»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.
Xliff CRAP 00:02
I knew there was a reason I wasn't working on that module.
There was a recent change in rakudo (end of summer, maybe) that introduced that.
It wasn't there in July, and the XML::LibXML codebase used it extensively. 00:03
MasterDuke btw, don't know if you just typed something wrong, but the line i commented out was "use XML::LibXML::CStructs;" without the ":types" 00:04
(obviously who cares if it's commented out, but just in case you're on a branch or something"
00:05 raiph left
Xliff OK. 00:05
in XML::LibXML dir, "git checkout -b 15nodelist"
Then pull new changes. 00:06
00:06 giraffe left
MasterDuke ===SORRY!=== Expected MAST::Frame, but didn't get one 00:07
Xliff YES!!
Er.... Um...
Youknowwhuddimean
So, there could be something screwy with either the XML::LibXML codebase or XML::XSLT 00:08
I just don't know.
and XML::LibXML needs serious overhaul due to bitrot
00:13 espadrine left 00:17 pmurias left
MasterDuke Xliff: well, so far golfed it to this: gist.github.com/MasterDuke17/a2c8e...5c90220441 00:21
maxp_ is there some kind of problems about capturing groups and ratchet adverbs? 00:23
I've this:"<ciao>mamma</ciao>" ~~ /:r '<' (<[\w\W]>+?) '>'/ 00:24
and I expect the word "ciao" but it returns nil
whitout parentesis it works 00:25
AlexDaniel bisect: "<ciao>mamma</ciao>" ~~ /:r '<' (<[\w\W]>+?) '>'/ 00:26
bisectable6 AlexDaniel, On both starting points (old=2015.12 new=29b228c) the exit code is 0 and the output is identical as well
AlexDaniel, Output on both points:
AlexDaniel ah, right
ok this is the simplified version 00:27
m: "<ciao>mamma</ciao>" ~~ /:r '<' (.+?) '>'/
camelia ( no output )
00:27 giraffe joined
MasterDuke Xliff: actually, you can even remove "use XML::LibXML::Parser;", but you do still need to include "p6-XML-LibXML/lib/" in your -I 00:27
AlexDaniel m: say "<ciao>mamma</ciao>" ~~ /:r '<' (.+?) '>'/
camelia rakudo-moar 29b228: OUTPUT«Nil␤»
AlexDaniel m: say "<ciao>mamma</ciao>" ~~ /:r '<' .+? '>'/
camelia rakudo-moar 29b228: OUTPUT«｢<ciao>｣␤»
AlexDaniel in fact, it does not have to capture 00:28
m: say "<ciao>mamma</ciao>" ~~ /:r '<' [.+?] '>'/
camelia rakudo-moar 29b228: OUTPUT«｢<ciao>｣␤»
AlexDaniel I think this is part of RT #130117 ?
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=130117
MasterDuke maybe rt.perl.org/Ticket/Display.html?id=123934 also 00:29
Xliff Huh!
AlexDaniel maxp_: while there is probably a bug, I think it is wrong to expect <ciao> here 00:30
Xliff MasterDuke: So it is something in XML::XSLT::Subs
maxp_ ok thanks
AlexDaniel m: say "<ciao>mamma</ciao>" ~~ /:r '<' (<-[>]>+) '>'/
camelia rakudo-moar 29b228: OUTPUT«｢<ciao>｣␤ 0 => ｢ciao｣␤»
AlexDaniel maxp_: ↑ that's how you get it :)
Xliff Because I have the 2 sub definitions commented out in my current version.
maxp_ ok thanks again
Xliff And most of XML::XSLT::Subs is commented out as well. 00:31
At least here
samcv more uniprops work in new PR :) will have 'Name' 'Titlecase_Mapping' 'Lowercase_Mapping' 'Uppercase_Mapping' and 'Numeric_Value' working
i think something has to happen in Moar though for Name1 the unicode version 1 name to work though, and probably a bunch of other ones
there's a lot of properties
MasterDuke Xliff: MAST::Frame errors may be above my paygrade, but at least it's been golfed a bit 00:32
Xliff Yes. Thank you so much for looking into it, MasterDuke++ 00:33
So it looks like it is something with XML::LibXML and XML::XSLT
But now I am more leaning toward XML::LibXML
Which... again... bitrot 00:34
*sigh*
00:37 maxp_ left
Xliff Yup. After that bit of code culling, it looks like it is XML::LibXML and not anything with XML::XSLT at all! 00:37
wow
00:38 cfa left 00:43 aborazmeh joined, aborazmeh left, aborazmeh joined 00:57 BenGoldberg joined 00:58 aborazmeh left 01:00 captain-adequate left 01:04 canopus left
samcv ok so there are unicode properties to find the matching bracket Bidi_Paired_Bracket is the thing 01:07
and atm it isn't supported by MoarVM or rakudo
also:
For legacy reasons, the characters U+FD3E ORNATE LEFT PARENTHESIS and
# U+FD3F ORNATE RIGHT PARENTHESIS do not mirror in bidirectional display
# and therefore do not form a bracket pair.
also idk if that means we should remove it from our list. but i think it was used in 6.c ... 01:08
but i doubt anybody has actually used it
AlexDaniel u: U+FD3F
unicodable6 AlexDaniel, U+FD3F ORNATE RIGHT PARENTHESIS [Ps] (﴿)
samcv and there's a bracket property where you can find out if it's open or close
phatbot u: U+FD3E
unicodable6 phatbot, U+FD3E ORNATE LEFT PARENTHESIS [Pe] (﴾)
samcv ftp://ftp.unicode.org/Public/UNIDATA/BidiBrackets.txt this is the file 01:09
let me check 6.c tests...
i remember it faliing when i switched the order of the ornate pars in nqp
AlexDaniel commit: all say Q﴾hello﴿
committable6 AlexDaniel, gist.github.com/ae5fddbcc12cba4451...0e14842b7a 01:10
AlexDaniel hehehe
samcv :(
i think we should really only support that list i linked to... 01:11
let me check roast though
01:12 canopus joined
samcv yeah.. it's in roast 01:12
phatbot Then it'd have to wait until 6.d 01:13
huggable: 6.d
huggable phatbot, Proposals for 6.d language: github.com/perl6/specs/blob/master/v6d.pod
phatbot ^ stick it in there, if it's worth it.
samcv u: “ 01:14
unicodable6 samcv, U+201C LEFT DOUBLE QUOTATION MARK [Pi] (“)
AlexDaniel it's something to think about
samcv yeah
the quotation marks aren't listed in that unicode properties but also weird that the closing and opening don't match for those ornate parens either 01:15
SmokeMachine I used my class (that instantiated a singleton) on a module, then I used the same class on my main program and tried to use the singleton... it got a different object... how should I do that? 01:17
phatbot No idea 01:21
phatbot frequently reaches to .try
Anyone else the same?
'fasdfsd'.EVAL.try; ....
probably doesn't make any sense :}
01:24 mawkish_ left
samcv and we will def need the BIDI mirrored glyphs thing if we want to do RTL text well 01:28
i'm looking up what it would take to support the Unicode Bidirectional spec
samcv puts on her heavy duty gear for reading this spec 01:29
01:29 aborazmeh joined, aborazmeh left, aborazmeh joined
samcv there's a BIDI unicode spec test included with unicode sa well... so someday would be good if we can pass it 01:29
phatbot www.youtube.com/watch?v=LNKocMiehZA 01:30
samcv i think the spec is most important for DISPLAYING RTL text.. but I will see what we need in all of our levels to support that kind of thing 01:31
01:32 unicodable6 left
samcv but we need to have some way in perl 6 to be able to figure out the order such characters should be displayed in 01:32
However, because these right-to-left scripts use digits that are written from left to right, the text is actually bidirectional
01:32 unicodable6 joined
samcv lots to consider 01:32
[Coke] docs - should the perl5-style USAGEs in docs follow the 'space after comma' rule? 01:34
01:35 unicodable6 left
phatbot Yes, but I'm biased since I always write that way :D 01:35
01:35 unicodable6 joined
AlexDaniel what should whateverables do when gisting something that contains control characters? 01:36
dalek c: 7dafc5f | coke++ | xt/words.pws:
learn new function param names
 01:38
AlexDaniel MasterDuke: ↑ ? 01:40
01:40 pyrimidine joined 01:43 aindilis left 01:44 pyrimidine left
samcv ok 'Bidi_Mirroring_Glyph', atm nqp::getuniprop_str returns a string, which is the hex codepoint of the mirrored glyph 01:45
i'm thinking probably Moar is storing it as a string not a number?
how do we want this to work? have nqp::getuniprop_int return the number of the codepoint?
01:46 travis-ci joined
travis-ci Doc build failed. Will "Coke" Coleda 'learn new function param names' 01:46
travis-ci.org/perl6/doc/builds/183459374 github.com/perl6/doc/compare/a4292...afc5f0000f
01:46 travis-ci left
dalek c: c23878a | coke++ | doc/Language/5to6-perlfunc.pod6:
prefer space after comma
 01:49
synopsebot6 Link: doc.perl6.org/language/5to6-perlfunc
01:51 skids joined 01:53 sufrostico left 01:57 travis-ci joined
travis-ci Doc build failed. Will "Coke" Coleda 'prefer space after comma' 01:57
travis-ci.org/perl6/doc/builds/183461255 github.com/perl6/doc/compare/7dafc...3878a5be8d
01:57 travis-ci left 01:59 aborazmeh left 02:04 dataangel joined, kalkin- left, kalkin-_ joined 02:18 rpburkholder joined, gfldex left 02:19 gfldex joined, rburkholder left 02:29 loveperl joined, rpburkholder left
loveperl How to know the name of a class? 02:30
SmokeMachine there are 2 different modules in memory when I load it at compile time and I load it at run time?
m: say 42.^name
camelia rakudo-moar 29b228: OUTPUT«Int␤»
AlexDaniel m: say 42.HOW 02:31
camelia rakudo-moar 29b228: OUTPUT«Perl6::Metamodel::ClassHOW.new␤»
loveperl how?
AlexDaniel m: say 42.WHAT
camelia rakudo-moar 29b228: OUTPUT«(Int)␤»
loveperl ok tenk you
tank you
AlexDaniel m: say 42.WHO
camelia rakudo-moar 6fc769: OUTPUT«Int␤»
SmokeMachine when I get a instance of a singleton inside of a trait_mod code, I get a object and when I get it on a regular code I get a different object... is that expected? 02:33
www.irccloud.com/pastebin/3zFvLdlH/ 02:37
loveperl What do you think of this syntax?
if $saludo.WHO == Humano { say "hola humano"; }
SmokeMachine loveperl: try using the smartmatch... 02:38
m: if 42 ~~ Int { say "Hi Int" }
camelia rakudo-moar 6fc769: OUTPUT«Hi Int␤»
AlexDaniel loveperl: what are you trying to do?
huggable: xy
huggable AlexDaniel, nothing found
AlexDaniel huggable: xy problem 02:39
huggable AlexDaniel, You want to do X, but don't know how. You think you can solve it using Y, but don't know how to do that, either. So, you ask about Y in order to solve X, which doesn't make sense. You should ask about X.
loveperl I want to know if the name of the class is human? 02:41
? 02:43
what? 02:44
02:45 ilbot3 left
SmokeMachine but when I do that in the same "file" it gets the same object: 02:45
www.irccloud.com/pastebin/NK78LBlF/
AlexDaniel loveperl: then do $saludo ~~ Humano 02:46
loveperl: note that this is not just checking the name (Y), it actually checks the class (X) 02:47
02:47 wamba left
SmokeMachine AlexDaniel: do you know if thats expected? 02:47
02:47 ilbot3 joined
dalek c: 96761b2 | coke++ | xt/space-after-commma.t:
Add xt test to insure space after comma.

  * Only checks non-code.
  * Has several exceptions for text already in docs/
Closes #666		 02:48
[Coke] Still no feedback on github.com/perl6/doc/pull/1047 - fuzzy search.
samcv ok so to implement the Unicode Bidirectional Algorithm, that would look like: given a string, determine: 02:49
1: whether the paragraph direction is LTR, RTL or autoLTR
AlexDaniel [Coke]: bring it on! You'll get a bunch of feedback ;)
samcv 2: the embedding levels of the text (embedded rtl, ltr text etc)
02:49 pyrimidine joined
loveperl AlexDaniel tank you 02:50
samcv 3: being able to return a list of the mapping of characters in the string to indexs how they map
like RTL text obviously the mappings would just be the opposite of from 1->length of string
or 0 or however we start 02:51
dalek c/coke/levenshtein: ea1230a | coke++ | / (3 files):
Allow fuzzy searching

For #999
AlexDaniel SmokeMachine: as always, it is a bit hard to tell if there's no simple example to reproduce the issue
samcv what is anybodies thoughts on this?
what kind of perl 6 routines or whatever would be want to have access to this information
SmokeMachine There are! 02:52
www.irccloud.com/pastebin/3zFvLdlH/
www.irccloud.com/pastebin/NK78LBlF/
AlexDaniel ahhh
let's see…
samcv any thoughts on this AlexDaniel ?
AlexDaniel I feel like a rubber duck :) 02:53
02:53 loveperl left
SmokeMachine AlexDaniel: the first one is the problem, and the second is it working on the "same file"... 02:53
02:54 pyrimidine left 02:55 pyrimidine joined
AlexDaniel samcv: wait, what problem are you trying to solve? 02:55
samcv is unsure
02:55 travis-ci joined
travis-ci Doc build failed. Will "Coke" Coleda 'Add xt test to insure space after comma. 02:55
travis-ci.org/perl6/doc/builds/183470235 github.com/perl6/doc/compare/c2387...761b22094d
02:55 travis-ci left
samcv problem of not implementing the unicode bidirectional algorithm? 02:55
and not being able to pass the test given in the unicode spec 02:56
for said algorithm
AlexDaniel is it for string operations or for the source code? :) 02:57
samcv well. string operations for sure
02:58 kyclark joined
samcv i'm just trying to think about how that would look... are there any perl 6 devs whose native language is RTL? 02:58
02:58 travis-ci joined
travis-ci Doc build failed. Will "Coke" Coleda 'Allow fuzzy searching 02:58
travis-ci.org/perl6/doc/builds/183470746 github.com/perl6/doc/compare/7ebb0...1230a397f9
02:58 travis-ci left 02:59 cyphase joined, pyrimidine left
samcv AlexDaniel, also source code would be interesting... hah 03:00
use v6; use RTL; hah 03:01
or even maybe not as something you have to use... hm
m: say "\x[05D0]\x[05D1]\x[0028]\x[05D2]\x[05D3]\x[005B]\x[0026]\x[0065]\x[0066]\x[005D]\x[002E]\x[0029]\x[0067]\x[0068]"
camelia rakudo-moar 6fc769: OUTPUT«אב(גד[&ef].)gh␤»
samcv like this string is one of the unicode spec strings
AlexDaniel SmokeMachine: the two numbers in the first case…
SmokeMachine: I only get one on any subsequent runs 03:02
SmokeMachine: so one of them is from precompilation, right?
samcv paragraph direction is supposed to be LTR
1 0 2 4 3 5 6 7 8 9 10 11 12 13
this is the order they are shown in
03:02 [vlad] joined
samcv and that parenthesis should prolly be shown so that there's open and closing. at least for me it displays two right parens 03:03
SmokeMachine AlexDaniel: sorry, I didn't get... 03:04
AlexDaniel SmokeMachine: well, your first example with -MFuzz -MSum
SmokeMachine: so I run it and get two different numbers
SmokeMachine: BUT! If I run it again, I only get one number 03:05
SmokeMachine yes... me too... shouldn't be the same?
AlexDaniel SmokeMachine: if I delete .precomp, I'd get two numbers again
SmokeMachine yes, the same with me!
AlexDaniel that's just how the precompilation works, I think
SmokeMachine so, I can't use singleton on trait_mod? 03:06
03:07 kyclark left
AlexDaniel SmokeMachine: why not? 03:08
SmokeMachine: you can discard what happens during the precompilation
03:09 kyclark joined
SmokeMachine in my case I was pushing the functions with the trait on a array on the singleton to use it in another part of the code... 03:09
but it returns another object, so I can't get the funds... 03:10
*funcs
03:10 noganex_ joined
SmokeMachine AlexDaniel: My "real" case is this one: github.com/FCO/Test-Fuzz 03:11
github.com/FCO/Test-Fuzz/blob/mast...st/Fuzz.pm
03:11 kyclark left 03:13 noganex left 03:14 kyclark joined
SmokeMachine on github.com/FCO/Test-Fuzz/blob/mast...zz.pm#L115 and here github.com/FCO/Test-Fuzz/blob/mast...zz.pm#L104 I save the function on the %!fuzzers and on github.com/FCO/Test-Fuzz/blob/mast...zz.pm#L157 I use that... but its empty, because its another object... 03:15
AlexDaniel: ^^
samcv m: my @o = "\x[05D0]\x[05D1]\x[0028]\x[05D2]\x[05D3]\x[005B]\x[0026]\x[0065]\x[0066]\x[005D]\x[002E]\x[0029]\x[0067]\x[0068]".ords; my @n; my @nums = "1 0 2 4 3 5 6 7 8 9 10 11 12 13".split(' ').».Int; for ^@o -> $i { @n[ @nums[$i] ] = @o[$i] }; say @n.».chr; 03:16
camelia rakudo-moar 6fc769: OUTPUT«[ב א ( ד ג [ & e f ] . ) g h]␤»
03:17 kyclark left
samcv the hebrew characters are the opposite order in this if you look closely 03:18
tho that's more boring cause it's a LTR paragraph 03:19
AlexDaniel SmokeMachine: this is some weird stuff going on there! :)
03:22 cdg left
SmokeMachine AlexDaniel: So, thats not expected? 03:22
03:26 rurban joined
AlexDaniel SmokeMachine: I don't think I know enough about traits to answer :( 03:27
SmokeMachine :(
AlexDaniel: do you know who could help me?
AlexDaniel SmokeMachine: I think just asking this question here at a different time of a day is a good idea :) 03:31
03:31 pyrimidine joined
SmokeMachine AlexDaniel: OK! Ill do that!!! Thank you very much! 03:31
AlexDaniel SmokeMachine: so in other words, you want to push some data into %!fuzzers during the compilation? 03:34
SmokeMachine thats is!
thats it!
AlexDaniel: ^^
03:36 pyrimidine left
AlexDaniel yea, how to save some data from a trait… hmm! 03:40
I don't know! Time to sleep :) 03:41
03:42 khw left 03:47 AlexDaniel left, [vlad] left
samcv i want a uniprop-enum ... 03:55
hm
tailgate how do I match [\w -]+ (alphanumeric + dashes, at least one)
can't quite get the syntax
samcv m: say '-a' ~~ /[\w -]+/ 03:56
camelia rakudo-moar 6fc769: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unrecognized regex metacharacter - (must be quoted to match literally)␤at <tmp>:1␤------> 3say '-a' ~~ /[\w7⏏5 -]+/␤ expecting any of:␤ infix stopper␤»
samcv m: say '-a' ~~ /[\w '-']+/
camelia rakudo-moar 6fc769: OUTPUT«Nil␤»
samcv m: say '-a' ~~ /[<Alpha> '-']+/
camelia rakudo-moar 6fc769: OUTPUT«No such method 'Alpha' for invocant of type 'Cursor'␤ in block <unit> at <tmp> line 1␤␤»
samcv oh
m: say '-a' ~~ /<[\w -]>+/
camelia rakudo-moar 6fc769: OUTPUT«｢-a｣␤»
samcv there
m: say '-a' ~~ /<[<Alpha> -]>+/
camelia rakudo-moar 6fc769: OUTPUT«｢-a｣␤»
samcv that is what you want 03:57
er alpha num tho
tailgate thanks
samcv m: say '-a1' ~~ /<[<Alpha> -]>+/
camelia rakudo-moar 6fc769: OUTPUT«｢-a｣␤»
samcv m: say '-a1' ~~ /<[<AlphaNum> -]>+/
camelia rakudo-moar 6fc769: OUTPUT«｢-a｣␤»
samcv forget which that one is
m: say '-a1' ~~ /<[<Alpha>\d -]>+/ 03:58
camelia rakudo-moar 6fc769: OUTPUT«｢-a1｣␤»
samcv that works
04:12 aborazmeh joined, aborazmeh left, aborazmeh joined
samcv pythons unicodedata module that comes with it is kinda lame docs.python.org/2/library/unicoded...nicodedata 04:15
SmokeMachine I just started my "Generative testing" draft. Could someone take a look and give me some feedback please? 04:22
04:25 Ven left
samcv is this the same thing you were talking about earlier with AlexDaniel or somethnig different? 04:27
i'll look at it though, curious not sure how much I can help though 04:28
SmokeMachine samcv: That's my draft for perl6 advent calendar... talking about the module that I was talking with AlexDaniel... 04:29
samcv oh i'll look at it then!
SmokeMachine samcv: thats on the wordpress...
samcv oh 04:30
:(
SmokeMachine samcv: pastebin.com/McHWeum7 04:32
samcv nice 04:33
SmokeMachine samcv: did you like it?! :) 04:35
samcv does Test::Fuzz know what the result should be?
or does it just tell you if it gets an error?
SmokeMachine by default it just test if it did break... 04:36
samcv ah kk :) 04:37
SmokeMachine but you can set sub function(...) is fuzzed(test => *.is-prime) {}
for example...
if you want that the function only return prime numbers...
like the function blu here: github.com/FCO/Test-Fuzz/blob/master/test.p6 04:38
samcv: ^^
samcv nice 04:39
SmokeMachine did you find any terrible english error? 04:40
samcv: ^^
samcv nope
SmokeMachine great! 04:41
thank you very much!
samcv: Ill continue tomorrow!
samcv good start so far :)
SmokeMachine samcv: :D 04:42
samcv: thank you very much!
04:45 kyclark joined 04:53 kyclark left
MasterDuke .tell AlexDaniel it looks like you implemented something already, but i'd say just return the name for them or something like that, i don't see anybody ever wanting them for real 05:06
yoleaux MasterDuke: I'll pass your message to AlexDaniel.
05:09 geraud1 left 05:12 geraud joined 05:25 pyrimidine joined 05:29 pyrimidine left 05:37 telex left 05:38 telex joined 05:43 rurban left 05:44 geraud left 05:46 geraud joined 05:48 Cabanossi left 05:49 Cabanossi joined 05:55 geraud left, geraud joined 06:00 rurban joined, rurban left 06:01 brillenfux joined 06:16 BenGoldberg left 06:17 aborazmeh left 06:18 BenGoldberg joined 06:22 Vynce joined 06:25 BenGoldberg left 06:31 pyrimidine joined 06:34 darutoko joined 06:35 pyrimidine left 06:44 Vynce left
dalek href="https://perl6.org:">perl6.org: cb115bf | samcv++ | source/whatever/index.html:
Add links for atom-language-perl6 and script

For syntax highlighting and running Perl 6 from Atom		 06:49
06:50 RabidGravy joined 07:03 jonas1 joined 07:08 ufobat joined 07:09 imcsk8_ left, imcsk8_ joined 07:11 cooper_ left, cooper_ joined
RabidGravy morning 07:12
I see I managed to get the scheduled post thing right in the end
07:16 skids left 07:21 domidumont joined 07:26 domidumont left 07:27 domidumont joined
moritz RabidGravy++ 07:31
07:36 pyrimidine joined 07:40 pyrimidine left 07:45 Actualeyes left 07:54 kurahaupo joined 08:01 labster left 08:03 dataange` joined, [particle]1 joined, labster joined 08:04 dataangel left 08:07 ilbelkyr joined, Actualeyes joined 08:13 abraxxa joined, djbkd left 08:14 wamba joined
ufobat morning 08:22
the module Task::Star does not have a travis-ci integration, is that on purpose? 08:23
i mean, isnt it a pretty good test if Task::Star is installable via zef and/or panda?
Xliff What are dynvars? Has anything been written up about them? 08:24
08:24 zakharyas joined
moritz ufobat: sure, go for it! 08:24
ufobat kk :)
08:25 MARTIMM joined 08:27 baest left, MARTIMM left
arnsholt Xliff: A dynvar (also called contextual in Perl 6 parlance) is a variable that is looked up through *dynamic*, rather than lexical, scope 08:29
Where dynamic scope follows the call stack, so if the name is found in the current scope, you look in your caller's scope, and so on
They're somewhat uncommon in day-to-day code, but when you need them you really need them 08:30
moritz you use them when in the bad old days, you'd use a global
arnsholt They're very useful for global variables for example. Instead of a global STDOUT, Perl 6 has a dynvar $*OUT
08:30 pyrimidine joined
moritz but they are more thread safe and generally safer than globals 08:31
arnsholt This way, rather than overriding a global before doing something and making sure it's reset correctly and all that crap when you leave, you create a new binding for it with "my $*OUT = ..."
That way, STDOUT will be altered *only* inside the current function and code called by it
And as moritz says, thread safety 08:32
They're also very useful in grammar code
Since Perl 6 grammars are recursive descent, lexical scope in the program being compiled will correspond to the dynamic scope of the parser
moritz arnsholt: that concept might deserve its own blog post some day 08:33
I've never seen it explained well outside of #perl6
arnsholt Am I being volunteered? ;)
But more seriously, I agree
08:34 pierre_ joined
arnsholt It's a useful concept, and not very complicated either. Just not very common 08:34
(Although older Lisps like elisp frequently only have dynamically scoped variables, not lexically scoped)
And apparently you can do some magic with macros and dynamic scope to implement lexical scope
08:35 pyrimidine left
Xliff arnsholt++ # Brilliant explanation. Thanks! 08:38
samcv what unicode properties are allowed for heredocs? anything? 08:39
except forward slashes or something?
08:39 dakkar joined
samcv m: say q:to/'/;␤test␤' 08:39
camelia rakudo-moar 6fc769: OUTPUT«test␤␤»
samcv m: my $var = "\0"; say q:to/$var/;␤test␤' 08:40
camelia rakudo-moar 6fc769: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Ending delimiter $var not found␤at <tmp>:3␤------> 3'7⏏5<EOL>␤ expecting any of:␤ whitespace␤»
samcv any nonwhitespace character?
psch m: say q:to{/};␤test␤/ 08:43
camelia rakudo-moar 6fc769: OUTPUT«test␤␤»
08:43 pierre_ left 08:44 pierre_ joined
psch samcv: <-[ws]> sounds reasonable to me, but you'd probably have to check Perl6::Grammar to be sure 08:44
samcv yeah it looks like <-ws> 08:45
probably
psch right, not the character class consisting of 'w' and 's' :)
or, actually, 'not "not the character class consisting of 'w' and 's'"' vOv 08:47
+plus proper nested quoting parsing :P
samcv heredocs do quoted parsing? what?
psch no, just for my statement... :) 08:48
samcv m: say q:to/{/;␤this{}␤}
camelia rakudo-moar 6fc769: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Ending delimiter { not found␤at <tmp>:3␤------> 3}7⏏5<EOL>␤ expecting any of:␤ whitespace␤»
samcv oh ok
psch heredocs parse whatever language you give them
well, from Q and its sisters anyway
samcv yeah i usually use Q or qq if i need interpolation
i like Q 08:49
very much
psch m: say qqw:to/foo/;␤{ "foo", "bar" }␤baz␤foo
camelia rakudo-moar 6fc769: OUTPUT«(foo bar baz)␤»
samcv m: say "\r"
camelia rakudo-moar 6fc769: OUTPUT«
samcv aww no carriage return chars?
err
psch .u carriage return
yoleaux U+000D CARRIAGE RETURN (CR) [Cc] (␍)
U+240D SYMBOL FOR CARRIAGE RETURN [So] (␍)
samcv yeah
lemme see what my bot does 08:50
psch could maybe stuff that into camelia as well, or whateverable maybe
eval: say "\r"
evalable6 psch, rakudo-moar 6fc7698: OUTPUT«»
samcv oh it doeqsn't hm
psch well, that at least doesn't eat the >>
samcv m: say "\t"
camelia rakudo-moar 6fc769: OUTPUT« ␤»
samcv :)
i mean :(
should be ↹
my bot deos that at least, ↹ 08:51
though i guess less important than doing \t
err \r i'm tired
08:56 kurahaupo left 08:57 andrzeju_ joined, andrzejku left 09:00 g4 joined 09:01 baest joined 09:05 rindolf joined
dalek osystem: 4a13e58 | (Takumi Akiyama)++ | META.list:
Add JSON::Hjson to ecosystem

See github.com/akiym/JSON-Hjson		 09:06
09:06 Vynce joined
dalek osystem: e349294 | RabidGravy++ | META.list:
Merge pull request #273 from akiym/master

Add JSON::Hjson to ecosystem		 09:06
09:08 ccntrq joined 09:09 zakharyas left 09:12 thayne joined, grondilu joined 09:13 ccntrq left 09:14 ccntrq joined 09:23 bjz joined 09:27 wamba left 09:29 ccntrq left 09:31 pyrimidine joined 09:41 astj joined
astj moritz: Hi. Thank you for inviting me to github/perl6! I accepted. But I wonder for which reason I've invited. Is there any reason? 09:42
09:42 ccntrq joined
moritz astj: you submitted a pull request to the ecosystem 09:44
astj: in future, you can push directly. And merge other's pull request, should you feel so inclined
astj moritz: Okay, thanks! 09:45
moritz astj: the ecosystem list was never meant to enable a gate keeper, we just did it because all our stuff is in github
astj: you're welcome
09:47 bjz left, ccntrq left 09:53 rurban joined, ccntrq joined 09:55 pmurias joined 09:57 pyrimidine left 10:02 bjz joined, Vynce left 10:03 pierre_ left 10:05 mniip joined, Actualeyes left, labster left 10:07 astj left, astj joined 10:10 pierre_ joined 10:11 pierre_ left 10:12 astj left 10:24 sena_kun joined
pmurias jnthn: does checking if objects need to be repossesed have a big performance price on MoarVM? 10:37
10:38 wamba joined
jnthn pmurias: iirc, most of the time once we're running code it's just an integer comparison (because by runtime where's no SCs on the compiling stack) 10:39
s/where/there/
Presumably that stays CPU cache hot, and the branch is predictable, so it's not too awful 10:40
10:44 lucasb joined 10:49 lukaramu joined
ufobat moritz, done ;) but i guess tadzik has to add the module to the CI test services github.com/tadzik/Task-Star/pull/12 10:51
tadzik oh, will do 10:52
ufobat cool 10:54
El_Che not enough sleep, not enough coffee and reading jvm docs for performance tuning a big jvm service. Bad combination 10:55
ufobat hmm on the other side, does it really make sense
the tests are triggered whenever you.. push something? 10:56
that is probably wrong for Task::Star
10:56 pyrimidine joined 10:58 rindolf left 11:03 bjz_ joined, bjz left
dalek c: ea1230a | coke++ | / (3 files):
Allow fuzzy searching

For #999		 11:04
c: 2586a29 | (Zoffix Znet)++ | / (3 files):
Merge pull request #1047 from perl6/coke/levenshtein

Allow fuzzy searching
11:07 zakharyas joined 11:10 cognominal left 11:11 cognominal joined 11:12 dogbert17 joined, travis-ci joined
travis-ci Doc build failed. Zoffix Znet 'Merge pull request #1047 from perl6/coke/levenshtein 11:12
travis-ci.org/perl6/doc/builds/183553537 github.com/perl6/doc/compare/96761...86a29a543b
11:12 travis-ci left 11:15 dogbert17_ left 11:19 pierre_ joined
phatbot Same fromspace collectable accesing error "MoarVM panic: Collectable 0x2ae0503e2918 in fromspace accessed" 11:22
.tell cygx reminder: your post is up next in ~12.5 hours 11:43
yoleaux phatbot: I'll pass your message to cygx.
11:51 pierre_ left
lucasb
.oO( hmm, a Collectable fromOuterSpace )
 11:58
brrt jnthn ^ what phatbot says 12:04
12:04 rurban left 12:17 nicky joined 12:18 sena_kun left
nicky ==> Testing File::Find Unhandled exception: Failed to write to filehandle: NULL string given 12:18
brrt /me wonders, just out of interest, how hard it would be to have a 'perl6-syntax-check' daemon process and an emacs mode to use it
psch brrt: something like "run perl6 -c on the current buffer after n seconds of idle" or what are you imagining?
i suspect that depends more on how emacs works than what perl6 offers :) 12:19
brrt something like that 12:20
well, emacs has flymake-mode, iirc
so that is just something to hack into
but the current wisdom is to use a client-server model for that, iirc
12:21 TEttinger left 12:26 bjz_ left, bjz joined
CIAvash[m] brrt: github.com/hinrik/flycheck-perl6 12:29
tbrowder RabidGravy: nice Advent post! It helps me get a feel for using concurrency in the real world. 12:31
nicky I'm getting a make install failure on rakudo star 2016.11 pastebin.com/7PzmcaTS 12:35
12:39 sena_kun joined
phatbot sena_kun: I don't think building %method_to_role is needed anymore, since we already know method's name by that point. 12:46
brrt ooh that is cool 12:47
thanks CIAvash[m] 12:48
sena_kun phatbot, method name - yes, but role?
phatbot sena_kun, that's what the added logic needs to do, look for a role with that method and tells us the role name when found. There's no need to store names of all methods for that 12:49
sena_kun phatbot, ah, so that's your point. Yes, you are correct. Another patch is coming. 12:50
phatbot \o/ 12:51
arnsholt Good point! 13:03
That should make it easier to report all roles at once too 13:04
13:06 rurban joined
sena_kun I've pushed update. arnsholt, do you think we need reporting of all roles with the unimplemented method after all? It is not hard to write, just asking. 13:08
13:08 canopus left
arnsholt Dunno 13:08
phatbot Wouldn't there be a merge conflict anyway?
or whatever it's called
m: role A { method foo {} }; role B { method foo {} }; class Z does A does B {}
camelia rakudo-moar 6fc769: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Method 'foo' must be resolved by class Z because it exists in multiple roles (B, A)␤at <tmp>:1␤»
phatbot m: role A { method foo {...} }; role B { method foo {} }; class Z does A does B {} 13:09
camelia ( no output )
phatbot m: role A { method foo {...} }; role B { method foo {...} }; class Z does A does B {}
camelia rakudo-moar 6fc769: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Method 'foo' must be implemented by Z because it is required by a role␤at <tmp>:1␤»
phatbot m: role A { method foo {...} }; role B { method foo {...} }; class Z does A does B { method foo {} }
camelia ( no output )
arnsholt Yeah, that's what I thought. The conflict only crops up if both roles have defaults
13:09 wamba left
phatbot sena_kun: from the above it does look like reporting all the roles could be useful. 13:10
arnsholt Yeah, I'd add it I think
psch m: role A { method foo {...} }; role B { method foo {} }; class Z does B does A {}
camelia ( no output )
arnsholt It's relevant knowledge about how to implement the required method, IMO
phatbot sena_kun: if it's not hard :) I know we've been kinda keeping you on the wheel... pushing all these changes :) 13:11
sena_kun phatbot, no problem, thanks for comments anyway. But all methods commit will be a bit later, since I'm busy right now. 13:12
13:13 matiaslina joined, canopus joined 13:14 wamba joined 13:20 pierre_ joined 13:21 kurahaupo joined
lucasb m: say (my %h{Int}).perl 13:28
camelia rakudo-moar 6fc769: OUTPUT«(my Any %{Int} = )␤»
dalek c: 3e88149 | coke++ | xt/space-after-commma.t:
remove trailing whitespace
c: 2af9049 | coke++ | xt/trailing-whitespace.t:
ignore this 3rd party file
lucasb ^^ no need for the equal sign
phatbot m: say (my %h{Int}).perl.EVAL.perl 13:29
camelia rakudo-moar 6fc769: OUTPUT«5===SORRY!5=== Error while compiling /home/camelia/EVAL_0␤Malformed initializer␤at /home/camelia/EVAL_0:1␤------> 3(my Any %{Int} = 7⏏5)␤ expecting any of:␤ prefix␤ term␤»
phatbot Yes, .EVAL.perl is meant to roundtrip
lucasb SourceBaby: help
SourceBaby lucasb, Use s: trigger with args to give to sourcery sub. e.g. s: Int, 'base'. See modules.perl6.org/dist/CoreHackers::Sourcery
lucasb s: Hash, 'perl'
SourceBaby lucasb, Sauce is at github.com/rakudo/rakudo/blob/6fc7...Mu.pm#L527
lucasb s: {}, 'perl'
SourceBaby lucasb, Ehhh... I'm too scared to run that code.
phatbot s: Hash, 'perl', \()
SourceBaby phatbot, Sauce is at github.com/rakudo/rakudo/blob/6fc7...Mu.pm#L528
13:30 ZoffixW joined
ZoffixW s: (my %h{Int}), 'perl', \() 13:30
SourceBaby ZoffixW, Sauce is at github.com/rakudo/rakudo/blob/6fc7...sh.pm#L704
13:30 ZoffixW left
phatbot It's not scared of me ;) 13:30
13:30 prammer joined
phatbot Err... I mean of this Zoffix guy 13:30
lucasb ah, it feels the k/v pairs after the equal sign 13:31
*it fills :)
sena_kun I'm going to switch for $must-to-do-it-work now, so the patch will be tomorrow, I suppose. Sigh. 13:34
phatbot sena_kun: I'll merge current version then. And you can submit another PR for multi-role thing when you get a chance 13:35
13:36 travis-ci joined, pyrimidine left
travis-ci Doc build failed. Will "Coke" Coleda 'ignore this 3rd party file' 13:36
travis-ci.org/perl6/doc/builds/183586638 github.com/perl6/doc/compare/2586a...f9049c11af
13:36 travis-ci left
sena_kun phatbot, got it. Thanks for help, everyone. 13:36
13:37 pyrimidine joined 13:38 nicq20 joined 13:42 kyclark joined, sufrostico joined 13:45 pierre__ joined 13:46 pierre_ left 13:49 sufrostico left
lucasb m: say Mu.is-lazy, Hash.is-lazy 13:50
camelia rakudo-moar 6fc769: OUTPUT«FalseFalse␤»
lucasb m: say List.is-lazy
camelia rakudo-moar 6fc769: OUTPUT«Cannot look up attributes in a List type object␤ in block <unit> at <tmp> line 1␤␤»
lucasb can List get a method is-lazy(List:U) { False } ? 13:51
phatbot s: List, 'is-lazy'
SourceBaby phatbot, Sauce is at github.com/rakudo/rakudo/blob/6fc7...t.pm#L1001
jnthn Should do, yes
psch m: say Any.is-lazy
camelia rakudo-moar bff051: OUTPUT«False␤»
phatbot I think a better approach would be for ^ that one to be a :D and let Mu handle the :U
lucasb phatbot: right 13:52
13:55 kurahaupo left, kyclark left
phatbot m: use MONKEY; my $r := ^2; nqp::getattr($r, Range, '$!is-int') = 0; 13:59
camelia rakudo-moar bff051: OUTPUT«Cannot modify an immutable int␤ in block <unit> at <tmp> line 1␤␤»
phatbot how can I change that attribute?
m: use MONKEY; my $r := ^2; nqp::bindattr($r, Range, '$!is-int', 0); 14:00
camelia rakudo-moar bff051: OUTPUT«P6opaque: representation mismatch when storing value (of type Int) to attribute (of type int)␤ in block <unit> at <tmp> line 1␤␤»
phatbot m: use MONKEY; my $r := ^2; nqp::bindattr($r, Range, '$!is-int', nqp::istype(42, Bool)); 14:01
camelia rakudo-moar bff051: OUTPUT«P6opaque: representation mismatch when storing value (of type Int) to attribute (of type int)␤ in block <unit> at <tmp> line 1␤␤»
jnthn _i
[Coke] phatbot: thanks for the merge. OOC, did you test or just assume we'd get some feedback only after mergeback? ;)
phatbot jnthn++ thanks 14:02
14:02 Ven joined
phatbot [Coke]: I didn't test, no. 14:02
oh shit... perhaps I should have... 14:03
I get freezes on my box :}
If I type "str" it takes it like 2 seconds to react to any further actions 14:04
Mmm
Or maybe it's something else entireely. Seems to be fine on another box in Chrome.
14:05 nicky left
phatbot oh, but same lag if I search for "a". And after it displays results, moving mouse around results has a delay before it reacts to anything. Is it just me? 14:05
sena_kun phatbot, confirmed on latest firefox. 14:06
14:06 kyclark joined 14:07 cibs left 14:09 cibs joined 14:10 pmurias left
phatbot [Coke]: so seems like there's a performance issue? Not sure if anything can be improved with it. Perhaps trigger levenshtein only after $x characters have been typed? 14:11
It's nice that it shows results for in_timezone, for example :)
[Coke] alexdaniel had suggested a faster levenshtein-like algorithm as well.
I can make it skip a few chars, one sec. 14:12
... er
14:12 cdg joined
[Coke] more like 10s of minutes. Have to find what the old default algorithm was, then check term.length before using getEditDistance() 14:13
14:13 prammer left
[Coke] I can do it after $DAYJOB if someone hasn't beaten me to it. 14:15
also, looks like someone changed the SASS processing, as external links are now being shown. 14:16
github.com/perl6/doc/issues/999 was the original ticket, commented about the slowdown and algorithm there. 14:20
phatbot That was me. I didn't change SASS processing, I just followed the proceedure we have for changing styles. You're meant to change assets/sass/style.sass (or whatever it is) and then run the dev app to update html/css/style.css (or whatver it is) and commit both files to repo. 14:21
And I guess I promised a few months ago to make the doc build do that stuff so we could nix html/css/style.css from repo
and still haven't done it :} 14:22
phatbot adds it onto Christmas Holiday Fixes list
14:26 kurahaupo joined 14:28 mawkish_ joined 14:31 mawkish_ left, mawkish_ joined 14:34 jonas1 left 14:36 wamba left
tbrowder ref p6 publicity: i think i've seen p6 described as a "large" core language, esp. compared to minimalist C. what's opinion of putting stats on perl6.org with numbers of built-in routines, classes, modules, dynamic vars, etc.? should, hopefully, be greppable, cacheable, and automatable. 14:37
phatbot So that all the people who compare us to PHP would have actual numbers to point to? :) 14:38
tbrowder roger! but maybe emphasize a subset for most new users. 14:39
phatbot I don't see anything posivite comming out of such an endeavour and only negative. Especially since a lot of routines haven't been moved to Rakudo::Internal yet
jnthn What do you count as the core of the C language?
libc isn't exactly small :) 14:40
14:40 prammer joined
jnthn And if that doesn't count, then nor does most of CORE.setting. :) 14:40
tbrowder hm, all inside rakudo seen by users?
phatbot yes
14:41 pierre__ left
phatbot m: use Test; is-deeply SEQUENCE(1, 6), (1...6) 14:41
camelia rakudo-moar 1d4600: OUTPUT«ok 1 - ␤»
phatbot Like that's not a user sub 14:42
jnthn Sure, but my question is more about where do you draw the line between language and standard library.
phatbot m: CORE::.values.grep({$_ ~~ Sub}).elems.say 14:43
camelia rakudo-moar 1d4600: OUTPUT«121␤»
14:43 kyclark left
jnthn It's for sure that Perl 6 has a decidedly larger grammar than C, Java, etc. 14:43
14:43 cdg_ joined
tbrowder of course that's a good question, but someone deeply familiar with the lang should be able to draw it 14:43
phatbot And well, you can override core stuff, as opposed to many other languages where you get a conflict
14:44 kyclark joined, Actualeyes joined
tbrowder part of the super language aspect, for advanced users 14:44
14:44 g4 left
phatbot Draw what? 14:44
tbrowder a line between "core" and library 14:45
phatbot Seems pointless semantics game to me.
tbrowder there are always folks who like such stats
14:46 cdg left
tbrowder when, say, comparing languages for suitability for their purposes 14:46
jnthn It could be nice to show the batteries that Perl 6 includes in its standard library.
tbrowder there is a lot appeal for a lang that can do a lot without hunting for lots of libraries 14:47
tadzik yep
tbrowder extra libraries
phatbot hah 14:48
You've not seen Zefram's talk then, I guess :)
tbrowder hm, is he yr brother?
lizmat clickbaits p6weekly.wordpress.com/2016/12/12/...rofilesql/ 14:49
Files=1155, Tests=53893, 200 wallclock secs (11.34 usr 4.53 sys + 1192.98 cusr 126.25 csys = 1335.10 CPU)
tbrowder no, i don't think so
jnthn fwiw, it depends what the "a lot" is, to me. Putting Set/Bag/Rat/Complex into the standard library is a non-brainer because math ain't going to suddently come out of date tomorrow. Same for stuff like async sockets/processes.
With serialization formats though, it's much more touch and go. The current fashionable ones come and go. 14:50
And some things are better just having different release cadences than the core language. 14:51
phatbot tbrowder: yes, all people with Z in their name are related.
14:51 dj_goku left
phatbot I can't find the URL right now, but his entire talk was about dissing Perl 6 for having too many features, while showing bugs with their interaction and coming to the conclusion that Perl 6 is inherently flawed and must be discarded. 14:52
psch i think one notable point was that assigning IterationEnd to an elemnt in an Array breaks iteration..?
+e
arnsholt That sounds a lot like the infamous "we found bugs in Bugzilla and CGI.pm, therefore Perl is inherently broken" talk that cropped up a few years back
phatbot So people like him, along with folks who like to make fun of PHP for having many built ins, would just love your stats, tbrowder
tbrowder well, one way to handle those types is to ignore them 14:53
phatbot arnsholt: yeah. The premise is inherently flawed. He'd have to reach for library XYZ to do feature X we do in core, giving him with as many "interactions", except now you have completely different projects with completely different styles and authors in play.
psch i think it was act.yapc.eu/lpw2015/talk/6436 14:54
jnthn phatbot: Yes, that's pretty much what I was going to write :)
phatbot :)
psch: aye, title looks familiar
[Coke] Why do talks like that get approved, I wonder. 14:55
14:55 kyclark left
[Coke] at a perl conference, I mean. 14:56
14:57 skids joined 14:58 sena_kun left, mawkish_ left 14:59 mawkish_ joined
jnthn [Coke]: Perhaps because having a community that won't listen to dissenting ideas isn't really healthy in general. :) 14:59
psch well, the talk has been recorded
tbrowder i couldn't find the talk itseif, but i think i've seen parts. he sounds like a large group of. nay-sayers over at perl monks
psch although i personally find the audio utterly incomprehensible
15:01 zakharyas left, zakharyas joined
tbrowder listening is fine, i leave it to smarter people to refute facts, and get on to helping advance my own interests with folks who share them--life is too short 15:02
15:03 kyclark joined, kyclark left, pierre_ joined
[Coke] jnthn: It's the question of dissent, but respect. 15:04
*sigh*. "it's NOT ..."
tbrowder i didn't mean refute facts...i meant argue as in debate 15:05
15:05 zakharyas left
[Coke] I wonder how many of the issues he raised were raised to the community before he felt the need to whine (his word) at a conference. if it's the same zefram, he's also a p5 release manager, among other things. 15:05
15:06 mawkish_ left 15:07 mawkish_ joined
phatbot Not sure about the timing of the talk, but Zefram did file a whole ton of tickets. Especially the ones with edge case math, like NaN handling, negative zeros, overflows. 15:07
I'd even say he's my favourite ticket reporter. 15:08
moritz phatbot: don't let masak hear that :-)
phatbot :D
jnthn I'm glad of both. :-) masak tends to report things that crop up in sometimes interesting but relatively normal programs. Zefram looks at how stuff can be broken on purpose. 15:10
[Coke] (report) ah, well that's awesome. Zefram++
lizmat fwiw, zefram liked MoarVM and has said that perl 6 should be implemented on top of that (from scratch) 15:11
perhaps he will change his mind when we have a fully functional JS backend :-) 15:12
15:12 brillenfux left 15:13 mawkish_ left
ilmari I thought he said he liked nqp as well 15:13
15:13 mawkish_ joined
ilmari (as a language-implementation language) 15:13
15:14 khw joined
tadzik LPW 2017: Zefram implements Perl 5 on nqp 15:14
15:15 Khisanth left, wamba joined
tadzik oh wow, SQL backend for profiler, that's nice 15:16
15:16 mawkish_ left
lizmat tadzik: perl 5 on nqp, isn't that "use v5" ? 15:17
tadzik quite :)
ilmari that still requires rakudo, no? 15:18
lizmat yes
eh, well
hmmm... technically, not I guess
jnthn You could feasibly do a Perl 5 built on the NQP compiler toolchain without re-using any part of Rakudo.
15:19 mawkish_ joined
brrt wonders whom he can challenge to do that 15:19
15:20 MorayJ joined, MasterDuke_ joined
perlpilot the next time any of you see Zefram, politely ask him to fix Data::Alias to work with perl 5.24 :-) 15:20
15:21 mohae_ joined
lizmat perlpilot: I've heard rumours that that is not going to be possible :-( 15:23
otoh, I've heard those rumours before
15:24 kyclark joined, mohae left 15:25 mawkish_ left, mawkish_ joined
lucasb multi role R[::T] {...} 15:26
multi role R[::T, ::U] {...}
^^ would such a feature make sense?
jnthn We already have it, just don't need to write the multi :) 15:27
15:27 Khisanth joined 15:29 thayne left
lucasb oh, I wasn't aware it already works, nice :) 15:30
arnsholt brrt: The I must confess that "how would I go about implementing Perl 5 on NQP" has been considered by me as an intellectual excercise, but I'm certainly not going to volunteer to actually *do* it!
(Steps one and two being implementing lex and yacc-a-likes for NQP, IMO) 15:32
lucasb m: role R1 {}; role R2[Int] {}; say R1[1,2], R2[Str] 15:33
camelia rakudo-moar 1d4600: OUTPUT«(R1[Int,Int])(R2[Str])␤»
lucasb ^^ If you try to '.new' these roles, it expectedly dies 15:34
15:34 mawkish_ left 15:35 mawkish_ joined
lucasb class C does R[invalid args] {} also dies 15:35
but... couldn't this check happen at compile time for the invalid ones?
IOW, should even be possible to refer to them by the names 'R1[1,2]', etc. no? 15:36
*shouldn't
lizmat arnsholt: I think the main issue is really: which perl 5 are you going to target? with all the recent syntactic changes :-(
15:37 dj_goku joined
lizmat I think it's going to be less important now that we have Inline::Perl5 15:38
jnthn lucasb: If you .new a role then we don't calculate the pun until runtime
lucasb: For compile-time composition it does die at compile time already
lizmat and once we can get "supersede" to work and include modules with from<Perl5>, one should be able to upgrade a module to a Perl 6 version transparently without needing to change code
lucasb jnthn: ok, understood, thanks 15:39
15:40 kyclark left 15:42 dj_goku left, brrt left 15:43 pyrimidine left, pyrimidine joined 15:54 kyclark joined
lucasb found some strange errors for shaped arrays, will post 3 lines here, ok 15:55
(my @a[2]).new.perl # Type check failed in binding to @dims; expected Positional but got Any (Any)
(my @a[2] = <a b>).new.Array # This type (Scalar) does not support elems
my @a[<a b c>] # P6opaque: get_boxed_ref could not unbox for the representation '20' of type Failure
also, for 'my @a[...]', idk what method it asks the value inside brackets (if .Int, .elems, or other), but for 'my @a[SomeEnum]', it returns a shaped array 15:56
m: dd my @a[Bool], my @b[Order] 15:57
camelia rakudo-moar 1d4600: OUTPUT«Array.new(:shape(2,), [Any, Any])␤Array.new(:shape(3,), [Any, Any, Any])␤»
lucasb I just wondered if that was expected and intended
phatbot no, crappy errors aren't expected :) 16:00
The my @a[<a b c>] one looks like it tries to convert a string to number, gets Failure, and tries to store isbig_I it
16:01 mad__ joined, mad__ left, mad__ joined
phatbot m: use nqp; nqp::isbig_I(nqp::decont(+'a')); 16:01
camelia rakudo-moar 1d4600: OUTPUT«P6opaque: get_boxed_ref could not unbox for the representation '20' of type Failure␤ in block <unit> at <tmp> line 1␤␤»
phatbot there are a couple more I have in my notes at home that give exception blah blah produce no messsage 16:02
Oh, I misread. 16:04
16:04 cooper_ left
phatbot lucasb: yeah, that's expected. Enums can be used as a shape 16:04
lucasb phatbot: because they numify to their number of elements? 16:05
phatbot m: dd +Bool
camelia rakudo-moar 1d4600: OUTPUT«Cannot resolve caller Numeric(Bool: ); none of these signatures match:␤ (Bool:D $: *%_)␤ in block <unit> at <tmp> line 1␤␤»
phatbot They don't. I just recall a commit fly through implementing that feature.
lucasb hm, interesting, ok 16:06
phatbot It's in 2016.10's changelog: github.com/rakudo/rakudo/blob/nom/...geLog#L279 16:07
lucasb oh, inserting the commit hash in the changelog only started in 2016.11. phatbot++ 16:08
phatbot And that's the commit: github.com/rakudo/rakudo/commit/55...05cb20ac7a 16:09
lucasb thanks
16:10 mawkish_ left 16:11 mawkish_ joined
lizmat lucasb: I think that was inspired by a blogpost by gfldex, if memory serves 16:13
gfldex.wordpress.com/2016/09/21/ar...your-keys/ 16:16
lucasb ah, you linked other post in you commit github.com/rakudo/rakudo/commit/fe...80812cbceb 16:17
but anyway, I wasn't aware the feature was intentional, lizmat++
dalek c: a972217 | MasterDuke17++ | doc/Type/Cool.pod6:
Add some missing linebreaks to create code blocks
synopsebot6 Link: doc.perl6.org/type/Cool
16:21 pierre_ left 16:25 acrussell joined, pierre_ joined
SmokeMachine I just started my "Generative testing" draft. Could someone take a look and give me some feedback please? 16:25
16:25 acrussell left 16:26 travis-ci joined
travis-ci Doc build failed. MasterDuke17 'Add some missing linebreaks to create code blocks' 16:26
travis-ci.org/perl6/doc/builds/183640038 github.com/perl6/doc/compare/2af90...722174d16d
16:26 travis-ci left
SmokeMachine there are 2 different modules in memory when I load it at compile time and I load it at run time? 16:27
when I get a instance of a singleton inside of a trait_mod code, I get a object and when I get it on a regular code I get a different object... is that expected?
00:37 www.irccloud.com/pastebin/3zFvLdlH/
but when I do that in the same "file" it gets the same object:
00:45 www.irccloud.com/pastebin/NK78LBlF/
16:27 abraxxa left 16:29 pierre__ joined, pierre_ left 16:30 wamba left 16:32 mr-foobar joined
SmokeMachine is that expected? 16:32
16:34 pyrimidine left 16:35 pyrimidine joined
phatbot SmokeMachine: no idea. Try hunting down nine 16:35
[Coke] fwiw, irccloud is blocked here.
perlpilot same here. gist.github.com is fine though :) 16:37
16:38 dj_goku joined, dj_goku left, dj_goku joined 16:39 mawkish_ left, domidumont left, mawkish_ joined, pyrimidine left 16:40 vike joined 16:42 wamba joined, cdg_ left 16:43 pyrimidine joined, cdg joined, dj_goku left 16:44 mawkish_ left 16:45 mawkish_ joined
SmokeMachine [Coke]: what do you mean with "irccloud is blocked here"? 16:48
16:48 girafe joined
SmokeMachine hum!! I got it! 16:50
perlpilot SmokeMachine: I (and presumably Coke) am at $work where certain URLs are blocked by our corporate rules.
SmokeMachine Ill paste it on GitHub...
[Coke], perlpilot: gist.github.com/FCO/65196d4bff0362...1bafa50ae8 16:53
16:55 mawkish_ left, mawkish_ joined
perlpilot SmokeMachine: an object's .WHERE gives a memory location which may change over time due to GC runs. You want .WHICH for object identity. 17:00
SmokeMachine perlpilot: ok!
phatbot oh damn...
phatbot should've actually read the code :P
SmokeMachine: FWIW, if you tell the gist extensionb is .p6 it'll add highlighting 17:01
[Coke] Looks like precomp is involved; run the perl 6 command twice...
(second time you only get one say.)
17:01 rindolf joined
SmokeMachine [Coke]: Yes, that happens... 17:02
[Coke]: but thats not what I want...
let me explain with some code...
[Coke] You want a singleton, that doesn't look like one, yes. :) 17:03
Is that not it?
SmokeMachine on github.com/FCO/Test-Fuzz/blob/mast...zz.pm#L115 and here github.com/FCO/Test-Fuzz/blob/mast...zz.pm#L104 I save the function on the %!fuzzers and on github.com/FCO/Test-Fuzz/blob/mast...zz.pm#L157 I use that... but its empty, because its another object... 17:04
[Coke]: I want a singleton where I .push() some data at compile time and I can get that data at run time... 17:05
rindolf Hi all. 17:07
phatbot Hi, rindolf 17:08
rindolf TimToady: are you there?
phatbot: hi
phatbot m: class Singleton { my @.stuff; BEGIN @.stuff.push: <some stuff>; method new { self }; }; my $foo = Singleton; my $bar = Singleton; $foo.stuff.push: <moar stuff>; dd $bar.stuff 17:09
camelia rakudo-moar 1d4600: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable @.stuff used where no 'self' is available␤at <tmp>:1␤------> 3ss Singleton { my @.stuff; BEGIN @.stuff7⏏5.push: <some stuff>; method new { self }␤ expecting any of:␤ term␤»
phatbot waat
you're trickin' me
oh
m: class Singleton { my @.stuff; BEGIN self.stuff.push: <some stuff>; method new { self }; }; my $foo = Singleton; my $bar = Singleton; $foo.stuff.push: <moar stuff>; dd $bar.stuff 17:10
camelia rakudo-moar 1d4600: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤'self' used where no object is available␤at <tmp>:1␤------> 3class Singleton { my @.stuff; BEGIN7⏏5 self.stuff.push: <some stuff>; method n␤ expecting any of:␤ term␤»
phatbot screw you
perlpilot phatbot: you're exhibiting the first signs of insanity :)
phatbot m: class Singleton { my @.stuff; BEGIN ::?CLASS.stuff.push: <some stuff>; method new { self }; }; my $foo = Singleton; my $bar = Singleton; $foo.stuff.push: <moar stuff>; dd $bar.stuff
camelia rakudo-moar 1d4600: OUTPUT«Array @.stuff = [("some", "stuff"), ("moar", "stuff")]␤»
phatbot perlpilot: that means I've been hiding it well for all these years 17:11
Muahahhaahaha
17:11 grondilu left
perlpilot in that case ... good job so far! 17:12
;)
SmokeMachine phatbot: when I try in the same "file" it works... the problem occurs when I use a different file... 17:14
phatbot SmokeMachine: so you got this singleton in a module... You `use Blagh` it two different files that are being run as a single program and you no longer have a singleton? 17:15
oh you pasted links 17:16
SmokeMachine phatbot: that's it!
phatbot: a small example: 17:18
phatbot SmokeMachine: well, I've no idea if that's even supposed to work, especially with the lexical loading branch we're merging Saturday. From what I understand, you could load two different modules with the same name, for example, and it would work. So I'm thinking the two different uses are two different compilation units and that's why it doesn't work like you expect it to
SmokeMachine: but I don't know this stuff, which is why my initial suggestion was to seek out nine :) 17:19
SmokeMachine gist.github.com/FCO/1561729d83b673...1d90199184
phatbot Add .p6 extension to files 17:20
gregf_ phatbot: m: say [*] 1..10
SmokeMachine phatbot: Im sorry, but what means "seek out nine"?
phatbot: thats not a .p6 file...
phatbot SmokeMachine: a person, nine. He's the expert on all this module stuff
gregf_: ? 17:21
Xliff m: say '"' ~~ / <[ '"' ]> /
camelia rakudo-moar 1d4600: OUTPUT«Potential difficulties:␤ Quotes are not metacharacters in character classes␤ at <tmp>:1␤ ------> 3say '"' ~~ / <7⏏5[ '"' ]> /␤ Repeated character (') unexpectedly found in character class␤ at <tmp>:1␤ ------> 3say '"' ~…»
Xliff m: say '"' ~~ / <[ " ]> /
camelia rakudo-moar 1d4600: OUTPUT«｢"｣␤»
SmokeMachine hum! nine is a person!!! thanks!
17:22 cibs left
phatbot gregf_: what was that? 17:23
m: say '"' ~~ / '"' /
camelia rakudo-moar 1d4600: OUTPUT«｢"｣␤»
17:24 cibs joined 17:25 cyphase left 17:26 toolforger joined
toolforger g'evening everybody :-) 17:26
phatbot \o
SmokeMachine .tell nine Hi! could you help me? Could you tell me if it's working as expected? use a module with a singleton in 2 different files (one of then at compile time) should return different object instances? like that: gist.github.com/FCO/1561729d83b673...1d90199184
yoleaux SmokeMachine: I'll pass your message to nine.
17:27 domidumont joined
phatbot gregf_: m: say [+] ^9999999999999999999999999999999999999999999999999999999999999999999999999 17:27
17:28 prammer left
phatbot hm, weird. perl6.fail lists tickets if they've been updated after being resolved :/ 17:28
toolforger Anybody available for discussing JVM desirables? 17:29
phatbot What are "desirables"?
toolforger E.g. what package names to use for JVM binaries for a given Perl source file/module 17:30
phatbot vOv 17:31
17:31 cyphase joined
toolforger I know how Java tools expect things to be, I'm not so sure about what's desirable for Perl 17:31
17:31 rurban left, nicq20 left
gregf_ phatbot: sorry, i thought you were a bot *runs* 17:31
toolforger I would like to implement some sensible defaults for that :-)
phatbot gregf_: so you see a bot and you start feeding him random crap? Pfft... That's racism! 17:32
m: say [+] ^9999999999999999999999999999999999999999999999999999999999999999999999999 17:33
camelia rakudo-moar 1d4600: OUTPUT«49999999999999999999999999999999999999999999999999999999999999999999999985000000000000000000000000000000000000000000000000000000000000000000000001␤»
phatbot That one actually works FWIW :)
jnthn
.oO( bigbotry )
17:33 berserk_guts joined
gregf_ m: ([*] 1..10).say # phatbot does that look like random crap? 17:33
camelia rakudo-moar 1d4600: OUTPUT«3628800␤»
gregf_ and racism is too strong a word :/ 17:34
17:34 berserk_guts left
phatbot I guess I think I'm funnier than I really am :( 17:34
gregf_ heh 17:35
17:35 pierre__ left
samcv good morning perl6 17:37
phatbot \o\
lizmat samcv o/
17:39 pierre_ joined, dj_goku joined
phatbot toolforger: ask psch, bartolin, or arnsholt if you see them around. 17:41
17:42 pyrimidine left, pyrimidine joined 17:43 pierre_ left
toolforger Being a chat noob: Is there a good way to invite two people to a common chat? 17:45
17:45 dj_goku left
toolforger I'm using Pidgin, FWIW 17:45
phatbot 's no idea
ilmari create a new channel, invite them
geekosaur pretty much just have to create a channel for yourself (by joining it; it's dynamic) and /invite
toolforger Ah, thanks, that's straightforward enough :-D 17:46
ilmari you can make the channel invite-only, /mode #channel +i
that way only people you /invite can join
17:47 pyrimidine left
geekosaur +s can also be useful (so it doesn't show in a channel list) 17:47
toolforger I think it's okay if I name it #perl6-jvm or something like that 17:48
MasterDuke_ toolforger: are you familiar with ObjectWeb ASM at all? 17:49
toolforger Sort of - I've been using libraries that use it for several years, and been looking at the docs at various times
I know my way around the JVM, though I did some brush-up around class file formats and such 17:50
17:50 domidumont left
toolforger Plus did a diagonal read of all the ASM manual pages 17:51
MasterDuke_ well, if you're interested in a hard problem, i've been trying to get r-j to support the original file+line annotations that i added to r-m 17:52
irclog.perlgeek.de/perl6-dev/2016-...i_13602100 has some chat about it (and the previous/next day or two)
toolforger Mmm... debug info should be easy to add 17:53
it might be hard to convince tools to pick up Perl sources though 17:54
17:54 dakkar left
MasterDuke_ i've got it added, it's reading it back out that's a problem 17:54
toolforger Wrong extension, and most likely wrong subdirectory, too
toolforger goes checking the old chat
any keywords I should search the chatlogs for? 17:55
phatbot RabidGravy++ good post
MasterDuke_ i'm adding ASM Attributes to the bytecode with file/line info 17:56
toolforger OK, that would be the way to do it
MasterDuke_ asm, line directives, attributes
around the timeframe of that link maybe plus/minus a couple days
[Coke] (#perl6-jvm) (#perl6-dev is probably fine for that, but if you do create another perl6- chat, be sure to mention it to moritz & mst.) 17:57
toolforger Did you try triggering a Java exception? The stack trace should show you what the JVM thinks about the info
MasterDuke_ well, i got an arrayaccessoutofbounds exception without trying to trigger one 17:58
*arrayindex
toolforger that should be easier to diagnose for mistakes than e.g. Eclipse simply not displaying the right file
MasterDuke_ but it wasn't very helpful
toolforger You can simply "throw new RuntimeException()" anywhere 17:59
then look how the stack trace looks like
MasterDuke_ i have to head out, but i'll be back on this evening
toolforger If the stack trace is too short, somebody up the call stack is not rethrowing the exception
I saw nqp-j has issues with dealing with exceptions that cross the Java/Perl boundary, maybe that's what you are (not) seeing 18:00
cu then :-)
MasterDuke_ before i leave, this is the code/line that was throwing the array exception: github.com/MasterDuke17/nqp/blob/s....java#L331 18:02
toolforger Oh, that's inside the JDK
I'd need to see the stack trace to say anything meaningful about that 18:03
samcv is UNPACK base something that should be documented in the docs? 18:04
it's not in roast or anywhere in the docs or in the design docs
toolforger Or maybe the array access is in NQP's class loaders, though I cannot remember having seen it use arrays
The kind of code in line 133 itself cannot throw an ArrayBoundsException, anyway 18:05
18:05 pyrimidine joined
phatbot samcv: what's UNPACK base? 18:06
grep -R UNPACK src/ gives me nothing
samcv err UNBASE
phatbot samcv: no, it should'nt be documented. It'
samcv: no, it should'nt be documented. It's an internal sub that should be moved to Rakudo::Internals. IIRC it's used by grammar to handle rad_number token
samcv ah ok 18:07
phatbot samcv: we have parse-base() as user-facing version
samcv gotcha
phatbot I think the convention is the ALLCAPS subs are all internal and we just haven't had a chance to move them all to Rakudo::Internals yet
samcv ah ok 18:08
18:10 pyrimidine left 18:11 prammer joined
phatbot m: CORE::.values.grep(Sub).grep({try .name.uc eq .name})».name.say 18:12
camelia rakudo-moar 1d4600: OUTPUT«WARNINGS for <tmp>:␤Useless use of "eq" in expression ".uc eq .name" in sink context (line 1)␤(RETURN-LIST GATHER GENERATE-ROLE-FROM-VALUE ORDER METAOP_TEST_ASSIGN:<&&> METAOP_HYPER_POSTFIX DUMP VAR DEPRECATED METAOP_HYPER_CALL MD-ARRAY-SLICE-ONE-POSIT…»
phatbot useless use of useless use 18:13
18:13 pyrimidine joined
phatbot m: say WHAT 42 18:13
camelia rakudo-moar 1d4600: OUTPUT«(Int)␤»
phatbot m: say HOW 42
camelia rakudo-moar 1d4600: OUTPUT«Perl6::Metamodel::ClassHOW.new␤»
phatbot I wonder why HOW doesn't show up in that list, but WHAT does 18:14
And these aren't internal despite all caps, it's the sub versions of .WHAT .HOW etc ummm.. what do we call them? pseudo-methods?
[Coke] discovers "leider hosen"
phatbot What's that? 18:15
[Coke] macro-ish]
phatbot: apparently "Unfortunately, pants."
as opposed to lederhosen.
toolforger "leather pants"
phatbot And I thought a ladder made of hoses...
toolforger "leider" = "alas" 18:16
18:16 ccntrq left
toolforger "ladder" = "Leiter" 18:17
18:17 zakharyas joined
[Coke] I like "Alas, pants." better. :) 18:17
toolforger Also, "Leiter" = "head" (e.g. of company) 18:18
So may be "Leiterhosen" might be "pants of the leader"
18:18 ccntrq joined 18:19 domidumont joined 18:20 ccntrq left 18:21 andrzejku joined
phatbot m: .say for 42.?dasdasdsa. 18:21
camelia rakudo-moar 1d4600: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Malformed postfix call␤at <tmp>:1␤------> 3.say for 42.?dasdasdsa.7⏏5<EOL>␤»
phatbot m: .say for 42.?dasdasdsa
camelia rakudo-moar 1d4600: OUTPUT«Nil␤»
phatbot Is there a nice-lookin' trick to make it loop only over stuff provided by the method, excluding Nil 18:22
18:22 ccntrq joined 18:32 FROGGS joined 18:34 ccntrq left 18:35 ccntrq joined 18:38 thalunil joined 18:40 domidumont left, dj_goku joined, dj_goku left, dj_goku joined 18:41 kyclark left 18:42 ccntrq left 18:44 ccntrq joined 18:46 dj_goku left, labster joined, ccntrq left 18:47 ccntrq joined 18:48 lucasb left
phatbot m: Nil.head 18:50
camelia rakudo-moar 1d4600: OUTPUT«Cannot resolve caller head(Nil: ); none of these signatures match:␤ (Any:D $: *%_)␤ (Any:D $: Cool $n, *%_)␤ in block <unit> at <tmp> line 1␤␤»
phatbot wonders why that fails
18:51 prammer left
phatbot s: [], 'head', \() 18:53
SourceBaby phatbot, Sauce is at github.com/rakudo/rakudo/blob/1d46...s.pm#L1880
b2gills You know whats annoying Ideone has Perl 6, but it is the 2014.07 version. I asked in the middle of June for it to be upgraded, they got back to me in July, and they still have the old version. 18:54
phatbot m: Nil.tail 18:55
camelia rakudo-moar 1d4600: OUTPUT«Cannot resolve caller tail(Nil: ); none of these signatures match:␤ (Any:D $: *%_)␤ (Any:D $: Cool $n, *%_)␤ in block <unit> at <tmp> line 1␤␤»
phatbot another one :/
lizmat samcv: undocumented ALLCAPS subs should remain undocumented: they will need to be moved to Rakudo::Internals at some point 18:56
El_Che . 18:57
18:57 wamba left 18:58 ccntrq left
samcv has rakudo been released yet for the month? 18:58
i'm using Readline and the REPL keeps like. not even coming up sometimes
so that commit adding Readline history for REPL i think should be reverted before release
i forget when we release again
phatbot samcv: no, it's this Saturday. How can I reproduce the issue? Just install Readline?
NeuralAnomaly: status
samcv oh ok
NeuralAnomaly phatbot, [✘] Next release will be in 3 days and 10 hours. Since last release, there are 50 new still-open tickets (6 unreviewed and 0 blockers) and 43 unreviewed commits. See perl6.fail/release/stats for details
samcv then i will try and fix it then :)
phatbot samcv: release is 3rd Saturday of the month
samcv++ ok cool :) 18:59
19:00 ccntrq joined
phatbot Ah, doh 19:01
19:01 prammer joined
phatbot the head/tail are an Any and Nil is an any too 19:01
[Coke] b2gills: sending them a feedback now. 19:02
explaining tht a lot's been done and I can help if they need.
nine ~~ 19:07
yoleaux 17:26Z <SmokeMachine> nine: Hi! could you help me? Could you tell me if it's working as expected? use a module with a singleton in 2 different files (one of then at compile time) should return different object instances? like that: gist.github.com/FCO/1561729d83b673...1d90199184
nine SmokeMachine: Precompilation can be tricky. Your singleton module is precompiled first. $instance (an empty scalar container) is stored in the precomp file. Use is precompiled next and causes $instance to be updated. But nothing changes the stored precomp file with the empty container. 19:08
SmokeMachine: when you then load ToUse and Use, all you get is an empty container. But how could it be otherwise? E.g. what would happen if two modules both get independently precompiled and both cause a singleton to be created. If your main program loads both these modules, which singleton object would you get? 19:09
phatbot m: say "$2" 19:12
camelia rakudo-moar a980eb: OUTPUT«Use of Nil in string context␤ in block <unit> at <tmp> line 1␤␤»
phatbot Um, looks like REPL wants fixing for warnings. That gives me "Control flow commands not allowed in toplevel" 19:13
19:13 thayne joined 19:14 kyclark joined 19:16 MasterDuke_ left, thalunil left 19:17 prammer left
phatbot samcv: FWIW, I'm not able to reproduce any hangs in my REPL with Readline installed. 19:18
I also learned that it's a bad idea to create ~/.inputrc and fill it with random crap
19:18 thalunil joined
phatbot clicks imaginary "Take" on the REPL warnings stuff 19:20
19:21 domidumont joined, thayne left 19:23 thayne joined 19:24 darutoko left, thayne left 19:25 thalunil left
phatbot hehe... down a rabbit hole 19:26
m: CONTROL { .gist.say }; emit; say "haha, no one is deadied";
camelia rakudo-moar a980eb: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Calling emit() will never work with declared signature (\value --> Nil)␤at <tmp>:1␤------> 3CONTROL { .gist.say }; 7⏏5emit; say "haha, no one is deadied";␤»
phatbot I mean 19:27
m: CONTROL { .gist.say }; emit 42; say "haha, no one is deadied";
camelia rakudo-moar a980eb: OUTPUT«Died with CX::Emit␤ in block <unit> at <tmp> line 1␤␤emit without supply or react␤ in block <unit> at <tmp> line 1␤␤»
19:28 effbiai left 19:29 effbiai joined 19:30 prammer joined 19:32 asd_ joined
asd_ hm... 19:32
19:32 asd_ left
phatbot heh 19:33
RabidGravy so I guess I'd better start the next advent post now 19:37
phatbot You're making two?
huggable: advent 19:38
huggable phatbot, github.com/perl6/mu/blob/master/mi...6/schedule
RabidGravy yeah, got a slot for sunday
phatbot RabidGravy++
interesting... now I rebuilt Rakudo and getting the hangs samcv mentioned :/ 19:39
samcv :/
method history-file() returns Str { ... } what 19:40
oh no
phatbot ?
.seen cygx
yoleaux I saw cygx 9 Dec 2016 14:23Z in #perl6: <cygx> first draft: perl6advent.wordpress.com/?p=5746&...eview=true
samcv that's in the REPL since the PR for adding the history file for readline... 19:41
that shouldn't do anything right unless there's some ... magic i don't know about in perl 6
19:41 dj_goku joined, dj_goku left, dj_goku joined
samcv oh linenoise has it set the same way hm 19:41
phatbot it's a stub
m: class { method meow { … } }.meow 19:42
camelia rakudo-moar a980eb: OUTPUT«Stub code executed␤ in method meow at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤Actually thrown at:␤ in block <unit> at <tmp> line 1␤␤»
samcv yeah i know that
phatbot Oh ok
How do you know so much Perl 6?
samcv me?
phatbot Yeah
samcv idk
learning it 19:43
RabidGravy how does anyone know so much Perl 6? ;-)
phatbot No, I mean I don't recall seein samcv here for more than a month, but in some things she knows more than me 19:44
samcv uhm ok it looks like history-file is a method that is defined in a different spot
phatbot I guess that's more a statement about me than anyone else heh
19:44 pyrimidine left
samcv i learn quickly ;) 19:44
i only started programming perl 5 5 months ago
19:45 pyrimidine joined
samcv my friends i've talked to are super suprised that even tho they've been perling for years i now surpass them haha 19:45
maybe i'm finding what i'm good at.. never been employed as a programmer. but i really enjoy it
phatbot :) 19:46
RabidGravy don't fancy a job? I'm going to be the only developer in a team that should have five in the new year
19:46 dj_goku left
samcv don't fancy? well i am looking for work atm 19:46
maybe i could do something in devops cause my last job was Jr Linux admin 19:47
samcv really needs to get a job
RabidGravy where are you based?
samcv in the bay area, in california 19:48
maybe i can point to contributing to perl 6 and say. pls give me job! i am good! i swear
19:48 dwarring joined
nine samcv: there are definitely worse ways to advertise than having public open source contributions 19:49
19:49 pyrimidine left
samcv like? lol 19:49
RabidGravy bit of a commute to London then :)
samcv RabidGravy, i will drive really fast 19:50
nine samcv: like....anything else *g*
samcv that's what i was thinking...
ugh so annoying. the perl language package in atom in built in, but it will go first before the perl 6 one... since it has perl 6 file detection too 19:51
nine If I've ever got to apply for a job again, I'll just point at the list of my conference talks and see who bites ;)
samcv so if i disable it, it detects the rakudo files... but if i do that i can't look at the perl files
ha
really liking it now that i got perl 6 code execution only a keyboard shortcut away :) 19:52
phatbot Yeah, I recall Madcapjake telling me I had to disable the built-in perl highlighter if I wanted perl6fe to work 19:53
samcv just really like automating things at least. i've been doing that a lot
you don't have to disable it, but
if it's not named .p6 or .pl6 or .nqp or whatever 19:54
the perl plugin will try and do its bad perl 6 highlighting
Xliff Is there any way I can block the default constructor of a class?
phatbot Ah
Xliff m: class A { has $.n; multi method new($a) { $!n = 1; }; multi method new($a, $b) { $!n = 2 }; }; my $a = A.new(:n(3));
camelia ( no output )
RabidGravy which reminds me I need to update the vim perl 6 syntax file at some point
phatbot Xliff: proto method new (|) {*}
19:54 djbkd_ joined
samcv i added some detection to detect if it starts with the perl 6 shebang or 'my class' and detect as perl 6 instead of perl 5, but 19:54
the perl plugin does that too 19:55
Xliff phatbot++ # Thanks!
samcv and goes first
dunno if i can get them to disable that, or something..
RabidGravy, it should just come with it
unless they updated it since then
19:56 loveperl joined 19:57 wamba joined
RabidGravy samcv it does, it just doesn't match the current state of the language 19:58
samcv yeah
you should try the atom one ;)
and if there's _any_ syntax highlighting errors please open an issue and i'll fix it quickly 19:59
unhighlighted methods, routines, etc even tiny things
19:59 andrzejku left
Xliff Hmm... phatbot, that works, but the more tricky among us might know that perl6 will allow constructor of the form A.new(1, :n(4)) -- don't want anyone to have the ability to set $.n via new. 20:00
samcv also the Readline plugin still has a problem it doesn't work with readline 7
it hardcodes version 6 in
Xliff samcv: I am using atom perl6-fe and loving it.
samcv nice!
\o/
Xliff I also managed to convert the snippets to something usable for Sublime.... although I have not texted that. ;)
s/texted/tested/ 20:01
phatbot Xliff: it won't get set
loveperl How did you call a class that is in another file?
Xliff Also: github.com/MadcapJake/language-per.../issues/53
phatbot Xliff: and well, reading your multi method new()... it's wrong. It needs to self.bless
Xliff phatbot: LOL! Yeah. Fixed that already in a camelia window. 20:02
m: class A { has $.n; proto method new (|) {*}; multi method new($a) { self.bless(:n(1)); }; multi method new($a, $b) { self.bless(:n(2)) }; }; my $a = A.new(1, :n(4));
camelia ( no output )
phatbot Xliff: right, so that's what will only be set. The extra args would be ignored. Any methods can take any number of named args
Xliff So how can I block the constructors from allowing $.n to be set like that?
Ooh! 20:03
phatbot m: say 'lulz'.uc: :42a :72meows
samcv phatbot, just made a PR to fix readline
camelia rakudo-moar a980eb: OUTPUT«LULZ␤»
samcv err the REPL
Xliff m: class A { has $.n; proto method new (|) {*}; multi method new($a) { self.bless(:n(1)); }; multi method new($a, $b) { self.bless(:n(2)) }; }; my $a = A.new(1, :n(4)); say $a.n;
camelia rakudo-moar a980eb: OUTPUT«1␤»
Xliff Right.
20:03 cygx joined
samcv Xliff, see the current readme for Perl6FE for how to get perl 6 scripts running from atom 20:03
cygx has scheduled his advent post
yoleaux 11:43Z <phatbot> cygx: reminder: your post is up next in ~12.5 hours
Xliff REPL... REPR... what's the difference?
phatbot Xliff: but if you want it to actually whine about the extra args, just add %_ and die "no args!"
cygx: \o/
cygx I'm not entirely happy with it, but it is what it is... 20:04
phatbot That's what my mom said about me!
samcv Xliff, github.com/perl6/atom-language-per...ntegration
Xliff samcv++ # \o/ 20:05
samcv :)
Xliff If only Atom was working under my Linux VM.
samcv you don't even have to save the file too
:(
what is the host?
Xliff Currently I am using it with Windows via SMB export to my host.
Windows
samcv atom is for windows too
Xliff I know. That's the only way I can use atom right now.
samcv ohhh
oh i see the problem i guess
perl 6 not working under windows properly?
Xliff I can't get it to run under Linux for some reason. The GUI will not start.
No. My perl6 install is on the linux vm. 20:06
samcv kk
Xliff So I export home dir to Windows so I can use Windows Atom for editing.
samcv ah
cygx our wordpress theme looks sub-optimal, though: it's too narrow, so code blocks end up with horizontal scroll bars :(
Xliff Sublime is working under Linux VM, so I will try getting the TextMate rules over there to see if they work better than ModernPerl.
20:06 AlexDaniel joined
samcv Xliff, maybe hack something together? 20:07
Xliff samcv: And I do have to say, Atom+language-perl6fe > Sublime+ModernPerl
samcv idk can you make a perl6 names executable which gets the vm to run it
Xliff samcv: Actually, I already have the .sublime_package. I just have been too lazy to try it!
samcv and returns the result ha
what .sublime_package?
Xliff language-perl6fe
samcv how did you create it 20:08
Xliff I converted it.
samcv how? pls tell
Xliff cson => json => plist => zip
samcv you're using the updated plugin too right? cause the dl package moved
Xliff There are scripts.
samcv hmmmmm
links
Xliff But I haven't tested it under Sublime, yet.
Yeh. I am using the repo.
samcv kk
tbrowder Anyone using Komodo for p6? 20:10
Xliff samcv: Actually, it is more convoluted. 20:11
cson => json => yaml => plist => zip 20:12
samcv well send it to me
i will test it
nice
haha
Xliff Uh. OK. I will create issue.
And attach package.
Or can I email you?
samcv you can email me
Xliff kk
samcv pm'd you
phatbot Xliff: if you make sublime highlight perl6 decently, I'll promise not to assimilate you :) 20:13
sublime > atom
samcv kk 20:17
yeah i need perl 6 highlighting to be happy tbh
AlexDaniel phatbot: source 20:18
yoleaux 05:06Z <MasterDuke> AlexDaniel: it looks like you implemented something already, but i'd say just return the name for them or something like that, i don't see anybody ever wanting them for real
samcv github.com/drforr/perl6-readline/pull/15 hmm if this can be pulled would we want to include Readline in Star? 20:19
because linenoise doesn't support utf-8
phatbot AlexDaniel: github.com/zoffixznet/perl6-phatbot
samcv oh AlexDaniel
you need to add something to unicodable6
AlexDaniel MasterDuke: the problem is, if you do ｢say ‘hello’ ~ 0.chr ~ ‘world’｣ or something like that, what should committable/evalable gist?
Xliff phatbot: LOL! You'll try to do that anyways. I wonder what else I can get out of you. :D 20:20
RabidGravy Hmm, now do I either split some of the utility functions of Test::META into a separate module or make Test::META a dependency the thing I want to use them in
AlexDaniel phatbot: bug tracker
the link is broken
I wanted to submit a PR that will change the nickname…
samcv: alright, what?
phatbot AlexDaniel: github.com/zoffixznet/perl6-phatbot/issues
AlexDaniel samcv: dumping of all unicode properties 20:21
?
samcv no
well sorta
m: for 0..0x1FFFF -> $i { if $i.unimatch('LF') { say $i.uniname } } 20:22
camelia rakudo-moar a980eb: OUTPUT«LINE FEED (LF)␤»
samcv the LF is a unicode property called the 'Line_Break' property 20:23
20:23 Tonik joined
samcv unimatch will check if the codepoint has the specified property 20:23
AlexDaniel u: LF
unicodable6 AlexDaniel, U+000A LINE FEED (LF) [Cc] (control character)
AlexDaniel, U+00BD VULGAR FRACTION ONE HALF [No] (½)
AlexDaniel, U+02BE MODIFIER LETTER RIGHT HALF RING [Lm] (ʾ)
AlexDaniel, gist.github.com/dc202f8aaf0aa2ec27...2457309795
samcv LF is a bad example tbh
AlexDaniel u: LF Cc
unicodable6 AlexDaniel, U+000A LINE FEED (LF) [Cc] (control character)
phatbot AlexDaniel: what's the new nickname? 20:24
AlexDaniel samcv: alright. OK. What could be the syntax?
phatbot: Zoffix
samcv m: for 0..0x1FFFF -> $i { if $i.unimatch('Latin') { say $i.uniname } }
camelia rakudo-moar a980eb: OUTPUT«LATIN CAPITAL LETTER A␤LATIN CAPITAL LETTER B␤LATIN CAPITAL LETTER C␤LATIN CAPITAL LETTER D␤LATIN CAPITAL LETTER E␤LATIN CAPITAL LETTER F␤LATIN CAPITAL LETTER G␤LATIN CAPITAL LETTER H␤LATIN CAPITAL LETTER I␤LATIN CAPITAL LETTER J␤LATIN …»
samcv uh just search it when searching other things
phatbot AlexDaniel: I can't use that one. I have to stay undercover
AlexDaniel u: { .unimatch(‘Latin’) }
samcv or maybe not uh
unicodable6 AlexDaniel, U+0041 LATIN CAPITAL LETTER A [Lu] (A)
AlexDaniel, U+0042 LATIN CAPITAL LETTER B [Lu] (B)
AlexDaniel, U+0043 LATIN CAPITAL LETTER C [Lu] (C)
AlexDaniel, gist.github.com/27443d7c8b9213125e...0950abb8e7
phatbot u: { .unimatch("{use NativeCall; sub system (Str) is native {}; system 'ls'}") } 20:25
RabidGravy tadzik++ # mergey goodness
samcv but it will find all properties no matter what they are as long as moarvm has it i guess
AlexDaniel phatbot: :|
20:26 unicodable6 left
phatbot >:} 20:26
AlexDaniel it should say something like ‘meh, fuck off’
samcv AlexDaniel, you could just search it when searching others
it won't slow it down really
tho you could do a special case too
m: for 0..0x1FFFF -> $i { if $i.unimatch('9.0') { say $i.uniname } }
camelia rakudo-moar a980eb: OUTPUT«<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved>␤<reserved…»
AlexDaniel samcv: sure, but it's already possible now, and unless you can propose some syntax I'm not sure I will be able to add it :) 20:27
20:27 domidumont left
AlexDaniel phatbot: aren't you supposed to bring them back on when you crash them? 20:27
phatbot OK will do
20:27 unicodable6 joined
AlexDaniel no need 20:27
samcv u: match 'thing' 20:28
unicodable6 samcv, U+006D LATIN SMALL LETTER M [Ll] (m)
samcv, U+0061 LATIN SMALL LETTER A [Ll] (a)
samcv, U+0074 LATIN SMALL LETTER T [Ll] (t)
samcv, gist.github.com/a6a297d998af3ae114...d7d96ece98
samcv maybe like that?
idk something with the match word in it would be easiest to remember
AlexDaniel phatbot: I'm still waiting for a day I get at least some small bit of time to work on containerizing them… until that happens, can you please not advertise the holes in our RESTRICTED setting pseudofeature? :) 20:29
phatbot: or… perhaps consider disabling nativecall if it is on 20:30
phatbot You mean the NativeCall one or the nqp one? :D
AlexDaniel … both
phatbot ehehe
20:30 FROGGS left
loveperl Could not find form at line 1 in: 20:34
use form 20:35
mst AlexDaniel: perhaps talk to simcop2387 about how he's locked down perlbot's eval system
loveperl ?
mst AlexDaniel: I mean, 'stick it in a container' won't help
loveperl: you tried to load a module that doesn't exist
AlexDaniel mst: wait… why not?
simcop2387 rlimits, chroot, seccomp. I can give a better overview if you'd like
phatbot :o
mst AlexDaniel: you want security. containers are not designed for security.
phatbot simcop2387!
simcop2387 the one and only 20:36
phatbot \o/
mst AlexDaniel: every time somebody goes "I'll use docker for security", apeiron eats a kitten
AlexDaniel mst: sooo, are you saying that LXC is insecure or something?
20:36 toolforger left
simcop2387 AlexDaniel: LXC and docker won't protect you from kernel exploits. 20:36
loveperl Do I want to call a class that is in a file that is called form.p6? 20:37
mst AlexDaniel: ... no, I'm saying it's not designed for the sorts of things you'd want to do to secure an eval server
loveperl Do I want to call a class that is in a file that is called form.p6?
form.pm6
simcop2387 seccomp can got a LONG LONG way to making it mostly impossible to do most exploits
mst loveperl: trout.me.uk/vaguecat.jpg
20:37 matiaslina left
loveperl lol 20:38
mst AlexDaniel: I'm not sure if you're intentionally being a dick or just completely clueless here, but responding to "this is not designed for securing what you're securing" does not mean it's "Insecure or something", it just means you don't know what you're doing yet. maybe ask the sysadmin of the machine they run on for help?
AlexDaniel simcop2387: alright. alright… good that you're here
20:39 ccntrq left
mst loveperl: show us all the code you currently have and explain what you're trying to do? 20:39
AlexDaniel mst: well, that's kinda my machine…
simcop2387 AlexDaniel: here's the code i do in perlbot's eval to secure things with seccomp, github.com/simcop2387/perlbuut/blo...#L31..L106 , the idea is that it doesn't let the user supplied code open anything for writing, open any sockets etc. they're only allowed to do a few things.
mst AlexDaniel: then I recommend listening to simcop2387 and not being deliberately obtuse to him like you were to me, since he's a nicer person than I am :P 20:40
AlexDaniel mst: stop accusing me of shit, I'm listening already
simcop2387 i won't abuse anyone (unless they ask for it nicely) but i will start ignoring people.
20:42 ccntrq joined
loveperl call.pl : code=( use form; say form) form.pm6 is: code=( class form extends common{}) 20:42
AlexDaniel simcop2387: I'd have to read about seccomp to get the idea about this stuff. So I hope you don't mind if I ping you a bit later
phatbot loveperl: the current directory is not included by default. you need use lib <.>; or supply -I. on command line
simcop2387 AlexDaniel: oh no problem at all.
20:43 dj_goku joined
simcop2387 AlexDaniel: This one might or might not affect you, but i gave the perl interpreter CAP_SYS_CHROOT with `setcap cap_sys_chroot+ep $PERL`, that one let me do everything running from nobody:nogroup without ever touching the root user. probably not THAT big of a deal but it helps against the possibility of some module i'm using having code sitting around specifically for me to hit. 20:44
loveperl I do not understand?
simcop2387 AlexDaniel: also, if you're not familiar with strace, it's a big help in understanding what you can do with seccomp. basically you can filter system calls that show up in strace based on the arguments to them. 20:45
phatbot loveperl: perl6 -I. call.pl 20:47
loveperl: you need to tell it where to find form.pm6 and common.pm6 20:48
loveperl How do I say that? 20:49
20:49 thalunil joined
phatbot loveperl: Poner lib <.> En su archivo para decir perl6 para buscar los módulos en el directorio actual 20:50
phatbot relocates
AlexDaniel mst: thanks 20:52
20:54 dj_goku left
loveperl asi 20:56
lib <.>; 2 use form; 3 say form;
20:57 bartolin left
AlexDaniel mst: by the way, any news about the button? 20:59
loveperl =begin ⏏comment?
I get error 21:00
How to comment en bloc? 21:01
=begin comment; brief utilities for the creation of HTML forms. with special datatypes not found in HTML like date, etc todo fieldset+legend, label for field todo BLOCKs (field.block?) todo block show/hide =end comment;
mst loveperl: please use fpaste.scsys.co.uk/perl6 to show complete code
21:01 nowan_ joined
mst AlexDaniel: I'm probably going to have to email somebody, will try and remember to do that this week 21:01
21:01 bartolin joined
AlexDaniel mst: that would be nice 21:02
mst AlexDaniel: at this point I suspect "attempt to invoke a higher power" is going to be the right answer
shadowpaste "loveperl" at 217.168.150.38 pasted "como comentar en perl" (7 lines) at fpaste.scsys.co.uk/540217
mst key question: are you ok if it's "registered users only" and are you ok if we have to ask on a per-user basis?
(trying to pre-anticipate compromises that might be easier to win)
loveperl how to comment eb bloc? 21:03
How to comment in block?
fpaste.scsys.co.uk/540217?tx=on&...rmat+it%21
mst loveperl: what happens if you remove the ';' and just have '=begin comment' ? 21:04
loveperl: ... wait, there's one space in front of the = in your code. no spaces!
21:04 nowan left
phatbot mst, IIRC it was already registerer users only so that's fine. The other option isn't ideal but better than nothing 21:05
loveperl tank you
AlexDaniel mst: “registered users only” is how it has always been
mst AlexDaniel: right, so maybe it needs some sort of additional whitelist
21:06 xinming_ joined
AlexDaniel mst: not sure if that would be easy, but perhaps checking if a user has at least one open or resolved ticket is good enough 21:06
loveperl class form extends common{ 21:07
AlexDaniel or a script that builds a whitelist based on this, that could work too
mst loveperl: please always use the paste site to show code.
loveperl: random lines in channel do not help
loveperl ok 21:08
SmokeMachine nine: so, I cannot have a trait that saves data on a singleton because I'll never get that singleton again?
21:09 xinming left
jnthn loveperl: If you are trying to do inheritance, you may find docs.perl6.org/language/classtut#Inheritance helpful 21:09
21:09 thalunil is now known as thayne
bartolin AlexDaniel: you pinged me to rebase rakudo pr 565. I can do that, but I wonder what git commands are best to do that. should I merge current nom into my old branch (resolving the conflict while doing that) and push the updated branch to my repo? 21:10
RabidGravy SmokeMachine, sure you can
SmokeMachine RabidGravy: not inter files... 21:11
RabidGravy eh?
SmokeMachine nine: when the second file uses the module it's gotten by precompiled file? Not from the preloaded module in memory with the $instance? 21:13
bartolin AlexDaniel: or is it acceptable to just force push something without a conflict?
SmokeMachine RabidGravy: this: 17:07 <yoleaux> 17:26Z <SmokeMachine> nine: Hi! could you help me? Could you tell me if it's working as expected? use a module with a singleton in 2 different files (one of then at compile time) should return different object instances? like that: gist.github.com/FCO/1561729d83b673...1d90199184 21:14
AlexDaniel bartolin: you can force push into your own branch, yes
phatbot bartolin, you can forcepush into your own branch the PR is from
AlexDaniel in your own fork rather
loveperl constructor? 21:15
bartolin thanks! I'll look into it (but not today)
loveperl Constructor of a class 21:16
nine SmokeMachine: maybe you want to initialize your singleton in an INIT block? 21:18
21:20 RabidGravy left
SmokeMachine nine: in my real case of use I do that on a trait... or do you mean initialize the $instance on a INIT block? 21:20
21:21 setty1 left
SmokeMachine nine: this is the case: github.com/FCO/Test-Fuzz/blob/mast...st/Fuzz.pm 21:22
shadowpaste "loveperl" at 217.168.150.38 pasted "constructor" (26 lines) at fpaste.scsys.co.uk/540218
21:23 loveperl_ joined
loveperl_ How do I build a builder? 21:23
fpaste.scsys.co.uk/540218?tx=on&...rmat+it%21
SmokeMachine nine: I'm sorry the leg.. I'm on my doughter's ballet presentation 21:24
21:24 RabidGravy joined 21:25 loveperl left 21:26 zakharyas left, Tonik left 21:27 imcsk8_ is now known as imcsk8 21:30 thayne left
dwarring m: role R {}; my $a = 42 but R; say 42.WHICH 21:35
camelia rakudo-moar 7b84ad: OUTPUT«Int|42␤»
dwarring m: role R {}; my $a = 42 does R; say 42.WHICH
camelia rakudo-moar 7b84ad: OUTPUT«Int+{R}|42␤»
dwarring ^^ shot myself in the foot the other day 21:36
used 'does' on simple scalars 21:37
has a very global affect
cygx there's an integer cache, so stuff like that can indeed happen 21:39
dwarring yeah, a but of a trap
21:41 lukiramu joined
cygx I believe early Fortran version even had the ability to inadvertently redefine numbers dues to by-reference argument semantics 21:42
21:42 bjz left
dwarring heh, like redefining 0 in fortran 21:43
masak in Perl 6, all we can do is redefine 2 + 2 21:44
but in a very safe, lexically scoped way :)
21:44 lukaramu left
masak m: sub infix:<+>(2, 2) { 5 }; say 2 + 2 21:44
camelia rakudo-moar 7b84ad: OUTPUT«5␤»
masak oh, I guess that should be spelled `multi` for clarity.
geekosaur sounds like stupid haskell let tricks :p
masak m: multi infix:<+>(2, 2) { 5 }; say 2 + 2 21:45
camelia rakudo-moar 7b84ad: OUTPUT«5␤»
21:45 rindolf left
cygx m: 2 does role { method gist { "42" } }; say 2 21:46
camelia rakudo-moar 7b84ad: OUTPUT«42␤»
21:46 pyrimidine joined 21:47 kurahaupo_ joined 21:48 kurahaupo left, kurahaupo_ is now known as kurahaupo 21:49 rindolf joined, dj_goku joined 21:55 cdg_ joined, dj_goku left 21:57 cdg left 21:58 Ven left 21:59 Ven joined, rurban joined 22:00 cdg joined 22:02 skids left, bjz joined 22:04 cdg_ left, prammer left 22:05 cdg left 22:11 kyclark left 22:12 RabidGravy left 22:15 kyclark joined 22:16 kyclark left, bjz left
loveperl_ exit in perl 22:25
?
22:27 mawkish_ left 22:33 kurahaupo left 22:34 MorayJ left 22:37 kyclark joined 22:46 pyrimidine left, pyrimidi_ joined 22:50 dj_goku joined, dj_goku left, dj_goku joined
masak these random comments three years later are sometimes amusing, but rarely very deep: perl6advent.wordpress.com/2013/12/...ment-16238 22:54
22:55 AlexDaniel left 22:56 dj_goku left, rburkholder joined, kyclark left
loveperl_ my $s = any; here null? 22:58
masak m: my $s; say $s 22:59
camelia rakudo-moar 7b84ad: OUTPUT«(Any)␤»
cygx masak: that comment is hilarious 23:01
"the factorial function is worth putting in your product"
23:09 bpmedley joined 23:11 cdg joined
phatbot hehe 23:11
masak: nice one :)
phatbot is glad "infix_prefix_meta_operator:<[ ]>" didn't stick :}
loveperl_: exit is exit 23:13
m: say "hi"; exit; say "this doesn't print"
camelia rakudo-moar 7b84ad: OUTPUT«hi␤»
loveperl_ tank you
phatbot loveperl_: no, `any` is not null; it's a Junction perl6.party/post/Perl-6-Schrodinger...-Junctions 23:14
loveperl_ what is null? 23:15
phatbot loveperl_: depends on what you're trying to do. Perl 6 has infinite number of "nulls". There's Empty, Nil, and all the type objects (Int is undefined; but 42 is, for example) 23:16
b2gills m: say #`(There is a factorial function, its just called) 6.&permutations.Int 23:19
camelia rakudo-moar 7b84ad: OUTPUT«720␤»
phatbot m: say [*] ^6+1; #`(There is a factorial function, its just called) 23:21
camelia rakudo-moar 7b84ad: OUTPUT«720␤»
23:22 AlexDaniel joined 23:25 lukiramu left
samcv omg fuzzy search 23:25
finding wayyyy too many things
it needs to list the fuzzyly found things at the end of the list 23:26
(re the perl 6 docs)
also it seems to be realllyyy slow sometimes
and maxing out the cpu on my computer 23:27
23:27 loveperl_ left 23:28 cdg left
samcv try typing 'run' into it. it justs... kinda 23:28
doesn't do anything and uses all my cpu
oh hey it finally showed up... but has a billion pages all returned in that scrolling menu.. 23:29
i think it should search how it normally would and then have a link to click to go to a fuzzily searched list of things if nothing shows up 23:30
AlexDaniel samcv: fix it then! 23:31
this problem has been there for quite a while 23:32
it just got a little bit worse
samcv you mean since it was changed to fuzzy search?
AlexDaniel no, previously if you typed ‘e’ it would have given you waaaaay too many results
so now there are just a bit more ways to get too many results
but same problem
samcv ah 23:33
dalek c: d5c7777 | samcv++ | template/search_template.js:
Change levenshtein distance from 2 to 1

Try to get it for now to not return way too many things for short or common words		 23:35
23:38 kurahaupo joined
samcv ok that change made it like 10x better 23:39
ok maybe not 10x better but usable
AlexDaniel OK: github.com/perl6/doc/issues/1050 23:40
23:43 travis-ci joined
travis-ci Doc build failed. Samantha McVey 'Change levenshtein distance from 2 to 1 23:43
travis-ci.org/perl6/doc/builds/183766447 github.com/perl6/doc/compare/a9722...c77772c72b
23:43 travis-ci left
samcv MoarVM panic: Collectable 0x2b3668c81980 in fromspace accessed :( 23:44
was wondering why that would have broken the docs
23:45 skids joined 23:46 ufobat left
AlexDaniel mmmmm collectable 23:48
23:51 dj_goku joined, dj_goku left, dj_goku joined 23:53 rurban left
AlexDaniel u: Lo 23:53
unicodable6 AlexDaniel, U+00AA FEMININE ORDINAL INDICATOR [Lo] (ª)
AlexDaniel, U+00BA MASCULINE ORDINAL INDICATOR [Lo] (º)
AlexDaniel, U+01BB LATIN LETTER TWO WITH STROKE [Lo] (ƻ)
23:53 cygx left
AlexDaniel … that's it??? 23:53
there are only 3 LO characters???
Lo* 23:54
unicodable6: Lo
no, we have a problem!
unicodable6 AlexDaniel, U+00AA FEMININE ORDINAL INDICATOR [Lo] (ª)
AlexDaniel, U+00BA MASCULINE ORDINAL INDICATOR [Lo] (º)
AlexDaniel, U+01BB LATIN LETTER TWO WITH STROKE [Lo] (ƻ)
AlexDaniel or maybe not
23:54 unicodable6 left
samcv u: { .unimatch('Lo') } 23:55
did i do that right?
AlexDaniel :| it ate too much memory and died
phatbot There are 112804
samcv oh
that's not enough
phatbot m: ^0x110000 .grep(*.uniprop eq "Lo").elems.say
camelia rakudo-moar 7b84ad: OUTPUT«112804␤»
23:55 unicodable6 joined
AlexDaniel alright, so it died trying to gist it, I guess 23:56
or… I don't know
phatbot What does it use to gist?
samcv or get the remaining results?
phatbot Zofware?
AlexDaniel phatbot: your module I think
samcv ugh i got this error just now building docs MoarVM panic: Collectable 0x7f8879020cd0 in fromspace accessed
travis got it too
phatbot samcv: rt.perl.org/Ticket/Display.html?id...et-history 23:57
AlexDaniel yea, Pastebin::Gist
23:57 dj_goku left
AlexDaniel phatbot: by the way, that module does not do any sort of escaping, and it probably should 23:57
23:58 dj_goku joined
samcv escaping? 23:58
AlexDaniel e.g. if you put some control characters in there it will not work
samcv oh
why can't it just gist the control characters
AlexDaniel dunno, let's try something
samcv does it write it to a file before gisting it?
phatbot AlexDaniel: I never updated it since I wrote it and it was my second piece of code I ever wrote in Perl 6.... 23:59
AlexDaniel evalable6: say ‘hello’ ~ 0.chr ~ ‘world’; .say for ^999
evalable6 AlexDaniel, rakudo-moar 7b84adc: OUTPUT«hello␀world␤0␤1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤11␤12␤13␤14␤15␤16␤17␤18␤19␤20␤21␤22␤23␤24␤25␤26␤27␤28␤29␤30…»
AlexDaniel, Full output: gist.github.com/
AlexDaniel samcv: no
samcv it probably should

Powered by Cro and the Raku® Programming Language.

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

Thank you!