»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
TimToady ZoffixLappy: obviously we're already to Step 3: Then they fight you. 00:14
ZoffixLappy :) 00:15
Xliff Next stop: "You Win!" 00:16
00:19 pierre_ joined
BenGoldberg m: for 1..2 { say $^x, $^y }; for 1..2 { say $^y, $^x }; 00:22
camelia rakudo-moar 3cab7c: OUTPUT«12␤21␤»
BenGoldberg While having placeholder arguments sorted alphabetically makes intuitive sense, istr them being documented as being used in order of appearance... that is, both of the above two ought to produce the same output. 00:25
ZoffixLappy If they're documented that way, it's an error that should be fixed. Perhaps, you're confusing regex captures? 00:29
m: 'foobar' ~~ /(foo(bar))/; say $0; say $1
camelia rakudo-moar 3cab7c: OUTPUT«｢foobar｣␤ 0 => ｢bar｣␤Nil␤»
ZoffixLappy :/ not the output I expected....
m: 'foobar' ~~ /(foo) (bar)/; say $0; say $1 00:30
camelia rakudo-moar 3cab7c: OUTPUT«｢foo｣␤｢bar｣␤»
ZoffixLappy Interesting...
BenGoldberg en.wikibooks.org/wiki/Perl_6_Progr..._Arguments 00:31
m: dd 'foobar' ~~ /(foo(bar))/; 00:32
camelia rakudo-moar 3cab7c: OUTPUT«Match.new(ast => Any, list => (Match.new(ast => Any, list => (Match.new(ast => Any, list => (), hash => Map.new(()), orig => "foobar", to => 6, from => 3),), hash => Map.new(()), orig => "foobar", to => 6, from => 0),), hash => Map.new(()), orig => "foobar…»
ZoffixLappy Fixed. Thanks. 00:34
00:34 wamba left 00:36 Ben_Goldberg joined, BenGoldberg left, Ben_Goldberg is now known as BenGoldberg
ugexe is there a way to get around another subroutine calling `exit()` other than &*EXIT? 00:39
&exit.wrap({ return; }); still exits
ZoffixLappy m: sub foo {exit}; &exit.wrap({ ; }); foo; say "Meow" 00:43
camelia rakudo-moar 3cab7c: OUTPUT«Meow␤»
ZoffixLappy Seems to work?
00:58 Actualeyes joined 01:03 tardisx left 01:09 ZoffixWin_ joined, ZoffixWin_ left, ZoffixWin_ joined 01:11 tardisx joined 01:22 Herby_ joined
Herby_ Evening, everyone! 01:22
o/
hotel \o
ZoffixLappy \o 01:26
Herby_ o/
01:26 molaf joined
Herby_ ZoffixLappy: you're a blogging machine! 01:27
I'm learning quite a bit reading through them, hope you keep it up :)
hotel oh that's a good idea
make a blogging machine in p6
Herby_ when in doubt, blog it out
hotel someone made a shitpost generator in java for tumblr 01:28
ZoffixLappy Herby_, thanks. I'm doing it for quite selfish reasons of trying to learn the stuff myself :) This morning, I still had no solid idea of how to use the S/// operator and now I do :)
Herby_ hotel: I just looked it up, here's the tagline for it: "I am a JAVA program who's purpose is to churn out the worst posts imaginable" 01:33
thats great
hotel yep
Herby_ and the 5th post on that page: "I am the zodiac" 01:35
ZoffixLappy heh 01:36
hotel so like I said, why not make one in p6? :P
maybe that'll be my first project
Herby_ blog about the blog project. i'd read it
i don't have the chops for that yet
hotel I don't have any chops yet, still can't get rakudo to build right 01:37
though that's mostly my fault 01:38
>shiny tools have shiny problems< 01:40
dalek osystem: fcf6c5e | (Andy Weidenbaum)++ | META.list:
Add module TXN::Parser
 01:45
osystem: b119fb8 | (Zoffix Znet)++ | META.list:
Merge pull request #201 from atweiden/txn-parser

Add module TXN::Parser: github.com/atweiden/txn-parser
01:48 pierre_ left 01:55 colomon joined 01:56 AlexDaniel joined 01:57 pierre_ joined
AlexDaniel this is so awesome essenmitsosse.de/pixel/?showcase=tr...mp;slide=4 02:02
reminds me of perl a bit :)
m: use MONKEY-TYPING; augment class Int { method foo { say self } }; int.foo 02:05
camelia rakudo-moar 3cab7c: OUTPUT«(int)␤»
ZoffixLappy AlexDaniel, heh... I hope it's not this part that reminds you of Perl: i.imgur.com/jp9yT6v.jpg 02:10
:)
Pretty cool though... Other than the crash.... 02:11
AlexDaniel ZoffixLappy: no, it does not remind me that your browser is broken :)1 02:12
ZoffixLappy lol
02:14 Herby_ left 02:15 winger_ joined 02:18 kid51 left 02:24 noganex joined 02:29 pierre_ left 02:33 hotel_california joined
dalek c: 7ad327f | (Zoffix Znet)++ | doc/Language/operators.pod:
Fix heading levels
 02:35
02:47 winger__ joined 02:48 pierre_ joined 02:51 winger_ left, pierre_ left 02:52 pierre_ joined 02:57 pierre_ left 02:58 pierre_ joined, colomon left 03:05 colomon joined 03:07 colomon left 03:19 hotel_california is now known as hotel 03:29 vendethiel joined 03:41 Amnez777 joined 03:44 Actualeyes left 03:45 colomon joined, Amnez777- joined 03:47 Amnez777 left 03:50 BenGoldberg left 04:17 pierre_ left 04:19 pierre_ joined 04:22 pierre_ left 04:23 pierre_ joined 04:28 Amnez777- left, Cabanossi joined 04:29 Actualeyes joined 04:31 pierre_ left 04:33 Amnez777 joined 04:34 vendethiel left 04:35 Timbus joined 04:36 pierre_ joined 04:37 pierre_ left
MadcapJake m: PROCESS::<&EXIT> = -> $_ { Nil }; exit 04:39
camelia ( no output )
MadcapJake ugexe: ^^
04:40 hotel is now known as hotel|netteiru
jdv79 ZoffixWin_: are you an educator or well versed presenter? 04:40
nice posts lately. 04:41
04:41 Actualeyes left 04:55 Actualeyes joined 05:03 vendethiel joined 05:16 zhmylove joined 05:21 al2o3-cr joined 05:22 rurban joined 05:24 vendethiel left, jjido joined 05:26 zhmylove left 05:28 m0ltar left 05:29 zhmylove joined 05:30 m0ltar joined 05:37 pierre_ joined 05:40 chienjo joined 05:42 chienjo left 05:43 pierre_ left 05:44 aries_liuxueyang joined 05:45 domidumont joined, domidumont left 05:46 domidumont joined 05:50 domidumont left 05:51 domidumont joined, jjido left 05:52 finanalyst joined
finanalyst hi. Is there any good way to test code containing 'warn'? 05:52
yoleaux 22 Feb 2016 20:37Z <ab6tract> finanalyst: your 'Intro to Sixy Perl' pastebin is awesome! you should consider publishing it somewhere more permanent!
finanalyst i wrote to perl-users, but thought I would ask here too. 05:54
05:56 finanalyst left, finanalyst joined 05:57 CIAvash joined 05:58 cognominal left 05:59 sno joined 06:01 finanalyst left 06:02 _mg_ joined 06:05 finanalyst joined 06:07 domidumont left 06:10 zhmylove left 06:13 zhmylove joined 06:20 hotel joined 06:21 hotel|netteiru left 06:24 firstdayonthejob joined 06:25 domidumont joined 06:26 zhmylove left 06:30 zhmylove joined 06:31 CIAvash left 06:34 CIAvash joined, hotel_california joined
TimToady finanalyst: warn throws a control exception that can be caught with a CONTROL block 06:35
06:35 hotel left
finanalyst TimToady: that I understand. But it is how to do this in a test snippet that I'm not sure about. Where does the CATCH block go 06:36
it would seem intuitive to me that there is a 'warns-ok' routine in TEST 06:37
06:38 jjido joined, firstdayonthejob left 06:43 jjido left 06:44 pierre_ joined, jjido joined 06:45 rurban left, zhmylove left
masak mornin', #perl6 06:47
06:48 zhmylove joined 06:49 jjido left 06:52 RabidGravy joined
moritz \o masak, * 06:53
RabidGravy oink! 06:57
_nadim Morning all 06:58
07:02 jamesnek1 joined 07:04 cschwenz joined
cschwenz \o #perl6 :-) 07:04
07:08 sue joined 07:10 abraxxa joined
cschwenz in perl6, how do i get the computer to do nothing for 0.012 seconds? (i need to translate some perl5 code which has "select( undef, undef, undef, 0.012 );") 07:13
moritz cschwenz: sleep 0.012 07:14
07:15 zhmylove left
moritz I hope you're not worried that it only applies to the current thread, not the whole computer :-) 07:15
cschwenz \o/ perl6 has a sane sleep !
moritz: i'm very okay with it only applying to the current thread :-) 07:16
moritz m: sleep 0.012; say now - INIT now
camelia rakudo-moar 3cab7c: OUTPUT«0.013980␤»
RabidGravy or if you actually want to do something something in the future but could usefully be doing something in the meantime, see Promise.in($secs) 07:17
07:18 boegel joined
_nadim Do we have a module that given something that looks like a command line would parse it? The format of the command line itself is not known (user defined) 07:19
cschwenz isn't there a port of Getopt::Long?
RabidGravy I think so 07:20
_nadim There is, completely undocumented but i wrote "format of options are unknown"
RabidGravy modules.perl6.org/#q=getopt
07:21 tardisx left
_nadim I did that, Getopt:::Long is not documented, Getopt::Tiny makes the same mistake as P5 and the doc is, well, fuzzy, Getopt::Type doesn't fit the work. 07:21
I smell a Getopt::Whatever module coming 07:22
RabidGravy yes I think you need to read some code
and/or help make one of the existing modules do what you want 07:23
07:23 zhmylove joined
_nadim I read the codes, and if there is nothing available, Getopt::Whatever will come, but it would be a pitty to no use what already exists. 07:23
RabidGravy on the face of it, it doesn't seem that difficult 07:24
07:24 tardisx joined
_nadim No, a grammar and some actions. And not even that is necessary. 07:25
but I want to finish what I am doing not start another module in the middle of it. 07:27
RabidGravy life sucks like that
07:27 _28_ria joined
_nadim specially when the thing one is doing is a template to make starting new modules easier ;) 07:29
07:33 domidumont left 07:34 telex joined 07:36 domidumont joined 07:39 hotel joined 07:40 hotel_california left 07:42 daniel-s joined, aries_liuxueyang left, leont joined 07:43 ufobat joined
ufobat god morning! 07:43
RabidGravy erp
07:44 zhmylove left, jjido joined 07:49 tardisx left, jjido left, al2o3-cr left 07:52 darutoko joined, zhmylove joined 07:53 dolmen joined 07:55 zakharyas joined 07:56 sue left, maybekoo2 joined 07:57 zhmylove left, aries_liuxueyang joined 08:10 Amnez777 left, nebuchadnezzar joined, Amnez777 joined 08:12 g4 joined 08:15 ZoffixWin_ left 08:20 al2o3-cr joined 08:25 al2o3-cr left 08:26 dolmen left 08:27 CIAvash left
dalek osystem: 515bf6e | RabidGravy++ | META.list:
Add App::ModuleSnap

See github.com/jonathanstowe/App-ModuleSnap		 08:29
RabidGravy Xliff, ^
08:31 wamba joined
RabidGravy thus ensuring that the next one is my fiftieth 08:31
Any objections if I alter the doc.perl6.org/language/modules#Dist...ng_Modules to have META6.json rather than META.info? 08:33
reason being that the future PAUSE thingy requires that afaiu 08:34
also do both zef and panda support the specced "source" now? 08:36
08:39 leont left 08:43 Cabanossi left, ufobat left
_nadim jnthn: App-ModuleSnap++ 08:44
jnthn ...what? :) 08:45
08:46 jjido joined
_nadim hehe, the other jonathan :) 08:47
08:50 Cabanossi joined, jjido left 08:52 rindolf joined 09:04 tardisx joined 09:06 tardisx left
RabidGravy easy mistake to make, he's the one with the beard 09:13
but jnthn++ just because
_nadim RabidGravy++ (now I know who you are ;) ) 09:20
09:24 zhmylove joined
RabidGravy :-O 09:25
stmuk_, you about? 09:26
09:26 domidumont left
cschwenz so, from doc.perl6.org/language/operators#Op...Precedence , what do "~&", "~<", and "~>" do? (the only mention of them i can find is in the linked operator precedence table) 09:27
psch cschwenz: bit-wise and, bitshift left, bitshift right, in order
oh no
Str-wise, not bit-wise
i don't know if the semantics are actually clear
hm, char wise maybe, actually? 09:28
moritz they make sense on buffers
but on Grapheme clusters?
psch m: say "foo" ~& "fou"
camelia rakudo-moar 6cb231: OUTPUT«foe␤»
psch m: say "foò" ~& "fou"
camelia rakudo-moar 6cb231: OUTPUT«fop␤»
psch .u snowman
yoleaux U+2603 SNOWMAN [So] (☃)
psch .u rocket
yoleaux U+1F680 ROCKET [So] (🚀)
psch m: say "🚀" +& "🚀" 09:29
camelia rakudo-moar 6cb231: OUTPUT«Cannot convert string to number: base-10 number must begin with valid digits or '.' in '3⏏5🚀' (indicated by ⏏)␤ in block <unit> at /tmp/R2_bby_Znw line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/R2_bby_Znw line 1␤␤»
psch m: say "🚀" ~& "🚀"
camelia rakudo-moar 6cb231: OUTPUT«🚀␤»
psch oh ffs :)
m: say "🚀" ~& "☃"
camelia rakudo-moar 6cb231: OUTPUT«☀␤»
cschwenz :-D
so they would appear to work on NFG characters? 09:30
moritz they appear to work, yes 09:31
psch m: say ("🚀".ord +& "☃".ord).chr # looks equivalent from here
camelia rakudo-moar 6cb231: OUTPUT«☀␤»
RabidGravy yeah, bitwise is +< etc 09:32
m: say 2 +< 2
camelia rakudo-moar 6cb231: OUTPUT«8␤»
jnthn It works at codepoint level, iirc
So it operates on each codepoint in an NFG synthetic
Which means you may or may not get another synthetic out :)
cschwenz jnthn++
psch .u combining
yoleaux U+0300 COMBINING GRAVE ACCENT [Mn] (◌̀)
U+0301 COMBINING ACUTE ACCENT [Mn] (◌́)
U+0302 COMBINING CIRCUMFLEX ACCENT [Mn] (◌̂)
cschwenz that is what i was trying to ask about
psch m: say "🚀\c[COMBINING GRAVE ACCENT]" ~& "☃" 09:33
camelia rakudo-moar 6cb231: OUTPUT«☂␤»
jnthn Whether it's useful for anything outside of ASCII cute tricks is another matter :)
oh interesting.... 09:35
I'm not sure it actually is doing what I thought it would be
github.com/MoarVM/MoarVM/blob/mast...ps.c#L1373
moritz well, there is a use case for string bitwise ops
for example if you want to to find borders in a text table
you can start with chr(255) x 80, and ~& every line 09:36
and see what columns remain in the end
if your data is large enough, only separate columns have non-zero bytes in the end
09:37 domidumont joined
jnthn Cute :) 09:37
09:37 zhmylove left
moritz (or if you know the separator character in advance, you can see the string with that instead of the chr(255), and see which columsn remain the same) 09:38
jnthn: that's signed integer arithmetic in MVM_string_bitand, right? 09:40
09:41 zhmylove joined
moritz can this create the number of a not-yet-created grapheme cluster? 09:41
jnthn moritz: Yeah, that's what I fear might happen 09:42
It's not clear why it works this way
moritz m: say "नि" & "5"
camelia rakudo-moar 6cb231: OUTPUT«all(नि, 5)␤»
moritz m: say "नि" ~& "5"
camelia rakudo-moar 6cb231: OUTPUT«4␤»
moritz m: say "नि" ~& "\r\n"
camelia rakudo-moar 6cb231: OUTPUT«नि␤»
RabidGravy can anyone else remember what the github alternative thingy that stmuk_ is using is? 09:43
some GitFoo
moritz RabidGravy: bitbucket? gitlab?
RabidGravy gitlab
cheers, just changing the wording of the modules.pod to make it clear that github is per example rather than an absolute requirement 09:44
moritz m: say "नि" ~> 1
camelia rakudo-moar 6cb231: OUTPUT«infix:«~>» not yet implemented. Sorry. ␤ in block <unit> at /tmp/3CjJH3GbMW line 1␤␤»
moritz m: say "नि" ~< 1 09:45
camelia rakudo-moar 6cb231: OUTPUT«infix:«~<» not yet implemented. Sorry. ␤ in block <unit> at /tmp/rBvD9tOSOA line 1␤␤»
moritz not sorry at all
09:46 jjido_ joined 09:50 nowan_ joined 09:51 jjido_ left
stmuk_ RabidGravy: gitlab 09:54
RabidGravy and that works fine for modules doesn't it (before I commit this)
stmuk_ it pretty much resembles github and you can login with gh creds 09:55
RabidGravy cool
stmuk_ yes some features aren't visible on modules.perl6.org butapart from that that works fine
09:56 domidumont left
RabidGravy excellent 09:57
is it worth explaining *why* you probably want to use META6.json rather META.info in the modules doc? I'm torn 09:58
stmuk_ colour syntax in vim
RabidGravy :)
I think I'll leave it, not much point describing forward compatibility when it's not quite clear what the end state will be 10:00
10:01 zhmylove left 10:04 zhmylove joined 10:05 domidumont joined
dalek c: 28989a3 | RabidGravy++ | doc/Language/modules.pod:
A few small changes:

   * prefer META6.json to META.info
   * more explanation on 'provides'
   * make it clear that other repos than github may work		 10:06
10:08 Actualeyes left
RabidGravy what next? 10:17
psch full SysEx support for Audio::PortMIDI? :P 10:19
10:19 maybekoo2 left
RabidGravy I thought you were sending me a patch for that ;-p 10:19
psch yeah, i might eventually :) 10:20
RabidGravy lots of things that I could fix up in my existing modules, also some 5 to 10 modules that I haven't finished yet
cschwenz error message feature request: when perl6 cannot find a matching signature, perl6 prints all of the signatures it found for a multi-method; will we please add the signature it was trying to find? 10:21
RabidGravy it does doesn't it?
psch m: multi f($) { }; multi f($, $) { }; f 1, 2, 3
camelia rakudo-moar 6cb231: OUTPUT«5===SORRY!5=== Error while compiling /tmp/qtM5TzAf4I␤Calling f(Int, Int, Int) will never work with any of these multi signatures:␤ ($) ␤ ($,)␤at /tmp/qtM5TzAf4I:1␤------> 3multi f($) { }; multi f($, $) { }; 7⏏5f 1, 2, 3␤»
RabidGravy yeah
psch cschwenz: it does. at the very beginning :)
ah, we still have a bug with disappearing anon positionals there :/ 10:22
RabidGravy it's different for multi methods
psch m: say :($, $).perl
camelia rakudo-moar 6cb231: OUTPUT«:($, $)␤»
psch m: say :($, $).gist
camelia rakudo-moar 6cb231: OUTPUT«($,)␤»
RabidGravy eugh
cschwenz so, i am seeing "Cannot call AUTOGEN( […] ); non of these signatures match: […]" are you saying the "[…] AUTOGEN( […] ) […]" is the signature it was trying to find? 10:23
RabidGravy ah that old AUTOGEN thingy
cschwenz do i need to update/upgrade perl6?
psch doesn't know about any AUTOGEN thingies
RabidGravy if you add "proto method <name> (|c) { * }" you will get the right name 10:24
psch, under some circumstances it doesn't create the proto name right so it just leaves it as "AUTOGEN" 10:25
I've never figured out quite what those circumstances are sufficiently to RT IT
psch o.o
that sounds a bit bad
cschwenz: can you golf the code that gives you that error? 10:26
in case it stays around when golfed this might be an angle to try and figure out what's actually happening 10:27
cschwenz psch: probably
10:27 Actualeyes joined
RabidGravy it's something in the area of more than one multi with a certain number of candidates 10:28
cschwenz so, with that solved, next question… if i have multi-methods which take a single positional value, and said value has a default, why does "$obj.foo()" not do what i mean and use the default? (instead it complains about not finding a signature match) 10:29
psch m: class A { multi method f($ = "foo") { say "ok" }; }; A.f
camelia rakudo-moar 6cb231: OUTPUT«ok␤»
psch there most be more to that issue :) 10:30
RabidGravy I think I've encountered that one before too, can't remember the circumstances
10:30 zhmylove left
RabidGravy but yeah a golfed example of that one would be good 10:31
cschwenz p6: class Foo { multi method bar( Int $baz = 1 ) { say "Int $baz"; }; multi method bar( Rat $baz = Rat.new(1, 1) ) { say "Rat $baz"; }; }; my $obj = Foo.new(); $obj.bar(); 10:32
camelia rakudo-moar 6cb231: OUTPUT«Ambiguous call to 'bar'; these signatures all match:␤:(Foo $: Int $baz = 1, *%_)␤:(Foo $: Rat $baz = { ... }, *%_)␤ in block <unit> at /tmp/tmpfile line 1␤␤»
psch oh, that's easy. two candidates with one parameter with default are two candidates with 0 parameters
cschwenz hmm, that error message is different than the one i'm seeing
psch cschwenz: maybe you do have to upgrade your rakudo :)
cschwenz psch++ 10:33
10:33 zhmylove joined
psch cschwenz: if it sticks around you might simply have golfed too much, too 10:33
...although i'm at a loss to guess what else could cause what you're describing 10:34
10:39 cdg joined 10:43 cdg left 10:44 daniel-s left 10:47 robertle joined, jjido_ joined, tbrowder joined
tbrowder need help with env var PERL6LIB: rakudo (nom) seems not to recognize the colon path separator 10:49
...I set "export PERL6LIB=path1" and rakudo finds my module; I set "export PERL6LIB=path1:path2" and the module is not found; am I doing something stupid? 10:50
10:50 daniel-s joined
psch tbrowder: PERL6LIB expects specs seperated by comma 10:51
a spec doesn't have to be a path, either
which is why we don't use colon as a separator i guess, as it could be needed for a protocol
like url#git://github.com/someone/something.git 10:52
10:52 jjido_ left, daniel-s left 10:53 daniel-s joined, zhmylove left, Actualeyes left 10:54 Actualeyes joined
tbrowder psch: oh! thanks, that works! next question then: how do I get the rakudo to search the paths in another rakudo install where the resources are specified differently? 10:54
psch i don't understand the question. you want one rakudo executable to include the (sha1'd) module sources from another? 10:55
10:56 zhmylove joined
tbrowder psch: yes, or else tell me how to install them into the new rakudo build 10:56
psch tbrowder: RabidGravy++ recently published a module for exactly that 10:57
RabidGravy :)
robertle I have problems with understanding the very basics of modules as well: to get started I create a single module with a single sub with "is export" in it. I can use that module, and call the sub. great. but I am very confused about the namespacing: I seem to be only able to access exported stuff (which is ok), and the export puts it in the namespace of the loader. so the actual namespace of the sub does not
10:57 _28_ria left
robertle matter anymore. interestingly the "unit module xyz" at the top of my module does not need to match the path of it (also ok), so effectively the namespace does not matter at all! that can't be right, there must be more to it... 10:57
am I right in assuming that the "use" looks up by filename/path, so that does not have to match the namespace? 10:58
RabidGravy robertle, subs are "my" scoped by default, they aren't in a "namespace", if you want to refer to them with a full package name, qualify them with "our"
10:58 Actualeyes left
psch tbrowder: App::ModuleSnap, fyi 10:59
10:59 daniel-s left
RabidGravy but yes, the default behaviour of the repos is to derive the path from the module name supplied to "use" or "require" 10:59
of course a CompUnit::Repository could be made that does something else 11:00
robertle RabidGravy: right, so if I "need" then the contents of the file in question get loaded. if that's "our" then I can access it by namespace, if it is "my" then it's private to the namespace and inaccessible. if I use "use" instead, the loaded bits also are made availabel in the loading namespace, so no need to prefix anymore. that abut right? 11:01
RabidGravy for installed modules it is different again, as the name of the module isn't directly reflected in the installed module name
robertle, not quite
tbrowder psch: so how will App::ModuleSnap help me if I can't get it installed for the new rakudo? 11:02
robertle RabidGravy: right, the file vs namespace is a bit academic because you want them to match up anyway, everythign else is just plain confusing
psch tbrowder: i have no idea how it works, but i'd assume you install it for the old rakudo that already has the modules you need
11:02 Actualeyes joined
RabidGravy yes 11:02
psch tbrowder: RabidGravy probably appreciates any feedback re documentation or functionality :)
11:02 finanalyst left
RabidGravy it basically traverses the $*REPO.rep-chain, finding all the Distributions which aren't something it made or "CORE" and then writes out a META6.json with all those as dependencies 11:03
so in your other perl6 you do "panda install ." and voila
tbrowder psch: ah! okay, I'll give it a shot 11:04
I have to go now but will try it and report back--thanks!!
11:04 tbrowder left
RabidGravy because I seem to install a lot of modules for testing purposes I find myself with the need to completely eradicate everything once in a while and reinstall the modules 11:05
I've just discovered that "Flash Magic Eraser" is perfect for cleaning the white rubber parts of Converse All Stars, almost box fresh looking ;-) 11:08
11:09 kid51 joined
RabidGravy is there something other than github "organizations" that people use for organising e.g. a "special interest group"? I'm thinking of starting a "Perl6 Audio" thingy 11:11
psch haha 11:12
DrForr SIGAUD, ISAGN.
psch special interest group made me laugh :)
RabidGravy :)
psch i don't think they have anything like that on github though
i'd totally join an org for that though, even if it might be somewhat inappropriate as an org on github vOv 11:13
11:13 ufobat joined
RabidGravy well an "organisation" gives you all the github stuff, wiki, issues and a github.io site but yeah it's not so much about the code as collecting and promoting the stuff 11:14
ufobat ola! :D just in case someone cares! my notebook is alive again
RabidGravy and of course becoming an evil and secretive cabal
psch right, but a github-page would be a good start for promoting i think
11:15 wamba left
RabidGravy so yeah, create a github organisation and create the page 11:15
psch i mean, unless you have an appropriate domain lying around... :)
11:16 wamba joined
RabidGravy jeezus-knievel.com? 11:16
perlmonkeys.org? plastic-fernandez.com, its-going-to-be-fabulous.com all unused but not really suitable ;-) 11:17
goatponi.es
psch its-going-to-be-fabulous.com is pretty close though
robertle RabidGravy: let me try again for "our" and "need" only. "our" stuff gets put in a namespace according to the enclosing "module". "need" loads a unit and therefore makes the namespaces within and what is "our" in them available. "my" stuff from the loaded unit is unavailable because it is lexically scoped, and the unit is the lexical scope. right? 11:18
RabidGravy yes, but "need" is a red herring here,
however you load it if the sub in a module isn't our scoped you can't access it at all outside the module 11:19
robertle RabidGravy: right, I said "need" only because "use" does something on top of it. let's check my "use" and "is export" understanding: "is export" or the more elaborate versions thereof takes whatever we are talking about and places it in a special namespace of the module. if you "use" a unit, everything from that special namespace makes it available in the loader. about right? if so: does use put into the t 11:22
RabidGravy m: module Foo { sub bar() { say "bar" }; our sub baz() { say "baz" };}; Foo::baz(); Foo::bar();
camelia rakudo-moar 6cb231: OUTPUT«baz␤Could not find symbol '&bar'␤ in block <unit> at /tmp/JljyT__jh2 line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/JljyT__jh2 line 1␤␤»
robertle arget namespace or the target lexical scope?
psch "use" puts the used module into the current lexical scope, and imports all exported symbols into the current lexical scope 11:24
robertle RabidGravy: understood that part, but that's because bar() isn't part of the Foo namespace, just part of the lexical scope that happens to be created by the module/namespace block. rigt?
psch well, after (potentially compiling and) loading the source file
RabidGravy yes 11:25
11:25 zhmylove left
RabidGravy subroutines are lexically scoped by default 11:25
11:26 pmurias joined
psch m: say MY::.keys 11:26
camelia rakudo-moar 6cb231: OUTPUT«($=pod !UNIT_MARKER EXPORT $_ $! ::?PACKAGE GLOBALish $¢ $=finish $/ $?PACKAGE)␤»
psch and lexical scope is a namespace too /o\
robertle ok, great! somewhat off topic, but is there a way to "use" from a different namespace and put in this *namespace* rather than this lexical scope? there is such a concept in UML, and it is thoroughly confusing and interestingly impossiblt to implement in C++ and java, which makes you wonder why it is in UML in the first place...
psch robertle: i think reexportation could cover that at least partially, but i don't know in how far that's implemented 11:27
RabidGravy yeah, there are export tricks available
robertle psch: interesting! I don't really have any use for it, just wondering...
psch m: module Foo { use Test :EXPORT }; import Foo; 11:28
camelia rakudo-moar 6cb231: OUTPUT«5===SORRY!5=== Error while compiling /tmp/g5NffKCkaG␤Error while importing from 'Test': no such tag 'EXPORT'␤at /tmp/g5NffKCkaG:1␤------> 3module Foo { use Test :EXPORT 7⏏5}; import Foo;␤»
psch that's the speculated syntax, but yeah...
RabidGravy well at it's simplest you can define a "sub EXPORT" in your compilation unit which returns a hash that is keyed on the names you want to export with *what* as the value 11:29
and some <hand wave> other thing with the EXPORT package 11:30
psch m: module Foo { use Test; }; import Foo; ok 1 11:31
camelia rakudo-moar 6cb231: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ExKc1N3mzA␤Undeclared routine:␤ ok used at line 1␤␤»
psch hm, i don't get the test case we have for that
oh, we're not running it anyway :) 11:32
github.com/perl6/roast/blob/master...e-export.t this one fyi
llfourn_ m: module Foo { use Test; BEGIN trait_mod:<is>(&ok,:export) }; import Foo; ok 1 # maybe?
camelia rakudo-moar 6cb231: OUTPUT«ok 1 - ␤»
psch llfourn_: yeah, that i'd expect to work, but see the linked test. that's what was once thought to be the easy way to make it work 11:33
it's still in roast, but we're not running it
llfourn_ mmm 11:34
I wrote a module to do it fyi
github.com/LLFourn/p6-CompUnit-Uti...-exporting
11:35 wamba left
RabidGravy llfourn_++ # making stuff 11:37
11:37 daniel-s joined
llfourn_ RabidGravy++ xx 10 :) making stuff 11:37
11:39 Actualeyes left 11:44 tbrowder joined
RabidGravy but it's nearly may and no-one has made a mailling list manager yet 11:44
11:48 jjido_ joined
tbrowder psch & RabidGravy: I realize now (I think) if I want to hack on rakudo I should create my own branch and let rakudobrew manage it on my own machine. That way there shouldn't be any major problems with standard module handling. Am I right? 11:48
RabidGravy psch, having done a cursory survey of the alternatives I'm thinking that a github organisation is easiest and makes collaboration easier and chuck a custom domain on it later
psch tbrowder: uhh, for actual hacking on rakudo itself i at most branch rakudo inside its own repo 11:49
tbrowder: and build with Configure.pl
RabidGravy tbrowder, I've always used rakudobrew for day to day stuff then what psch said for hacking on rakudo
tbrowder but the problem with that has been finding modules
..installed modules 11:50
psch i don't think i ever needed modules to hack on the compiler
i'd suggest golfing the code module code that is buggy and writing tests from that
RabidGravy yeah I see it as two different enterprises
llfourn tbrowder: I just do panda --prefix=/path/to/forked_rakudo/lib if I want to install a module on non-rakudobrew rakudo 11:51
RabidGravy and when I really do need to do it I just use one or more -I<path to module> on checkouts of the modules
couple of times fixing breakage in my own or other modules from change in rakudo I do that 11:52
11:52 jjido_ left 11:53 pierre_ left
RabidGravy or bisection of said breakage 11:53
tbrowder psch, RabidGravy, llfourn: okay, sounds good--thanks; bye
11:54 tbrowder left 11:56 pierre_ joined 11:59 rurban joined, kid51 left, rurban left 12:00 Actualeyes joined
literal if I want to loop over an arrayref stored in a hash, do I always need @() ? as in: for @(%hash<foo>) -> $bar { ... } 12:02
RabidGravy psch (or anyone else who might be interested) any thoughts on what one might call said Perl6 Audio organization? "Perl6 Audio Cabal", "Perl6 Noise Gang", "Butterflies on Acid"? 12:03
llfourn literal: |%hash<foo> is more idiomatic imo 12:04
psch Butterflies on Acid feels more like the first collab album published by the org tbh
RabidGravy indeed
literal llfourn: ah, of course
RabidGravy I'm going with the middle one
psch i do like Perl 6 Noise Gang yeah
note the space though
the synopses even said it has to be an unbreaking one
12:04 rba_ joined
psch ...i put that in a comment in a nqp-j java file once and then the build stopped working :P 12:05
grondilu wonders if we could not have %hash<foo>@ to mean @(%hash<foo>) 12:07
12:07 domidumont left
literal and is there a more straightforward way of getting all the matched strings of a group than this? @matches = $<mygroup>.map({~$_}) 12:07
_nadim jnthn: I pushed a pull request for Grammar::Tracer, the output should be like this with the patch imgur.com/YoFIvU4
llfourn in p5 there is $hash{foo}->@* now
(or *@..)
rba_ How to avoid passing getting an error when passing and Inf from a Num to an Int? e.g. > say Inf.Int 12:08
psch m: sub postfix:<@>(\c) { @(c) }; my %h = a => 1, b => 2; say %h@
camelia rakudo-moar 6cb231: OUTPUT«(a => 1 b => 2)␤»
grondilu psch: nice
psch i don't think that's a useful CORE op, fwiw
rba_: there's been a long-standing no-final-word kind of situation around that afaik 12:10
llfourn literal: I haven't seen a better way than that. It mgiht be kinder to write it as .map(*.Str) or .map(~*)
psch rba_: that's to say that the design documents speculated Int to somehow be able to accomodate Inf, but most implementors find it either wrong or too hard to make work usefully
rba_: practically i'd suggest you add a multi candidate that explicitely accepts Inf and does what you would do if it would be allowed in Int 12:11
m: "foo" ~~ /(.)+/; say @0 12:12
camelia rakudo-moar 6cb231: OUTPUT«[｢f｣ ｢o｣ ｢o｣]␤»
rba_ psch: I did it like this "return $b != Inf ?? round $b !! False;"
psch rba_: i'd consider returning the type object Int instead of False
m: "foo" ~~ /(.)+/; say ~<<@0 12:13
camelia rakudo-moar 6cb231: OUTPUT«[f o o]␤»
psch literal: ^^^ that also works, fwiw 12:14
rba_ psch: like this "return $b != Inf ?? round $b !! Int;" ?
psch rba_: yeah. that way you stick to your type at least
12:14 domidumont joined
psch rba_: in the end it strongly depends on how what else is going on, of course :) 12:14
RabidGravy So "Perl 6 Noise Gang" it is, I'll just check in the other place to see if there are any objecttions
gfldex m: my %e = enum <one two three>; say one.WHAT, " It's a ghost!";
camelia rakudo-moar 6cb231: OUTPUT«() It's a ghost!␤»
rba_ psch: Got it. Thanks 12:15
psch RabidGravy: ...there's another place for talking about Perl 6 and using it for music?
RabidGravy no, no #p6dev ;-)
psch oh :P
llfourn psch: I think >>/<< is not guaranteed to return it in the same order
jnthn llfourn: It gives results in order, but side-effects can be parallel/out-of-order 12:16
psch ...which hopefully doesn't matter for stringification of Match objects :)
otherwise i'd think we'd have a problem somewhere 12:17
llfourn jnthn: Ah.. thanks that makes sense now
gfldex m: my %e = enum <one two three>; say one.WHAT; my $foo = class :: { has $.one = 0 }; say $foo.WHAT; # Is this difference in introspectiveness intentional?
camelia rakudo-moar 6cb231: OUTPUT«()␤(<anon|81105168>)␤»
psch m: enum f < a b c >; say a.WHAT 12:18
camelia rakudo-moar 6cb231: OUTPUT«(f)␤»
psch gfldex: well, considering the enum you assign is anonymous...
ufobat what does Int(Any) in a Signature mean? github.com/supernovus/perl6-psgi/b...GI.pm6#L15 12:19
12:19 rindolf left
gfldex psch: so is the class 12:19
grondilu m: talk: { say "hi" } # it occured to me that labels can be used as comments
camelia rakudo-moar 6cb231: OUTPUT«hi␤»
12:19 zhmylove joined, huf joined
dalek href="https://perl6.org:">perl6.org: e64aafc | (Zoffix Znet)++ | fetch-recent-blog-posts.pl:
Fix bug with recent blog updater

Not all blogs (e.g. masak's) have a `type="text/html"` on the `<link>`. Exclude that restriction from the selector.		 12:19
psch gfldex: oh, right. yeah, that looks like we want something similar for the enum... 12:20
gfldex i shall rakudobug
psch m: my %e = enum < a b c >; say a.WHERE
camelia rakudo-moar 6cb231: OUTPUT«140372698805376␤»
psch m: my %e = enum < a b c >; say a.WHAT.WHICH
camelia rakudo-moar 6cb231: OUTPUT«␤»
llfourn m: sub foo(Int(Any) $i) { $i.say }; foo("5") # In this case it will coerce the Str "5" to Int 5
camelia rakudo-moar 6cb231: OUTPUT«5␤» 12:21
llfourn ufobat: ^^
12:21 sno left
psch m: class A { method Int { 5 } }; sub f(Int(A) $x) { say $x }; f A # more expansive example 12:21
camelia rakudo-moar 6cb231: OUTPUT«5␤»
psch Int(Any) seems really misleading to me, 'cause it's Cool that allows it and all... vOvo 12:22
-o
ufobat ah! coercion,
psch wonders what exactly happened to the COERCE proposal...
llfourn yes, of sorts 12:23
ufobat i'd like to read the documentation about that :)
RabidGravy anyone else want an invite to Perl6-Noise-Gang ?
llfourn ufobat: docs.perl6.org/language/functions#Coercion_Types
dalek href="https://perl6.org:">perl6.org: b476ad6 | (Zoffix Znet)++ | fetch-recent-blog-posts.pl:
Make link finder smarter

Commit 061cbe2265a6ed9eb9d5bb3a8daee21f96c7bd8b suggests not restricting <link>s by text/html will give bogus URLs, but some blogs do not have text/html		 12:24
ufobat llfourn :D thank you
sjn RabidGravy: noise gang? 12:25
RabidGravy sjn, audio/music cabal
:) 12:26
sjn ah, cool
RabidGravy just thought it was about time we got some application domain specific thingies going
sjn fair enough
RabidGravy sjn, want an invite then while I've got the page open? 12:27
sjn isn't much of an audio geek at all
llfourn if I have a signature like :(Int $a,Str $b) is there a way to attempt to bind to it and get a useful exception if it doesn't work?
psch and off goes Music::Helpers
llfourn :($a,$b).try-bind(1,2,3) # eg throws and exception
psch now i can slack off on that again, 'cause there's an org to take care of it \o/
RabidGravy ;-p 12:28
github.com/Perl6-Noise-Gang 12:31
anyone a dab hand with photoshop/gimp/paint want to knock up a quick "camelia with some musical thing" logo 12:32
12:33 wamba joined
dalek c: 741f52b | (Wenzel P. P. Peppmeyer)++ | doc/Language/typesystem.pod:
doc anonymous enums
 12:36
12:37 abraxxa left 12:39 pierre_ left 12:41 pierre_ joined
psch RabidGravy: we also need all kind of domain-appropriate groups i think 12:41
...well, maybe not, i can't even think of one really fitting (and somewhat humorous) example vOv 12:42
_nadim RabidGravy: what would Perl6-Noise-Gang be? 12:43
RabidGravy for people interested in doing audio/music related stuff
_nadim Nice but I suck at music, can barely put a song on ;) 12:44
12:45 pierre_ left 12:48 m0ltar left
psch hmm, i wonder if travis still does Music::Helpers correctly now... 12:48
12:48 jjido_ joined
RabidGravy I ought to try putting Audio::PortMIDI in travis, I wasn't sure how much of alsa it had to have the default sequencer ports an shizzle 12:51
psch hm, shouldn't it just work if you send stuff back and forth across the loopback? 12:53
12:53 jjido_ left
_nadim Can someone look and see if there is a obviously wrong in the grammar? nopaste.linux-dev.org/?1096659 12:54
RabidGravy yeah, it was the presence or not of the "Midi Through" device
12:54 cdg joined
grondilu I have the error message: 12:54
Cannot find method 'package_at_key': no method cache and no .^find_method
and yet my results seems fine. I don't understand what's wrong. 12:55
I mean it does find the sub since the results are good.
(or method whatev)
12:56 domidumont left 12:57 domidumont joined
grondilu I'll try to golf it 12:58
13:02 rurban joined 13:05 daniel-s left, andreoss joined 13:06 rurban left 13:07 daniel-s joined 13:09 tbrowder joined
RabidGravy in HTML is it "Perl&nbsp;6" 13:11
13:11 pmurias left
perlpilot yes 13:11
RabidGravy yeah that works 13:13
tbrowder psch & llfourn & RabidGravy: yt hints helped to make my hackable rakudo work, but I found some gotchas: (1) panda needs to be executed from OUTSIDE the rakudo repo fork, (2) the best choice of place to install seems to be /path/to/rakudo-repo/install/share, and (3) I still needed to specify the install path with PERL6LIB; but I now have a working solution--thanks!
RabidGravy cool 13:14
llfourn tbrowder: yep I think I had to do all 3 of those too :)
andreoss what would be the reverse operation of .uniname? 13:15
m: "a".uniname.say
camelia rakudo-moar 6cb231: OUTPUT«LATIN SMALL LETTER A␤»
andreoss m: "a".uniname.reverse_uniname == "a"
camelia rakudo-moar 6cb231: OUTPUT«WARNINGS for /tmp/60ucTlAn_h:␤Useless use of "==" in expression ".reverse_uniname == \"a\"" in sink context (line 1)␤Method 'reverse_uniname' not found for invocant of class 'Str'␤ in block <unit> at /tmp/60ucTlAn_h line 1␤␤»
psch m: say "\c[LATIN SMALL LETTER A"]
camelia rakudo-moar 6cb231: OUTPUT«5===SORRY!5=== Error while compiling /tmp/1jd8KQYr5C␤Unrecognized character name LATIN SMALL LETTER A"␤at /tmp/1jd8KQYr5C:1␤------> 3say "\c[LATIN SMALL LETTER A"7⏏5]␤»
psch huh 13:16
andreoss as a method probably?
psch m: say "\c[LATIN SMALL LETTER A]"
camelia rakudo-moar 6cb231: OUTPUT«a␤»
psch oohh
"] vs ]"
andreoss: what do you propose that method does? 13:17
andreoss: i mean, it has to be a method on Str, right?
13:17 sufrostico joined
andreoss it takes uniname and return a string much like charnames::str_vianame in p5 13:18
13:18 cdg left
andreoss *string_vianame 13:18
13:18 cdg joined 13:19 Sgeo left
psch i don't think we have anything like that currently 13:19
13:20 _4d47 joined
tbrowder I think I'll propose the tidbits for the docs... 13:21
13:23 cdg left
andreoss m: "Hello".uninames>>.subst(/^ <?before LATIN | DIGIT >/, "FULLWIDTH ").map({ "\\c[" ~ $_ ~ "]" }).join.say 13:23
camelia rakudo-moar 6cb231: OUTPUT«\c[FULLWIDTH LATIN CAPITAL LETTER H]\c[FULLWIDTH LATIN SMALL LETTER E]\c[FULLWIDTH LATIN SMALL LETTER L]\c[FULLWIDTH LATIN SMALL LETTER L]\c[FULLWIDTH LATIN SMALL LETTER O]␤»
13:25 cdg joined
psch m: ('"' ~ ("Hello".uninames>>.subst(/^ <?before LATIN | DIGIT >/, "FULLWIDTH ").map({ '\\c[' ~ $_ ~ ']' }).join) ~ '"').EVAL.say 13:25
camelia rakudo-moar 6cb231: OUTPUT«Ｈｅｌｌｏ␤»
psch andreoss: well, if i'm right and we really don't have anything like that i think i'd agree we could use it
andreoss EVAL is bad 13:26
13:27 tbrowder left
jnthn andreoss: Please file a feature request for a sub/method to map a char name into its codepoint, we don't have one of those yet 13:29
For now either EVAL, or `use nqp` and then nqp::codepointfromname('FULLWIDTH LATIN SMALL LETTER E')
13:30 zwu joined
jnthn Guess TimToady++ will be able to supply us a nice name :) 13:31
zwu is using full path name of symbols bypass the compiler check, for example M1::call_foo() in script?
jnthn (Which reverse_uniname is not. :P)
13:32 _28_ria joined 13:44 Actualeyes1 joined, Actualeyes left 13:45 maybekoo2 joined 13:46 ZoffixWin joined, ZoffixWin left, ZoffixWin joined
ZoffixWin jdv79, no. Neither. 13:47
13:49 domidumont left, jjido_ joined 13:50 domidumont joined
MadcapJake RabidGravy: I have a repo of Audio::FluidSynth that I started a while back (needs some TLC) but I haven't yet added it to my github, perhaps I could put it up in the Noise Gang org? 13:52
13:53 pierre_ joined
timotimo what is this noise gang org you're speaking of 13:53
MadcapJake timotimo: github.com/Perl6-Noise-Gang
timotimo neat
RabidGravy MadcapJake, what's your gh id?
MadcapJake MadcapJake
13:54 jjido_ left
timotimo MadcapJake, MadcapJake. 13:54
RabidGravy timotimo, want an invite? It's "timo" isn't it
MadcapJake MadcapJake xx 2
psch ooh, the logo is actually already pretty neat
RabidGravy just some note clipart I blagged ;-) 13:55
RabidGravy invites timotimo anyway
timotimo no, just "timo"
psch hah, but it has 6 note lines
timotimo that's perfect 13:56
for perl 6
RabidGravy :)
zwu if both modules have exported a same name sub foo() is export {...}, in the main script use m1; use m2; will give error: Cannot import symbol &foo from m2, because it already exists in this lexical scope. It will stop us to use both of m1 and m2, though the actual foo is defined under m1:: and m2::.
RabidGravy no
psch m: sub f { }; sub f { } # this also never works
camelia rakudo-moar 6cb231: OUTPUT«5===SORRY!5=== Error while compiling /tmp/JgRfKimQeH␤Redeclaration of routine f␤at /tmp/JgRfKimQeH:1␤------> 3f { }; sub f { } # this also never works7⏏5<EOL>␤»
RabidGravy don't export them by default and make them "our" subs 13:57
zwu importance of the export alias is definitly necessary I think.
RabidGravy you can do that
psch so merging GLOBALs from two used modules with identical exported symbols can never work either
zwu sometime we cannot control the modules to not to export.
ZoffixWin m: use nqp; "Hello".uninames>>.subst(/^ <?before LATIN | DIGIT >/, "FULLWIDTH ")>>.&({nqp::codepointfromname($_)})».chr.join.say
camelia rakudo-moar 6cb231: OUTPUT«Ｈｅｌｌｏ␤»
13:58 neuron joined
zwu for example if I use two third party modules, both have exported a same symbol. 13:58
MadcapJake It should default to overwrite globals and then you can still access the other via m1::<&foo> right?
psch zwu: well, use is lexical. you can always use the modules in different scopes
MadcapJake: no, if it's < my ... is export > you can't access it via the namespace long name
MadcapJake oh yeah the `my` makes it non-global 13:59
13:59 skids joined
RabidGravy use them in different scopes and then re-export them from that scope with different names even 13:59
psch m: module Foo { sub ok is export { say "ok!" } } }; { import Foo; ok }; { import Test; ok 1 }
camelia rakudo-moar 6cb231: OUTPUT«5===SORRY!5=== Error while compiling /tmp/RAw0A8aLfo␤Unexpected closing bracket␤at /tmp/RAw0A8aLfo:1␤------> 3 Foo { sub ok is export { say "ok!" } } 7⏏5}; { import Foo; ok }; { import Test; ok␤»
psch m: module Foo { sub ok is export { say "ok!" } }; { import Foo; ok }; { import Test; ok 1 }
camelia rakudo-moar 6cb231: OUTPUT«5===SORRY!5=== Error while compiling /tmp/iJP2sqhtvh␤Could not find module Test to import symbols from␤at /tmp/iJP2sqhtvh:1␤------> 3" } }; { import Foo; ok }; { import Test7⏏5; ok 1 }␤»
psch m: module Foo { sub ok is export { say "ok!" } }; { import Foo; ok }; { use Test; ok 1 } # grr
camelia rakudo-moar 6cb231: OUTPUT«ok!␤ok 1 - ␤»
robertle witin a start { ... }, how can I get any idea of the thread I am running under? an id or so? 14:00
psch overwriting symbols is really a bad idea in any case i think 14:01
timotimo m: start { say $*THREAD }; say $*THREAD
camelia rakudo-moar 6cb231: OUTPUT«Thread #3␤Thread #1 (Initial thread)␤»
timotimo also interesting is $*SCHEDULER
psch m: say $*THREAD.WHAT
camelia rakudo-moar 6cb231: OUTPUT«(Thread)␤»
grondilu apparently it was the use of race that produced the error mentioned above. 14:02
robertle timotimo: thanks! I knew $*SCHEDULER, but not $*THREAD...
timotimo at the moment, .race and .hyper are not really stable at all
14:03 ZoffixWin left
RabidGravy m: module Foo { sub ok is export { say "ok!" } }; my &foo-ok; my &test-ok; { import Foo; &foo-ok = &ok }; { use Test; &test-ok = &ok }; foo-ok; test-ok 1; 14:04
camelia rakudo-moar 6cb231: OUTPUT«ok!␤ok 1 - ␤»
RabidGravy not quite "re-exporting" but doable 14:05
grondilu could have you done 'our &foo-ok is export'? 14:08
14:08 _4d47 left
moritz I don't think we can export variables at this point 14:09
RabidGravy you can do that as an export sub though
14:09 m0ltar joined
_nadim Can we Haz a substr that would do what a lot (well .. ) of people want? when giving an end that is before the start of the sub string, just return an empty one. Although it has not happened lately, the forever checking if one is withing range, if one os going to count from the end or what kind of error, if any, will be send back is tiresome. and it would add some huffman encoding points to the whole handling of strings. 14:10
grondilu I meant it as a way to re-export as you just tried
m: module Foo { module Bar { our sub f {} }; import Bar; our &f is export = &Bar::f }; 14:11
camelia ( no output )
grondilu m: module Foo { module Bar { our sub f {} }; import Bar; our &f is export = &Bar::f }; import Foo; f();
camelia ( no output )
grondilu m: module Foo { module Bar { our sub f { say "hi" } }; import Bar; our &f is export = &Bar::f }; import Foo; f();
camelia rakudo-moar 6cb231: OUTPUT«hi␤»
RabidGravy yeah I thought that worked
grondilu cool, I was wondering how to do this. 14:12
RabidGravy but then moritz made me doubt myself
moritz sorry for that
grondilu it'd be easier if we had 'import Bar; re-export Bar;' 14:13
or just 'export Bar;' 14:14
14:14 cdg left
psch _nadim: what do you mean "end"? substr takes start position and length 14:14
zwu thanks for ans, i tried to use "{ use m1; ...}" and "{use m2;...}" under individual lexical code, it works. but when I replaced use with require, it doesn't work. 14:15
RabidGravy require doesn't import
you need to do the import explicitly 14:16
14:16 tharkun joined
RabidGravy you can also specify the actual symbol to import with require - e.g. "require Foo <&foo>" 14:16
zwu I used require "full path name";
how can I do if I want to import dynamically? 14:17
grondilu m: module A { our sub term:<§> {} }; say &A::term<§>; # not sure if such syntax is ok
camelia rakudo-moar 6cb231: OUTPUT«(Any)␤»
zwu I used required "full path file name"
grondilu RabidGravy: how would I do it for an infix? 14:19
I mean I can't write &Bar::infix:<*>, can I?
RabidGravy the same
yeah
grondilu oh wait
jnthn Use the bracketed infix syntax 14:20
RabidGravy or that
grondilu m: module A { our sub term:<§> {} }; say &A::term:<§>;
camelia rakudo-moar 6cb231: OUTPUT«sub term:<§> () { #`(Sub|87002416) ... }␤»
RabidGravy but they're just subs with funy names
grondilu I had forgotten a colon earlier. YEs it works indeed.
14:22 loren joined 14:23 vendethiel joined
grondilu but now, what if they are 'multi' and not 'only'? 14:23
14:23 geekosaur left
zwu thanks. using the bracket <&foo> works. 14:23
grondilu I mean I can not add multi candidates to say infix:<*> without using the 'multi infix:<*>(|c) {...}' syntax, can I? 14:25
14:25 dalek joined, ChanServ sets mode: +v dalek, john51_ joined, sftp joined, AlexDaniel left 14:26 geekosaur joined
zwu can we use bracket in the "import" or "use" command too? 14:26
MadcapJake grondilu: correct
14:26 psch_ joined
grondilu to I can not use the trick above to re-export multis. It's a bit LTA. 14:27
s/to I/so I/
14:27 psch_ is now known as psch
MadcapJake grondilu: yeah I've run into that before too 14:28
14:28 tharkun_ joined
dalek c: d42ebaa | VZ++ | doc/Language/regexes.pod:
Add matching "/" at the end of regex

Trivial change to avoid unterminated regex in an example.		 14:28
c: 3333a9d | (Zoffix Znet)++ | doc/Language/regexes.pod:
Merge pull request #474 from vadz/patch-2

Add missing matching "/" at the end of regex
BooK is there a better way to write @a».Int ? 14:29
sorry, @a = @a».Int
grondilu +<<@a
ah no
BooK because nodal, I remember what masak said some days ago :-) 14:30
perlpilot grondilu: Have you seen github.com/perlpilot/p6-Test-Class...ass.pm6#L4 ? I dunno if that's helpful or not as I haven't been paying attention.
perlpilot meeting &
grondilu m: my @a = (rand xx 3) X*10; @a [>>.]= Int;
camelia rakudo-moar 6cb231: OUTPUT«5===SORRY!5=== Error while compiling /tmp/esVRA6WKkR␤Malformed postfix call␤at /tmp/esVRA6WKkR:1␤------> 3my @a = (rand xx 3) X*10; @a [>>.7⏏5]= Int;␤»
_nadim psch: youin, that sucks. can't give a negative length, so you must do all the computations before; again and aga 14:31
psch: you can't give a negative length, so you must do all the computations before; again and again, that sucks.
psch m: say "negative length".substr(*-10, *-4)
camelia rakudo-moar 6cb231: OUTPUT«ive le␤»
14:31 remmie2 joined
grondilu m: my @a = (rand xx 3) X*10; @a [X.]= Int; 14:31
camelia rakudo-moar 6cb231: OUTPUT«5===SORRY!5=== Error while compiling /tmp/k9i9Bv2DD6␤Malformed postfix call␤at /tmp/k9i9Bv2DD6:1␤------> 3my @a = (rand xx 3) X*10; @a [X.7⏏5]= Int;␤»
psch _nadim: that doesn't do what you need? 14:32
grondilu m: my @a = (rand xx 3) X*10; @a X[.=] Int;
camelia rakudo-moar 6cb231: OUTPUT«5===SORRY!5=== Error while compiling /tmp/IvfvBwL_WZ␤Cannot cross with .= because dotty infix operators are too fiddly␤at /tmp/IvfvBwL_WZ:1␤------> 3my @a = (rand xx 3) X*10; @a X[.=]7⏏5 Int;␤»
grondilu gives up
_nadim m: say "negative length".substr(0, -4)
camelia rakudo-moar 6cb231: OUTPUT«Number of characters argument to substr out of range. Is: -4, should be in 0..Inf; use *-4 if you want to index relative to the end␤ in block <unit> at /tmp/TPuU6HoGeM line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/TPuU6HoGeM line 1␤␤»
psch m: say "negative length".substr(0, *-4)
camelia rakudo-moar 6cb231: OUTPUT«negative le␤»
psch _nadim: it even says "use *-4 if you want to index relative to the end"
14:33 loren left, tharkun left, huf left, aries_liuxueyang left, molaf left, uruwi left
RabidGravy SPLITTERS! 14:33
14:33 ribasushi joined
_nadim it's not relative to the en, or we'd be doing computation again, it's a length that is negative, big difference 14:33
psch _nadim: but a negative length is a semantically invalid concept. it doesn't make any sense 14:34
RabidGravy er
_nadim they said the same thing about the root of negative numbers ;)
psch yeah, no, it's not the same thing...
anyway, what's your use-case ooc?
grondilu m: module Foo { module Bar { our sub f is export { say "hi" } }; sub EXPORT { %(Bar::EXPORT::DEFAULT) } }; 14:35
camelia ( no output )
tony-o MadcapJake: ++ for myhtml
grondilu m: module Foo { module Bar { our sub f is export { say "hi" } }; sub EXPORT { %(Bar::EXPORT::DEFAULT) } }; import Foo; f()
camelia rakudo-moar 6cb231: OUTPUT«hi␤»
14:35 pierre_ left
grondilu perlpilot: oh that works 14:35
_nadim psch: well, often it is when you have a string to display and a certain width of display.
MadcapJake tony-o: thanks!
grondilu and it's fairly elegant as well
RabidGravy aah gah doo doo doo 14:36
tony-o just saw it as i'm checking the cron settings for modules.zef.pm
_nadim psch: I actually met the case earlier when patching Gramar::Tracer
psch _nadim: why not use comb for that?
instead of substr
RabidGravy or sprintf
14:37 huf joined, aries_liuxueyang joined, molaf joined, uruwi joined, richi235 joined, JimmyZ joined, vike joined, vytas joined, M-Illandan joined, gtodd joined, masak joined, [Coke] joined, chris2 joined
MadcapJake tony-o: 1.6 million downloads wooohoo! ;) 14:37
psch or that i suppose, yeah
14:37 loren joined
_nadim EG: $sniplen > 0 ?? colored(' ' ~ $snippet.substr(0, $sniplen).perl, 'white') !! '' 14:37
sprintf doesn't take negative length either, back to the board 14:39
as for comb, I need to check again what it accepts for arguments
This is _really_ not a big deal, just something Ihave seen a lot of times
tony-o MadcapJake: wait until you refresh :-p
14:39 daniel-s left
RabidGravy _nadim, 90% of programming is bounds checking ;-O 14:40
14:40 Juerd joined
grondilu perlpilot: trying this trick in one of my modules I got a spectacular fail: "Interrupted command". Never seen that before. 14:40
14:41 khw joined
grondilu probably on shell level since it was "commande interrompue" (in my language) 14:41
psch m: my $foo = lines()[0..2].join; $foo.comb(/. ** 10/)>>.say
camelia rakudo-moar 6cb231: OUTPUT«Céad slán ␤ag sléibht␤e maorga C␤hontae Dhú␤n na nGall␤Agus dhá c␤héad slán ␤ag an Eire␤agal ard i␤na stua os␤ cionn cao␤r is coll;␤Nuair a gh␤luais mise␤ thart le ␤Loch Dhún ␤Lúich’ go ␤ciúin sa g␤hleann ina␤»
14:41 ggherdov joined
RabidGravy yeah, that looks like something bit the dust under the hood 14:42
psch ...making up a regex that doesn't split words is a bit harder - i mostly don't remember the syntax for the anchor right now :P
grondilu m: module Foo { our class Bar { our multi infix:<*> is export { -1 } }; sub EXPORT { %(Bar::EXPORT::DEFAULT) } }; import Foo; say Foo::Bar*Foo::Bar 14:43
camelia rakudo-moar 6cb231: OUTPUT«5===SORRY!5=== Error while compiling /tmp/4sfINh6CVr␤Cannot use 'our' with individual multi candidates. Please declare an our-scoped proto instead␤at /tmp/4sfINh6CVr:1␤------> 3r { our multi infix:<*> is export { -1 }7⏏5 }; sub EXPORT { %(Ba…»
_nadim m: '1234567'.comb(-4 ,1)
camelia ( no output )
_nadim m: '1234567'.comb(-4 ,1).say
camelia rakudo-moar 6cb231: OUTPUT«(6)␤»
grondilu m: module Foo { our class Bar { multi infix:<*> is export { -1 } }; sub EXPORT { %(Bar::EXPORT::DEFAULT) } }; import Foo; say Foo::Bar*Foo::Bar
camelia rakudo-moar 6cb231: OUTPUT«Use of uninitialized value of type Foo::Bar in numeric context in block <unit> at /tmp/sSilg5F3QI line 1␤Use of uninitialized value of type Foo::Bar in numeric context in block <unit> at /tmp/sSilg5F3QI line 1␤0␤»
_nadim m: '1234567'.comb(-4 ).say 14:44
camelia rakudo-moar 6cb231: OUTPUT«(1 2 3 4 5 6 7)␤»
grondilu m: module Foo { our class Bar { multi infix:<*> is export { -1 } }; sub EXPORT { %(Bar::EXPORT::DEFAULT) } }; import Foo; say Foo::Bar.new*Foo::Bar.new
camelia rakudo-moar 6cb231: OUTPUT«Cannot call Numeric(Foo::Bar: ); none of these signatures match:␤ (Mu:U \v: *%_)␤ in block <unit> at /tmp/Bd7IbQAo5h line 1␤␤»
_nadim m: '1234567'.comb(-3 ).say
camelia rakudo-moar 6cb231: OUTPUT«(1 2 3 4 5 6 7)␤»
zwu I found there is a different behavior between the import and require: module m1 { our sub foo() {...}; our class TestClass {...}}; //// use m1; I can call m1::foo(); and my $tc = m1::TestClass.new(); //// however, when I use require "m1_filename"; I can call m1:foo() too but my $tc=m1::TestClass.new() gives me error of no symbole of &TestClass.
14:44 pierre_ joined
_nadim m: '1234567'.comb(-3, 2 ).say 14:44
camelia rakudo-moar 6cb231: OUTPUT«(5 6)␤»
gtodd has perl6 scaled enough to begin monetizing ? :-) 14:45
_nadim m: '1234567'.comb(3, 2 ).say
camelia rakudo-moar 6cb231: OUTPUT«(123 456)␤»
_nadim m: '1234567'.comb(-3, 0 ).say
camelia rakudo-moar 6cb231: OUTPUT«()␤»
BooK how do I see the canonical form of a Rat? (as numerator/denominator)
_nadim m: '1234567'.comb(3, 0 ).say
camelia rakudo-moar 6cb231: OUTPUT«()␤»
_nadim BooK: .perl does that I believe 14:46
timotimo BooK: you'd have to .nude.join("/") probably
BooK ah nude
I had tried with gist and perl
_nadim Data::Dump::Tree
it shows num/denum 14:47
14:47 rindolf joined
BooK thanks _nadim & timotimo 14:48
_nadim and "$r.numerator ~ '/' ~ $r.denominator" of course 14:49
zwu why does it say no symbole of &TestClass when using m1::TestClass.new(), for which m1 module is loaded by require "m1_filename.pm6", module m1 {our class TestClass ...} 14:50
14:50 jjido_ joined, lostinfog joined
gtodd _nadim: over time you can make all the .things() that do stuff with "strings" be perfectly consistent and better :-) 14:54
14:55 jjido_ left
RabidGravy ooh the domain "noisegang.com" is actually available 14:56
14:56 ZoffixWin joined
timotimo neato 14:57
ZoffixWin I'm looking at this: docs.perl6.org/language/5to6-perlfunc#exec .... is there no way at all to basically restart your own script?
I want Bailador to watch my source for changes and restart itself when it detects a change
_nadim yoohoo!!! putting \n* at a start of a grammar makes things happen! why? I need more time. 14:58
Xliff OK. I have a release branch in git. How do I go about adding that branch to the ecosystem?
perlpilot ZoffixWin: you could use IO::Notification for that.
ZoffixWin Xliff, click Pencil icon and add URL to *raw* meta file here: github.com/perl6/ecosystem/blob/ma.../META.list
perlpilot, thanks, but watching isn't a problem.. it's restarting that's a problem :) 14:59
RabidGravy Xliff, what's your GH id? I gonna invite you to a party
perlpilot oh
Xliff Xliff [email@hidden.address]
RabidGravy well actually a github organisation but the other thing is more fun
no, no github id 15:00
Xliff ZoffixWin, I've already cloned ecosystem and done that, but before I release, I want to know if I can specify a release branch, or if I will need to create another branch for further development.
psch Xliff: the links inside the META.list contain the branch 15:01
ZoffixWin Xliff, I think if you link to the meta from the branch, that'll work
RabidGravy yeah a release branch is fine as long as specify the correct thing in the META.list and the correct urls in the META6.json
_nadim feedback please, this is the modified output of Grammar::Trace, anything that could be done better? imgur.com/Wimuoso
gtodd Grammar::Trace or Grammar::Tracer ? 15:03
ZoffixWin exit shell($command); doesn't really do the trick... 'cause the socket is still in use when the second copy of bailador starts :/ 15:04
_nadim Grammar::Tracer, jnthn's module
Xliff So this should work as a source_url in META.info: git://github.com/Xliff/p6-audio-oggvorbis.git#v0.0.1 15:07
ZoffixWin That won't get indexed 15:08
hm
perlpilot ZoffixWin: when you figure this out, I expect a module to make it easier for the next guy and a blog post about your adventure ;-) 15:09
ufobat ZoffixWin++ :D
perlpilot ZoffixWin++ btw. Your blog posts have been excellent
Xliff Do panda and zef use the URL in META.list, or the one in META.info? 15:10
ufobat ZoffixWin, funny, yesterday i tried to kill my webserver with fast requests and it didnt happen as well
15:10 kaare_ joined 15:11 sno joined
gtodd _nadim: I remember being amazed by jnthn's presentation (from youtube ?) and thinking it would be neat if there were Grammar::Tracer plugins that would output the trace in different formats 15:11
ZoffixWin Xliff, I believe they use a JSON file generated by this script: github.com/perl6/ecosystem/blob/ma...atelist.pl
15:11 colomon left
gtodd _nadim: wow ... I just checked and that was a long time ago 8-} 15:12
ZoffixWin Xliff, and this part mangles source in META: github.com/perl6/ecosystem/blob/ma...ist.pl#L62
_nadim gtodd: It's a replacement of his format, IMVHO, clearer and much more compact. But maybe it is a good idea to have different formats
RabidGravy Xliff, yeah the META file in the META.list is retrieved and then the installer will use the source in that to retrieve the actual module source 15:13
perlpilot sethgodin.typepad.com/seths_blog/20...-gate.html seems like something #perl6 should be continually cognizant of 15:14
_nadim ZoffixWin's post are great, except for his aversion for parenthesis, he puts ' :' everywhere ;)
perlpilot _nadim: feature!
ZoffixWin Yes, I hate parentheses with passion :)
zwu hello, is it a bug? it says no symbol of &TestClass when calling m1::TestClass.new() for which m1 module is loaded by require "m1_filename.pm6", module m1 {our class TestClass ...} 15:15
perlpilot ZoffixWin: were you accosted by a LISP variant at some point? :)
Xliff OK. So the META.info has to match the release branch. Got it. 15:16
Thanks.
gtodd _nadim: I remember (again a long time ago) someone making a video of a web framework of some kind (maybe Django) which showed how a request was handled by the code ... highlighting the actual lines of code and showing how different modules were loaded and handed off the request and reponse etc. It was really well done but probably a lot of work. Then a few years later jnthn's module (tiny I think) Grammar::Tracer almost was able to do the sam
RabidGravy zwu, better show us the actual code as I'm pretty certain that works 15:18
ZoffixWin Xliff, well, I found just one module that doesn't use the default branch as the META file URL, but in their source, they still listed just the URL without branch: github.com/nkh/P6-Data-Dump-Tree/b.../META.info
I guess you get to be the pioneer to figure out whether this stuff actually works :) 15:19
psch the /release/ bit in the path *is* the branch, no?
15:19 vendethiel left
ZoffixWin psch, yes. But that's not part of the `source` URL in the META 15:19
psch ZoffixWin: ooh, right. i misunderstood which level the branch has to be tracked 15:20
_nadim gtodd: Grammar::Trace is very useful, I am trying to figure out why my rammar sucks right now ;), but I think it can be polished a bit more. Conway had a great regex visualizer, there are ideas to get from there.
RabidGravy ZoffixWin, that should work as long as the META file doesn't change differently in the branch 15:21
gtodd _nadim: exactly ... I was thinking a sort of typescript playback of the parse ... a deep dive into perl6's *brain* .... a plugin architecture for Grammar::Tracer of some kind might lead to something cool ... Grammar::Tracer::Plugin::Animated::MP4 ? :-D
ZoffixWin RabidGravy, but branch information from the META file URL is not preserved anywhere: ecosystem-api.p6c.org/projects.json 15:22
15:23 domidumont left
RabidGravy ah I see 15:23
yeah, that's not so useful then
_nadim gtodd: reminds me of hen I made movies of a build system, where you see the rulestrigger and the graph build up
15:23 colomon joined
ZoffixWin And seems that at least panda doesn't switch to any branches: github.com/tadzik/panda/blob/2beab...her.pm#L46 15:23
gtodd _nadim: wow cool ... btw I agree about ZoffixWin's posts ... I think they should have footnotes showing the parenthetcial alternative to ": " :)
zwu here is the test code I have gist.github.com/anonymous/35b0e607...b0513a0112 15:24
gtodd also there's too much blood :-D
ZoffixWin heh
gtodd ZoffixWin: ++ !
RabidGravy I'm kinda mixed about the parens on methods 15:25
15:25 colomon left
RabidGravy at first I thought the colons was ugly, now I've warmed to it for single argument methods 15:25
zwu RabidGravy: can you also take a look on the code at gist.github.com/anonymous/35b0e607...b0513a0112 15:26
tadzik hmm
psch i kind of want to make an effort to get used to indirect method invocation syntax :P
tadzik would you like panda to be able to checkout branches (or specific commits) when you do 'panda install git://foo.bar/baz.git@devel' or something?
_nadim The problem is that the examples are for beginers too and when you have a colum instead for a well delimited , parenthesis, scope and then you shuffle in adverbs, like Zoffix likes to do, then you need concentration to match things up. 15:27
gtodd: I'll try to find the movie I made, or maybe I could generated a new one.
15:28 colomon joined, ZoffixWin left
gtodd RabidGravy: it will take a while for P6BP to emerge :) 15:29
15:29 neuron left
gtodd with perl5 it's always educational to see how perl -MO=Deparse "sees" your script ... or how perlcritic would rewrite it :) 15:31
RabidGravy zwu, if you change the line to "my $tc = ::('m1::TestClass').new()" it will work, the reason it doesn't work how you have it is that the symbol isn't there at compile time 15:36
"require" being a runtime thing
zwu oh, that ::('m1::TestClass') is to by pass the compiler, but why the sub works m1::foo() ? 15:37
Xliff *Whew*
I'm off master and on to v0.0.2 15:38
Now to get v0.0.1 into the ecosystem.
zwu RabidGravy: I just tested the code ::('m1::TestClass').new(), it works.
RabidGravy zwu, well it's a "run time symbol lookup", the sub works because the sub isn't looked up until runtime when it's package qualified 15:39
TimToady jnthn: given that we overload \c[] for both \c65 and \c[LATIN CAPITAL LETTER A], I can't offhand think of a really good reason not to overload chr to take either chr(65) or chr("LATIN CAPITAL LETTER A") 15:40
15:40 g4 left 15:41 zakharyas left
RabidGravy and the reason it blows up the way it does is that rather than dying in compilation when it can't find the symbol it just evaluates to the Failure which is what you are seeing 15:41
15:41 _mg_ left
zwu ah... thanks RabidGravy. But why the m1::TestClass with packaged qualified does not the same at runtime as the sub? 15:41
15:41 maybekoo2 left
RabidGravy because types are handled differently 15:42
psch m: A.new; class A { } 15:43
camelia rakudo-moar 1acf80: OUTPUT«5===SORRY!5=== Error while compiling /tmp/71re6Rxjxj␤Illegally post-declared type:␤ A used at line 1␤␤»
psch m: ::('A').new; class A { }
camelia ( no output )
zwu thanks!
Xliff How often does raw.githubusercontent.com/ get updated? I just pushed a change on META.info to master and I don't see it reflected in the URL from ecosystem/META.list 15:44
RabidGravy I guess the compiler optimistically thinks that something may come along and actually provide the symbol but makes it a Failure that will happen at run-time if it doesn't
Xliff, every twenty minutes or so I think 15:45
Xliff Yep. Just saw it. Submitting PR now.
Aaand. Done. 15:47
RabidGravy ooh who killed dalek
dalek osystem: ca2ba7a | Xliff++ | META.list:
New Perl6 Module

Add Audio::OggVorbis to the perl6 ecosystem.
osystem: 2faa6a2 | RabidGravy++ | META.list:
Merge pull request #202 from Xliff/patch-1

New Perl6 Module
Xliff \o/
OK. I'm already late. I wanted to get this in before I had to leave. 15:49
See ya in the afternoon, #perl6.
jnthn m: say chr("65")
camelia rakudo-moar 1acf80: OUTPUT«A␤»
jnthn TimToady: Only thing is we'd break ^^ :)
TimToady: Unless we try numification as a fallback 15:50
15:50 cschwenz left 15:51 jjido_ joined
zwu what's the disadvantage is we change to the runtime lookup for types when they are with package qualified? 15:52
TimToady can just look at the first char to see if numeric; don't think any uninames start numeric
15:53 robertle left
TimToady m: .say if .uniname ~~ / ^ \d / for 0..0x10ffff 15:54
camelia ( no output )
TimToady would probably be a bit faster than a fallback 15:55
15:55 maybekoo2 joined
TimToady though of course we're depending upon the Unicode Consortium to continue following this (unwritten?) policy 15:55
15:55 jjido_ left, pecastro joined
RabidGravy zwu, well typically you should check the result of the lookup for being a Failure rather than it just blowing up if the symbol really isn't defined 15:56
people are doing that in modules when something is optionally loaded at run time 15:57
see e.g. github.com/sergot/http-useragent/b...t.pm6#L345 15:58
16:02 uruwi left 16:04 vendethiel joined
TimToady well, something in there is already making the decision to distinguish \c[65] from \c[LATIN CAPITAL LETTER A] 16:07
and we should use that mechanism, or emulate it
jnthn TimToady: OK, sounds good. :) 16:09
I'll see if I can sneak it in at some point...will be a nice break from hunting GC/concurrency issues. :)
RabidGravy snark hunting 16:10
16:12 nd3i joined
nd3i eli5? p6: say {my $s ="Camelia"; $s.subst(/<[aeiou]>/, { .uc }, :g)}() 16:13
timotimo sure 16:14
so, when you call the subst method on the string $s, it'll search for the regex
the :g at the end specifies that it's supposed to replace as often as it can, rather than stopping after the first change
nd3i Oops. I thought the bot would pick that up. I get: CAmElIA 16:15
timotimo right
it only reacts when the m: or p6: is at the very beginning; i think even a space character will prevent it from running
that setting has its pros and cons :)
nd3i But when I want to modify the string, I have to use the s/// form
p6: say {my $s ="Camelia"; $s ~~ s:g/<[aeiou]>/{ .uc }/; $s}() 16:16
camelia rakudo-moar 1acf80: OUTPUT«CCAMELIAmCAMELIAlCAMELIACAMELIA␤»
timotimo you can use S/.../.../ when your string is set to $_ (which the "given" statement-suffix does, for example)
ah, hehe
the smartmatch operator sets the value of $_ to what you match against, i.e. $s
so you're replacing every vowel with $s.uc 16:17
p6: say {my $s ="Camelia"; $s ~~ s:g/<[aeiou]>/{ $/.uc }/; $s}()
camelia rakudo-moar 1acf80: OUTPUT«CAmElIA␤»
nd3i so is there something I can use in the s/// form to call a method on the current match?
Ahh, u beat me to it! Thanks.
timotimo :) 16:18
you're welcome
BooK what's this signature? (Mu:U \v: *%_) 16:19
nd3i Ack! but $/ doesn't work in the method form
16:19 uruwi joined
BooK and where can I read up on signatures? 16:19
TimToady m: say {my $s ="Camelia"; $s.subst-mutate(/<[aeiou]>/, -> $_ { .uc }, :g); $s }()
camelia rakudo-moar 1acf80: OUTPUT«CAmElIA␤»
nd3i p6: say {my $s ="Camelia"; $s.subst(/<[aeiou]>/, { $/.uc }, :g)}()
camelia rakudo-moar 1acf80: OUTPUT«CAmElIA␤»
nd3i Oops. It works here but not in my REPL ... 16:20
Ok, it works for me in a new session
16:21 ZoffixWin joined
nd3i Maybe the $/ was left over from a previous line? 16:21
RabidGravy BooK, that is a method signature for a method that explcitly requires a type object and has no explicit arguments 16:22
TimToady BooK: Mu:U is a type that accepts any indefinite type object, \v is a sigil-less declaration of a "variable" named "v", the ":" means that's the invocant, and *%_ will slurp up any named arguments into the %_ array
RabidGravy it would like like "method whatever(Mu:U \v) { ... }" 16:23
and doc.perl6.org/type/Signature seems like a good place to read on the matter ;-) 16:24
it wouldn't like what I typed of course because it is missing a colon after the \v
16:25 loren left
ZoffixWin Looks like IO::Notification is buggy.... If I run this script, it prints all the directories it finds, but then says 'no such directory'. BUT, if I comment out that 0.1 second sleep on line 7, then the problem doesn't happen: gist.github.com/zoffixznet/e07d896...4675ff3aef 16:25
So it seems like creating a lot of IO::Notifications too fast is problematic 16:26
And if I use design-docs instead of . as the initial watch path, it crashes only about 15% of the time 16:28
16:29 maybekoo2 left
timotimo nd3i: you don't need $/ in the method variant, because the method will invoke your code-block with the match as its first (only?) positional parameter; in other words: you can just use $_ inside your block 16:30
ZoffixWin Yes, the Match object is the only arg 16:31
nd3i Ok, so I just have to remember that the s/// form and the method form handle a code block on the RHS differently 16:32
Sorry, on the replacement
ZoffixWin m: my $s ="Camelia"; say $s.subst: /<[aeiou]>/, *.uc, :g
camelia rakudo-moar 1acf80: OUTPUT«CAmElIA␤»
ZoffixWin m: say "Camelia".subst: /<[aeiou]>/, *.uc, :g 16:33
camelia rakudo-moar 1acf80: OUTPUT«CAmElIA␤»
ZoffixWin nd3i, BTW, the method form of s/// is .subst-mutate. .subst is the method form of S///
timotimo well, it's not truly the RHS of the s operator; it's more like the second half inside 16:34
at least that's how it feels to me
nd3i I think I'm at the "just enough knowlege to confuse myself" stage
timotimo the thing is, the right part of the s/// and S/// are regexy, not cody, so you have things like "the $/ has a value here" 16:35
RabidGravy nd3i, I'm permanently at that stage
nd3i ok, that makes sense; similar to p5
ZoffixWin m: say {my $s ="Camelia"; $s ~~ s:g[<[aeiou]>] = { $/.uc }() ; $s}()
camelia rakudo-moar 1acf80: OUTPUT«CAmElIA␤»
timotimo TimToady: did you have an opinion on what should be allowed to work in the listy part of an enum declaration? 16:36
16:37 rba_ left
nd3i right. I'm playing with the different ways to write it and getting different results. Sometimes different results from the same expression! 16:38
I'll study the suggestions here and see if the fog lifts any. TYVM 16:40
16:41 nd3i left
ZoffixWin .tell perlpilot sent you some fixes: github.com/perlpilot/p6-IO-Notific...ive/pull/1 There's still a problem with it: irclog.perlgeek.de/perl6/2016-04-29#i_12411080 16:42
yoleaux ZoffixWin: I'll pass your message to perlpilot.
mst ZoffixWin: parens are evil because hitting the shift key is hard, that's why I wrote a lisp that uses JSONY as its format so you can use [] 16:45
16:46 zwu_ joined
ZoffixWin heh 16:48
16:49 pierre_ left 16:52 jjido_ joined 16:53 zwu_ left 16:57 jjido_ left
RabidGravy there I can see the screen again 16:57
16:59 firstdayonthejob joined
ZoffixWin This seems to be in the wrong queue: rt.perl.org/Ticket/Display.html?id=128021 17:09
17:10 diakopter joined
RabidGravy seems fine to me, perl 5 ticket in the perl 5 queue 17:11
ZoffixWin Oh... brainfart. I just clicked on it in rt.perl.org/ forgetting the list includes both queues >_< 17:12
RabidGravy :)
zwu RabidGravy: thanks! 17:19
17:23 uruwi left 17:24 zwu left
[Coke] sees that someone donated to the perl QA hackathon 2016 as "the perl6 community". Thank you, whoever you are. (I probably know who you are. :P) 17:26
17:34 wamba left
stmuk_ two there are! 17:35
17:41 leont joined
MadcapJake ZoffixWin: you can change the dashboard to only show the Perl 6 queue (though it's slightly roundabout) 17:41
ZoffixWin head desks... 17:45
Can't figure out why my sub is called in duplicate :/
Any idea? gist.github.com/zoffixznet/50e0a7a...a521ed4f0a Line #32 prints and you can see there are two dirs, but the output from line #35 is repeated :S 17:46
It should 'Starting watch...' on each dir just once 17:47
vendethiel ZoffixWin what's |@dirs.map: { find-dirs $_ } ?
ZoffixWin vendethiel, it looks for each dir in @dirs to find all subdirectories 17:48
17:48 rba_ joined
vendethiel what's the result of that, I mean 17:49
ZoffixWin vendethiel, it's printed on line 32... the result is "DIRS: [lib|bin]"
Just two dirs, but I get 4 calls to watch-it :S
17:50 adu joined
RabidGravy there perl6-noise-gang.github.io/ - worst possible template I could have chosen but hey 17:50
please feel free to nit-pick, patch, fix or completely re-do 17:51
17:51 rba_ left, Relsak joined
adu RabidGravy: can I bootstrappify it? 17:52
17:52 jjido_ joined
RabidGravy feel free, I shouldn't be let anywhere near web pages 17:53
ZoffixWin Seems the supply {} block is executed twice :/
adu RabidGravy: do you know what bootstrappify means?
ZoffixWin I print "ZOMFG" just as I enter supply {} and it gets printed twice, even though the "CMON" I print right before it gets printed just once: gist.github.com/zoffixznet/66411b0...8f193dde04 17:54
adu, add it :) The current typography is hurting my eyes :)
adu getbootstrap.com/css/#buttons-options 17:55
when I think of bootstrap, I usually think of the buttons first :P
17:57 jjido_ left
RabidGravy oh I've done stuff with bootstrap, I tend to do so when I start something from scratch because I'm lazy, but GH presented me with a set of templates and I just chose the least offensive looking one ';-) 17:57
ZoffixWin, the block is executed every time the the supply is tapped, which is possibly what is happening there 17:59
ZoffixWin Yeah, I just noticed if I remove the await .Promise; the duplication goes away
But then... I'm not sure how to wait for it to be .done 18:00
18:06 rindolf left
ZoffixWin *sigh* a rogue loop {} at the end of the program it is.... 18:06
18:06 _mg_ joined 18:08 al2o3-cr joined
ZoffixWin Heh. Thanks to my program, I now know Atom does two writes whenever you save a file, for some reason. 18:09
perlpilot ZoffixWin++ thanks. 18:10
yoleaux 16:42Z <ZoffixWin> perlpilot: sent you some fixes: github.com/perlpilot/p6-IO-Notific...ive/pull/1 There's still a problem with it: irclog.perlgeek.de/perl6/2016-04-29#i_12411080
perlpilot ZoffixWin: I'm sure I had updated that module post-xmas, but I guess I never pushed the changes.
18:11 dwarring joined
ZoffixWin perlpilot, I think there's still a bug with that code. I'm using what I believe is the same code as in your module, but in my app, when I remove lib/x directory, I'm getting three events: "Change detected [lib/x/x, FileRenamed]. app" 18:14
"Change detected [lib/x/x, FileRenamed]. app
" "Change detected [lib/x, FileRenamed]. app
"
lib/x/x ?? :S Not even sure how it comes up with that
ZoffixWin is too tired to debug that too right now
perlpilot Did you make an issue for that?
I might have time tonight to look at it.
ZoffixWin No, I just found that
And it's with this code, not your module exactly: gist.github.com/zoffixznet/a726ea0...a59c3c1b61
perlpilot Hmm. ok 18:15
ZoffixWin Ah, I see. It just creates multiple watchers on the same dir. I'm gonna create an Issue with the description of the fix. 18:17
18:19 lostinfog left
perlpilot ZoffixWin: if you've got any ideas on how to fix the race condition in the conditional of that if statement, that would be nice too :) 18:21
18:21 rindolf joined 18:24 andreoss left
tadzik hm, I'm considering making :force a default when doing CUR.install() in panda and alikes 18:29
is there ever any harm done by this action, other than potentially wasting some time on it that wouldn't otherwise needs to be spent? 18:30
though it may de-motivate people from actually versioning their stuff properly
yeah, I'm not in favour of de-educating folks. Thanks for help, #perl6 :) 18:31
RabidGravy couldn't do what it does with the deps?
or something 18:32
tadzik hm?
RabidGravy only even try to install if the version you got is newer than an installed on?
one? 18:33
tadzik it does that, yes
18:33 uruwi joined
tadzik but I found that in module development I often go through a cycle of "install this stuff for me please even though I'm not ready to bump the version number yet" 18:33
or "I'd rather not bump the version number every time I make a change" 18:34
so now I always do `redpanda -du .`, but it feels like there's a minor annoyance here somewhere that's easily solvable with software :)
perhaps tools should default to "yes, really update regardless of the version" when you give it a location (path, git url etc) rather than a module name 18:35
since you're likely to know what you're doing and this is exactly what you mean
18:39 AlexDaniel joined
dalek osystem: cdb74b0 | (Zoffix Znet)++ | META.list:
Add Bailador::Dev::AutoRestarter

script to detect source code changes and auto-restart Bailador app: github.com/zoffixznet/perl-Bailado...oRestarter		 18:40
AlexDaniel ZoffixWin: not sure if you got an answer to your thingy but
m: ‘foobar’ ~~ /(foo(bar))/; say $0; say $0[0]
camelia rakudo-moar 1acf80: OUTPUT«｢foobar｣␤ 0 => ｢bar｣␤｢bar｣␤»
AlexDaniel ZoffixWin: I guess that ↑ is what you wanted
18:40 pochi joined
tadzik ermagherd 18:41
AlexDaniel m: ‘foobar’ ~~ /(foo(bar))/; say ~$0; say ~$0[0]
camelia rakudo-moar 1acf80: OUTPUT«foobar␤bar␤»
tadzik ZoffixWin: that's amazing!
psch m: ‘foobar’ ~~ /(foo(bar))/; .say for @0
camelia rakudo-moar 1acf80: OUTPUT«｢bar｣␤»
psch aw
ZoffixWin AlexDaniel, thanks
psch well, only slightly misleading, isn't it
m: ‘foobar’ ~~ /(foo(bar))/; put $0; put @0 18:42
camelia rakudo-moar 1acf80: OUTPUT«foobar␤bar␤»
AlexDaniel what is misleading?
psch that a nested capture is the first multi-capture
m: 'foobar' ~~ /(...) ** 2/; say @0
camelia rakudo-moar 1acf80: OUTPUT«[｢foo｣ ｢bar｣]␤»
psch AlexDaniel: the overload between @0 meaning "the first element of the first capture" as well as "the first capture that captures more than once" feels weird, is my point 18:43
AlexDaniel oh 18:44
indeed
psch m: 'foobar' ~~ /(...) ** 2/; say @() 18:45
camelia rakudo-moar 1acf80: OUTPUT«([｢foo｣ ｢bar｣])␤»
psch m: 'foobar' ~~ /(...) ** 2/; say $()
camelia rakudo-moar 1acf80: OUTPUT«foobar␤»
psch well, those are all magical i suppose... vOv
AlexDaniel whoa… 18:46
TimToady m: say "bicycle race".substr(2, -1) 18:49
camelia rakudo-moar 1acf80: OUTPUT«Number of characters argument to substr out of range. Is: -1, should be in 0..Inf; use *-1 if you want to index relative to the end␤ in block <unit> at /tmp/HvO12Bmtpt line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/HvO12Bmtpt line 1␤␤»
TimToady m: say "bicycle race".substr(2, -1) // ''
camelia rakudo-moar 1acf80: OUTPUT«␤»
TimToady _nadim: why not just do that? ^^^
18:57 ufobat left
dwarring m: <🚲 🚲 🚲>.race 19:00
camelia ( no output )
ZoffixWin .u 🚲 19:01
yoleaux U+1F6B2 BICYCLE [So] (🚲)
ZoffixWin heh
m: "$_ won".say and exit for <🚲1 🚲2 🚲3>.race: :1batch 19:03
camelia rakudo-moar 1acf80: OUTPUT«🚲1 won␤🚲2 won␤»
19:03 ufobat joined
literal class Foo { constant BAR = 3; }; my $foo = Foo.new; how do I get to BAR from $foo? 19:04
ZoffixWin m: class Foo { constant BAR = 3; }; my $foo = Foo.new; say Foo::BAR 19:05
camelia rakudo-moar 1acf80: OUTPUT«3␤»
literal ZoffixWin: I mean without hardcoding the class name
$foo.some.method.chain.BAR
ugexe m: class Foo { constant BAR = 3; method BAR { BAR }; }; Foo.new.BAR.say 19:06
camelia rakudo-moar 1acf80: OUTPUT«3␤»
ZoffixWin ^name gives Foo in this case.. um... unsure how to get bar from that
m: class Foo { constant BAR = 3; }; my $foo = Foo.new; say "$foo.^name()::BAR".EVAL 19:07
camelia rakudo-moar 1acf80: OUTPUT«3␤»
ZoffixWin heh
m: class Foo { constant BAR = 3; }; my $foo = Foo.new; say ::('::')
camelia rakudo-moar 1acf80: OUTPUT«Start argument to substr out of range. Is: 1, should be in 0..0; use *1 if you want to index relative to the end␤ in block <unit> at /tmp/cktr011TA1 line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/cktr011TA1 line 1␤␤»
ZoffixWin giggles 19:08
TimToady timotimo: haven't looked at that code in a while, but I expect some of the original enum parse/actions hackery might be interfering with the more general case, which should be able to define pretty much any terms/values from pairs, modulo semantic constraints imposed by the enum type
timotimo OK, cool :) 19:14
19:15 rindolf left
dalek osystem: 9c51126 | (Zoffix Znet)++ | META.list:
Move Audio::MIDI::Note to Perl6-Noise-Gang
 19:21
timotimo what'll the noise gang distribution of rakudo be called? 19:22
ZoffixWin Racket :) 19:24
jnthn Rockudo :P
ZoffixWin jnthn++
19:25 uruwi left
timotimo damn, that's good 19:25
literal is it possible to topicalize a parameter in method's signature? 19:26
jnthn literal: Just name a parameter $_ ?
psch m: sub f($_) { .say }; f 2
camelia rakudo-moar 1acf80: OUTPUT«2␤»
literal ah
19:27 jjido_ joined
stmuk_ whose going to be a Rockudo Star? 19:29
ZoffixWin :D
19:29 yqt joined
ugexe a 10x Rockudo Star 19:32
19:32 AlexDaniel left
stmuk_ full stack! 19:32
19:34 lostinfog joined
ZoffixWin FYI: Yapc::eu 2016 <amalia> And we are also lauching the call for speakers medium.com/yapc-europe-2016/call-f....kspsclkzo 19:37
19:40 jjido_ left
ufobat strange world, my wive asked me to work on bailador because she has a bad conscience 19:52
ZoffixWin :)
ufobat ZoffixWin, i like your restart script a lot :) 19:54
19:54 spider-mario joined
ZoffixWin Thanks. 19:54
looks like panda doesn't know how to --force install scripts. The install succeeds, but the script doesn't update :/ 20:05
20:07 maybekoo2 joined
_nadim TimToady: That's definitely a way! 20:07
20:08 dolmen joined
_nadim TimToady: what surprises me with >> say "bicycle race".substr(2, -1) // '' << is that the warning is gone! 20:12
20:12 user3 joined
tadzik ZoffixWin: please tell me more 20:13
user3 at this time, how much slower is the perl6 rakudo compiler compared to perl5?
for general processing, and in particular for regex
2x 10x slower/ 20:14
?
psch m: my $x = "foo".substr(1, -1); say "ok"
camelia rakudo-moar 1acf80: OUTPUT«ok␤»
tadzik ZoffixWin: the script handling in general is quite sketchy in CURs :/
ZoffixWin tadzik, you know the wrapper script that's generated when you install scripts inside the dist? Here's my modified one. gist.github.com/zoffixznet/d92b530...61bd0720f1
psch _nadim: the error "disappears" because substr returns a Failure that only complains when it's evaluated
ZoffixWin tadzik, I had to add sort { $^b<ver> cmp $^a<ver> }, on line 10 for the wrapper to actually pick the latest version
psch m: my $x = "foo".substr(1, -1); say $x.WHAT 20:15
camelia rakudo-moar 1acf80: OUTPUT«(Failure)␤»
psch _nadim: ^^^
tadzik ZoffixWin: hm, interesting
ZoffixWin And I have like a dozen version installed, because I did --force install several times. 20:16
tadzik hehe
BooK when I put a class in a module, should I do unit module Foo; class Foo { ... } ?
20:16 darutoko left
tadzik so they did get installed, just that the generated script was broken? 20:16
BooK and then use Foo; in the code needing it? 20:17
ZoffixWin BooK, you can. unit class Foo; is also a thing
tadzik, yes
BooK I have some private functions and variables
_nadim psch: OHHH I like it!
tadzik oh noes 20:20
ZoffixWin m: sub defuse {so $^a}; my $x = "foo".substr(1, -1); defuse $x; say $x.words[1..4].join: ' '; say "meow"
camelia rakudo-moar 1acf80: OUTPUT«Number of characters argument␤meow␤»
ZoffixWin :)
tadzik ZoffixWin: speaking of 'install --force', may I interest you in being a betatest of what will soon [tm] become part of the new, big, major panda update? :)
ZoffixWin not really lol
I started working on perl6.party at 9AM... it's 4:21PM now and I still don't have anything ready. Have been messing with bailador autorestart script lol 20:21
perl6.party
tadzik hehe 20:22
ZoffixWin Cool, hexchat recognizes it as a link
20:25 Praise joined, Praise left, Praise joined 20:28 remmie2 left 20:30 remmie joined
timotimo well, if you put http:// in front ... 20:32
20:32 _mg_ left
timotimo oh 20:32
ZoffixWin :)
timotimo ... never mind
20:33 user3 left
dalek c: cd3dd6d | (Wenzel P. P. Peppmeyer)++ | doc/Language/typesystem.pod:
fix enum return type in example
 20:33
20:33 rurban joined 20:34 Praise left
masak hm, is github.com/perl6/perl6.org/commit/e64aafc918 somthing I should do something about? 20:36
20:36 Praise joined
masak I recall producing my feed.atom with quite some care, using a validator... 20:37
but is there an advantage (for me or others) to have that `type="text/html"` on the `<link>` ?
20:38 rurban left 20:41 Ven joined
RabidGravy anyone else interested in being press ganged into perl 6 noise gang? 20:45
ZoffixWin masak, that attribute declares the type of content the link points to. It's up to the tool parsing it to make use of that information. 20:48
20:48 adu left 20:49 uruwi joined 20:50 sufrostico left
ZoffixWin BTW, did anything happen to the whole make Bailador auto-serve static files from a specific dir thing? 20:50
I recall there was a conversation
masak ZoffixWin: my question was more -- why was the expectation of `type="text/html"` there in the first place? is it a common expectation? can I read up on this somewhere? :)
ZoffixWin masak, ah, it was there because there are multiple <link>s in the XML it's parsing, so adding the extra restriction let it avoid wrong links. 20:51
20:51 zwu joined
ZoffixWin As for reading, the spec I guess: www.w3.org/TR/2011/WD-html5-author...-link-type 20:52
"The type attribute gives the MIME type of the linked resource. It is purely advisory"
ufobat whats the difference between my sub something and sub something ?
masak ufobat: none 20:53
ZoffixWin I don't think there is any. It's `my` by default.
masak ufobat: a sub is lexical by default
ZoffixWin m: class Foo { our sub bar { say 'bar' }; sub cow { say 'cow' }; my sub meow { say 'meow' } }; Foo::bar; Foo::cow; Foo::meow
camelia rakudo-moar 1acf80: OUTPUT«bar␤Could not find symbol '&cow'␤ in block <unit> at /tmp/Q2GwJtI1Sh line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/Q2GwJtI1Sh line 1␤␤»
ZoffixWin m: class Foo { our sub bar { say 'bar' }; sub cow { say 'cow' }; my sub meow { say 'meow' } }; Foo::bar; Foo::meow
camelia rakudo-moar 1acf80: OUTPUT«bar␤Could not find symbol '&meow'␤ in block <unit> at /tmp/FMSFPqLVyC line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/FMSFPqLVyC line 1␤␤»
ufobat good! thanks! 20:54
masak ZoffixWin: ok, so this was only on the level of writing a query that found all the right <link> elements. got it.
ZoffixWin Yeah
zwu Is the S11-modules.pod? Because the example code use Sense <&common @horse>; expecting a positional parameter to export sub instead?
I mean if the S11-moudules.pod documentation is outdated? 20:55
21:02 skids left
ufobat oh yeah! bailador almost runs on HTTP::Server::Tiny 21:03
ZoffixWin How do I use this %?RESOURCES thing? I'm dumping it, but it's empty 21:07
21:09 TEttinger joined, Ven left 21:10 _28_ria left
ZoffixWin OK. The issue was resource/resources. The spec says "resource" directory but it's actually looking in resources 21:10
21:10 sufrostico joined 21:18 grondilu left
ZoffixWin Does Perl 6 has some sort of construct to ensure a path is *inside* a directory? I thought I saw an example. e.g., "foo/bar" is inside, but "foo/../../../bar" is not. You know what I mean? 21:19
I need some safe way to open a file
perlpilot $*SPEC might have a method for that 21:21
dalek ar: 87c3a4a | (Steve Mynott)++ | / (2 files):
fix PATH for make modules-install under windows
 21:22
21:23 dolmen left
ZoffixWin Don't see one. 21:23
I guess I should splitpath filter out any updir/curdir and then join back together
perlpilot ZoffixWin: It still sounds like something that should exist in IO::Spec. Maybe get lizmat to add it :) 21:26
ZoffixWin m: say $*SPEC.catdir: $*SPEC.splitdir('foo/../../bar/foo.txt').grep({ $_ ~~ $*SPEC.curupdir }) 21:27
camelia rakudo-moar 1acf80: OUTPUT«foo/bar/foo.txt␤»
ZoffixWin uggly
ugexe m: my $path = "foo/../bar".IO; say $path.resolve; say $path.resolve.starts-with('bar') 21:31
camelia rakudo-moar 1acf80: OUTPUT«"/home/camelia/bar".IO␤False␤»
ugexe somethin like that anyway 21:32
then a good ol `use MONKEY-TYPING; augment IO::Path { method contains-path($path) { <check code> };` 21:33
ZoffixWin ugexe++ those don't seem to be documented
ugexe theres a warning in resolve though: "# XXXX: Not portable yet; assumes POSIX semantics" 21:34
(and contains-path is not an original method, thats just a way to add it to your IO::Path objects) 21:40
ufobat good night :) 21:42
21:42 Relsak left 21:44 ufobat left 21:51 sufrostico left 21:52 Sgeo joined 21:59 ZoffixLappy joined 22:00 hotel_california joined 22:01 nchambers joined, hotel left 22:03 sufrostico joined
dalek osystem: 6e9e9f1 | (Zoffix Znet)++ | META.list:
Fix incorrect repo name
 22:03
22:08 kid51 joined 22:12 grondilu joined 22:16 skids joined
ZoffixLappy eyes github.com/ufobat/Bailador/blob/ma...ticFile.pm 22:17
I wonder what's that for... Not used anywhere.
22:21 tomboy64 joined
ZoffixLappy wtf... we *still* can't serve binary files with Bailador? 22:21
22:23 perlawhirl joined
perlawhirl RabidGravy: the link to 'module list' is borked 22:23
22:24 lostinfog left
ZoffixLappy is baffled 22:26
22:27 RabidGravy left
ZoffixLappy Ugh.. good god: github.com/supernovus/perl6-psgi/pull/7/files 22:30
literal maybe a silly question...how do I flatten a Pair? 22:34
or an entire hash of them
ugexe lol
github.com/davepagurek/perl6-psgi/...m6#L21-L24
then later does $output ~~ Buf 22:35
ZoffixLappy literal, flatten into what? The answer is likely |
(which is a slip)
literal ZoffixLappy: to go from {1 => 2, 3 => 4} to (1,2,3,4)
ugexe so no matter what $output gets assigned a string value, so will never ~~ Buf
perlawhirl .kv
m: say (this => 'that').kv 22:36
camelia rakudo-moar 1acf80: OUTPUT«(this that)␤»
literal ah, of course
ZoffixLappy m: my $p = {1 => 2, 3 => 4}; say ['a', |$p.kv]
camelia rakudo-moar 1acf80: OUTPUT«[a 1 2 3 4]␤»
ZoffixLappy ugexe, yeah, and it needs to be a Buf when serving binary files. I opened an Issue saying it needs to accept Buf but the "fix" changed it to Stringy :/ 22:38
22:38 tomboy64 left
ugexe Stringy can take a buf. the problem is even if it *is* a buf at initialization of $output, its immediately turned into a string in the following conditional by assigning a string no matter what 22:38
ZoffixLappy hm 22:39
ugexe m: my Stringy $output = Buf.new(1,2,3,4); say $output.WHAT; $outout = "asdf"; say $output.WHAT 22:40
camelia rakudo-moar 1acf80: OUTPUT«5===SORRY!5=== Error while compiling /tmp/NHD3uiZGMS␤Variable '$outout' is not declared. Did you mean '$output'?␤at /tmp/NHD3uiZGMS:1␤------> 3t = Buf.new(1,2,3,4); say $output.WHAT; 7⏏5$outout = "asdf"; say $output.WHAT␤»
ugexe m: my Stringy $output = Buf.new(1,2,3,4); say $output.WHAT; $output = "asdf"; say $output.WHAT
camelia rakudo-moar 1acf80: OUTPUT«(Buf)␤(Str)␤»
22:40 tomboy64 joined
ZoffixLappy Hm. I'm getting "Type check failed in assignment to $output; expected Str but got Buf (Buf.new(72,84,84,80,4...)" 22:40
huh.. on this line: github.com/supernovus/perl6-psgi/b...GI.pm6#L47 22:42
ZoffixLappy starts to suspect maybe the installed version isn't the latest for some reason
22:44 leont left
ZoffixLappy *sigh* Gonna nuke everything and try to use zef instead of panda 22:44
ugexe i dunno, from what i can tell that code should not work properly 22:45
ZoffixLappy BTW, ISTR someone asking if there was a difference between a sub and pointy block. There's this: 22:47
m: sub {return}()
camelia ( no output )
ZoffixLappy m: -> {return}()
camelia rakudo-moar 1acf80: OUTPUT«Attempt to return outside of any Routine␤ in block <unit> at /tmp/oxwDvtWBYK line 1␤␤»
22:56 zwu left, perlawhirl left, perlawhirl joined 22:58 Herby_ joined
Herby_ Afternoon, everyone! 22:58
o/
ZoffixLappy \o 22:59
perlawhirl :q 23:00
lol
23:00 perlawhirl left 23:01 firstdayonthejob left
ZoffixLappy rakudobrew build zef doesn't work 23:08
Well.. it installs zef, but when I run it I get "===SORRY!=== Failed to open file /home/zoffix/.perl6/precomp/27E0B8A90A583776A17689697FE506415BB69CBE.1461970578.11573/F9/F94CA4E91B2AF324B8A925E1F065F40258AE4D90.deps: no such file or directory" 23:09
ZoffixLappy is exhausted from all the issues experienced today
ugexe did you `rakudobrew rehash` first?
otherwise you probably have to delete .perl6/precomp for some reason 23:10
23:11 dolmen joined
ZoffixLappy ugexe++ that did the trick 23:12
grondilu m: use MONKEY-SEE-NO-EVAL; (EVAL '-> ($x, $y) {}')(); 23:14
camelia rakudo-moar 1acf80: OUTPUT«Too few positionals passed; expected 1 argument but got 0␤ in block at EVAL_0 line 1␤ in block <unit> at /tmp/fYTPaOL2tt line 1␤␤»
grondilu m: use MONKEY-SEE-NO-EVAL; (EVAL '-> ($x, $y) {}')(1, 2); 23:15
camelia rakudo-moar 1acf80: OUTPUT«Too few positionals passed; expected 2 arguments but got 0 in sub-signature␤ in block at EVAL_0 line 1␤ in block <unit> at /tmp/akFOn5UOKA line 1␤␤»
grondilu ^not sure why that does not work
m: use MONKEY-SEE-NO-EVAL; (EVAL 'sub ($x, $y) {}')(1, 2);
camelia ( no output )
grondilu meh
23:16 wamba joined
ZoffixLappy m: my $x = -> ($x, $y) {}; $x(1, 2) 23:18
camelia rakudo-moar 1acf80: OUTPUT«Too few positionals passed; expected 2 arguments but got 0 in sub-signature␤ in block <unit> at /tmp/1PwAuRxrkO line 1␤␤»
ZoffixLappy shrugs
Ahhh
m: use MONKEY-SEE-NO-EVAL; (EVAL '-> $x, $y {}')(1, 2); 23:19
camelia ( no output )
ZoffixLappy grondilu, blocks don't need parens. And if you include them, it means destructuring
m: use MONKEY-SEE-NO-EVAL; (EVAL '-> ($x, $y) {}')([1, 2]);
camelia ( no output )
grondilu oh, good to know. 23:20
I managed to successfully generate code for my Clifford module :) 23:27
github.com/grondilu/clifford/blob/...imized.pm6
ZoffixLappy \o/
grondilu now I'm anxious about testing it though. Kind of scared of finding out it's all crap :P
geekosaur it's all crap, yyou will realize this in about 6 months. (approximately a universal truth :p ) 23:30
23:30 dolmen left
ZoffixLappy :) 23:30
TimToady soulds about 90% true :) 23:32
*sounds
arnsholt Definitely. These last few days, I've had to look at code I haven't touched since December 2012 and early 2013, and I've had much the same reaction 23:33
23:33 zwu joined
grondilu unfortunatly it's something that is often said about perl. 23:35
but I do believe it's only a matter of writing sensible code, not about the language itself.
23:35 hotel_california left, hotel joined
grondilu First drafts are always fugly. 23:36
TimToady well, it's true about everything, if you believe Sturgeon.
ZoffixLappy nonsense. I have a ~100,000 LOC app at work with zero documentation, but I manage it just fine
TimToady it's in assembly, right?
ZoffixLappy I hate the code. I wish I had the time to rewrite it. But it's not line noise
Nah, Perl 5
geekosaur remembers maintaining about 20,000 LOC of m4 macros around a 4GL, and another 30,000 or so LOC written using those macros 23:37
ZoffixLappy rages 23:39
Well, reinstalling everything fixed my issue with the Buf vs Str stuff, but now it seems content_type 'whatever' doesn't work in Baildator ~_~ my PNG is getting served as text/html :/
grondilu m: say class A::B::C {}.new.WHAT # a bit surprised about the shortening here 23:44
camelia rakudo-moar 1acf80: OUTPUT«(C)␤»
ZoffixLappy m: say class A::B::C {}.new.^name
camelia rakudo-moar 1acf80: OUTPUT«A::B::C␤»
grondilu oh ok 23:47
ZoffixLappy Aw, cmon... we have arrays, people :( github.com/ufobat/Bailador/blob/a7...nse.pm#L22 23:48
Well, this app doesn't set Content-Type and calling content_type doesn't set content type :( gist.github.com/zoffixznet/cd07431...465beee7c4 23:49
I'm really bummed I wasted about ~2 hours so far on crap errors and issue 23:50

Powered by Cro and the Raku® Programming Language.

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

Thank you!