lizmat not in there 21:43
kmel: but yeah, Not Yet Implemented
sometimes I wonder whether we shouldn't move S99 to doc
psch m: sub f($, :$!) { "ok" }; say f 1 # still one of my favorite WATs in that context 21:44
camelia rakudo-moar 8a8bf6: OUTPUT«ok␤»
psch a required anonymous named parameter doesn't have to be passed vOv
lizmat yeah, not going to change for xmas... 21:45
psch i'm really not sure what it even means
21:45 hohoho joined
psch but i suspect it somewhat relates to RT #119929 21:45
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=119929
lizmat yeah :-) 21:46
21:46 rurban joined 21:47 llfourn joined 21:50 hohoho left
kmel quit 21:50
21:50 kmel left 21:52 firstdayonthejob joined, apotheon left 21:53 Hor|zon joined, llfourn left 21:57 apotheon joined, Hor|zon left 22:02 tokuhiro_ joined, thaljef joined 22:05 thaljef left 22:06 tokuhiro_ left 22:08 kid51 joined, pmurias left 22:19 Ch0c0late joined, kmel joined, sufrosti1o left 22:23 sufrostico left
uruwi Is it currently possible to sandbox Perl 6 code? 22:25
m: shell("ls"); 22:26
camelia rakudo-moar 8a8bf6: OUTPUT«shell is disallowed in restricted setting␤ in sub restricted at src/RESTRICTED.setting:1␤ in sub shell at src/RESTRICTED.setting:15␤ in block <unit> at /tmp/2eRoTeDwwf:1␤␤»
22:26 kid51 left
dalek kudo/nom: 44a7ffa | lizmat++ | src/core/Any-iterable-methods.pm:
Add :end,:k,:v,:p adverbs to List.first

  :end indicates you want searching from the end of the list (like last-index)
  :k return the ordinal number, instead of the value
  :v return the value (default)
  :p return a pair of ordinal number => value
This effectively makes first-index and last-index obsolete. Perhaps we should deprecate them.		 22:32
lizmat PerlJam: ^^^ 22:33
uruwi: depends on what you want to sandbox, perhaps look at how the RESTRICTED.setting is made :-)
22:34 leont joined
lizmat good niight, #perl6! 22:35
uruwi is there any example? 22:36
RabidGravy toodlepip
lizmat uruwi: look at src/RESTRICTED.setting
uruwi where? 22:37
RabidGravy in the source of rakudo
lizmat github.com/rakudo/rakudo/blob/nom/...ED.setting 22:38
really sleep&
22:38 spider-mario left, hes joined
uruwi And the only way to change it is to compile rakudo from source? 22:39
RabidGravy as it currently stands yes 22:41
22:44 samb1 left 22:49 rurban left 22:53 kmel left 22:54 Hor|zon joined 22:55 samb1 joined 22:57 zacts joined 22:58 Hor|zon left 23:00 skids joined, firstdayonthejob left 23:09 zacts left 23:14 margeas left 23:15 uruwi left 23:16 hohoho joined 23:20 hohoho left 23:35 TEttinger joined 23:36 mattprelude left
timotimo lizmat: didn't we introduce first-index and last-index because we felt "adding adverbs" was a design smell? 23:40
what has changed?
23:43 thundergnat joined 23:50 llfourn joined 23:51 RabidGravy left 23:55 Hor|zon joined, llfourn left 23:59 Hor|zon left

Powered by Cro and the Raku® Programming Language.

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

Thank you!