»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
00:00 AlexDaniel left
viki kyclark: never mind, it's not a bug. 00:00
Hmm
Or maybe it is.
Here's the crux of the issue: 00:01
m: my $c = "\x[33F]"; say $c.ord; say $c.ord.succ; say $c.ord.succ.chr; say $c.ord.succ.chr.ord;
camelia rakudo-moar 15f51a: OUTPUT«831␤832␤̀␤768␤»
viki it gets normalized or whatever we call it, but that normalized form is a lower .ord, so from that it soon reaches the original form and the loop continues
00:03 eisen74 joined
viki m: "\x[340]\x[300]".uninames.say 00:05
camelia rakudo-moar 15f51a: OUTPUT«(COMBINING GRAVE ACCENT COMBINING GRAVE ACCENT)␤»
viki m: .succ.chr.ord < $_ and say "{.base: 16} is problematic 'cause .succ is {.chr.ord.base: 16}" for ^0x110000 00:09
camelia rakudo-moar 15f51a: OUTPUT«33F is problematic 'cause .succ is 33F␤340 is problematic 'cause .succ is 300␤342 is problematic 'cause .succ is 342␤343 is problematic 'cause .succ is 313␤373 is problematic 'cause .succ is 373␤37D is problematic 'cause .succ is 37D␤386 is pro…»
viki m: .succ.chr.ord < $_ and say "{.base: 16} is problematic 'cause .succ is {.succ.chr.ord.base: 16}" for ^0x110000 00:10
camelia rakudo-moar 15f51a: OUTPUT«33F is problematic 'cause .succ is 300␤340 is problematic 'cause .succ is 301␤342 is problematic 'cause .succ is 313␤343 is problematic 'cause .succ is 308␤373 is problematic 'cause .succ is 2B9␤37D is problematic 'cause .succ is 3B␤386 is prob…»
viki m: my $x = 0; .succ.chr.ord < $_ and $x++ for ^0x110000; say "$x problematic codepoints"
camelia rakudo-moar 15f51a: OUTPUT«1094 problematic codepoints␤»
viki So basically such issue exists anytime your sequence has any of these 1094 codepoints in the middle of it 00:11
m: say ("\x[33F]" ... "\x[341]").elems
camelia rakudo-moar 15f51a: OUTPUT«63␤»
viki ... why didn't that hang :/ 00:12
m: say ("\x[33F]" ... "\x[341]")
camelia rakudo-moar 15f51a: OUTPUT«(̿ ̾ ̽ ̼ ̻ ̺ ̹ ̸ ̷ ̶ ̵ ̴ ̳ ̲ ̱ ̰ ̯ ̮ ̭ ̬ ̫ ̪ ̩ ̨ ̧ ̦ ̥ ̤ ̣ ̢ ̡ ̠ ̟ ̞ ̝ ̜ ̛ ̚ ̙ ̘ ̗ ̖ ̕ ̔ ̓ ̒ ̑ ̐ ̏ ̎ ̍ ̌ ̋ ̊ ̉ ̈ ̇ ̆ ̅ ̄ ̃ ̂ ́)␤»
viki oh 341 is also messed up heh 00:13
m: say ("\x[A5A]" ... "\x[A5C]").elems 00:14
camelia rakudo-moar 15f51a: OUTPUT«(timeout)»
viki m: say (*.parse-base(16) - *.parse-base(16))("A5C", "A5A")
camelia rakudo-moar 15f51a: OUTPUT«2␤»
viki hehe
Well, I'd call that a bug :)
Or is it? 00:15
It's Quantum Bug... It's both a bug and ain't one ;)
00:15 eisen74 left 00:23 cdg left
viki God damn.. Digital Extremes ruined my day today :( 00:24
They changed a thing in a game and I don't like it...
viki goes to fix Rakudo bugs instead
geekosaur disappointment-driven development
viki heh 00:25
00:32 bjz joined 00:35 labster left 00:55 aborazmeh joined, aborazmeh left, aborazmeh joined 00:59 seatek joined
viki m: nqp: nqp::say(nqp::islt_n(nqp::div_n(1e0,0e0), 0e0)) 01:04
camelia rakudo-moar 15f51a: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Could not find nqp::div_n, did you forget 'use nqp;' ?␤at <tmp>:1␤------> 3nqp::say(nqp::islt_n(nqp::div_n(1e0,0e0)7⏏5, 0e0))␤»
viki nqp: nqp::say(nqp::islt_n(nqp::div_n(1e0,0e0), 0e0))
camelia nqp-moarvm: OUTPUT«0␤»
viki nqp: nqp::say(nqp::islt_n(nqp::div_n(1e0,-0e0), 0e0))
camelia nqp-moarvm: OUTPUT«1␤»
viki Don't need an op. ^ that is pretty obvious (compared to atan2)
wc
01:05 sufrostico left 01:06 labster joined 01:10 kyclark_ joined 01:12 adu joined, kyclark_ left 01:14 finanalyst joined 01:18 pierre__ joined, pierre__ left, pierre__ joined 01:19 pierre__ left 01:21 labster left 01:30 Sgeo__ joined
viki is rocking to youtu.be/TGDbkrWXzhw?t=43m39s 01:31
01:31 Sgeo_ left
viki Hacking Party \o/ 01:31
01:37 avalokite joined, avalokite left 01:46 labster joined
viki m: say IO::Handle ~~ Cool 01:48
camelia rakudo-moar 15f51a: OUTPUT«False␤»
viki Is printf(IO::Handle:D, Cool:D $format, *@args) candidate a bad idea? that's adding to current candidate printf(Cool:D $format, *@args) 01:50
MasterDuke (this is not an answer to your question) i have no idea why some things are Cool and others aren't 01:51
viki Cool things make sense as both a number and a string
01:52 labster left
MasterDuke yeah, i'm actually bothering to read the documentation 01:52
viki
.oO( TFW you're scrolling through old logs and wondering if that nick is Zoffix... )
01:53 adu left
viki I'm basically wanting to add IO::Handle printf, because the current docs make me cry ( docs.perl6.org/routine/printf ) 01:53
And to mirror the IO interface, it'd make sense that the printf() sub can do filehandles too. 01:54
MasterDuke FWIW, Perl 5 printf takes a filehandle
viki I think we had a similar talk about comb()
Alright, so there's some exisitng precedent/expectations for my idea. 01:55
MasterDuke comb()ing a filehandle?
viki Nah, defaulting for comb 'fooo' to be comb 1, 'foo' or something like that
MasterDuke seems reasonable 01:57
viki At the end it was ruled that was too magical 01:58
MasterDuke don't know how it's any less magical than 'foo'.comb 01:59
02:04 BenGoldberg joined
zilti did someone succeed to install rakudo on netbsd? 02:15
ugexe i installed rakudo on netbsd on a raspberry pi months ago 02:18
zilti I always get a segfault when compiling, not sure if it's rqp or MoarVM 02:19
02:24 kalkin-_ joined, kalkin- left
ugexe are you asking us to guess, or waiting for someone to ask you to paste the output somewhere? 02:25
02:26 labster joined 02:27 kalkin-_ is now known as kalkin-, wamba left
viki option c: debug on your own and submit a PR fixing the issue \o/ 02:34
02:38 bjz_ joined, araujo left 02:39 araujo joined 02:40 bjz left
zilti ugexe, viki: ah, I did open a ticket: github.com/MoarVM/MoarVM/issues/441 02:43
viki Thanks!
timotimo can we actually work with the core dump? 02:45
like, if we don't have a netbsd of our own?
02:46 ilbot3 left
zilti Uh. Good question. I have no idea what's actually in the core dump. 02:47
ugexe its easier to debug on your own if you know the compiler/3rd party libraries/etc
02:47 ilbot3 joined
timotimo it'd be a good first step if you could just gdb --args the_full_commandline_here, and just "bt full" when it reaches the crash 02:48
also, i'm not sure if valgrind runs on netbsd, but it probably does. so try to run "valgrind the_full_commandline_here" and give us the output
that'd be good
zilti Ok, will do that! Thanks 02:51
02:53 labster left
timotimo goes to bed 03:02
03:06 nchambers joined
samcv how can I make attributes inherited by other classes? 03:07
the methods get inherited but I would like to access the data without having to copy a large data structure back and forth each time. is this possible?
timotimo private attributes are private to the one class only, derived classes can use the method accessors
using the accessors doesn't force you to copy stuff around, though 03:08
seatek you have to use $.varname
samcv ok, can i write to the variables though?
seatek you have to use $.varname is rw
samcv yeah ok
hmm. but what if i only want inheriting classes to use it and not the main program?
timotimo if you don't want the attribute to be publically writable, you can have a private accessor method and call that from subclasses
i think
samcv how do I do that timotimo ? 03:09
seatek yeah i haven't found a nice way to do that yet
samcv ah
i'll just use `is rw` for now 03:10
timotimo m: class Parent { has $.attr; method !dothething is rw { $!attr } }; class Child is Parent { method do-evil-magic($a) { self!Parent::dothething = $a } }; my $c = Child.new(); say $c.attr; $c.do-evil-magic(42); say $c.attr
camelia rakudo-moar 2efeda: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Cannot call private method 'dothething' on package Parent because it does not trust Child␤at <tmp>:1␤------> 3evil-magic($a) { self!Parent::dothething7⏏5 = $a } }; my $c = Child.new(); say $c.a␤ …»
seatek purists seem to believe you should never be able to access attributes
timotimo okay, not like that.
seatek except through methods
timotimo private attributes are for separating your logic from the API your class provides
seatek and i haven't found a good way to "override" the "automatic" accessors created for attributes
samcv well seatek i am inheriting methods. and the base class holds the data, but 03:11
but most methods aside from loading/unloading and such are in the dependent classes
timotimo bedtime for real
samcv night
seatek hehe
nite nite
samcv hmm is it threadsafe though to do .data{'something'} = 'something else'; in a dependent class? 03:12
seatek yeah i've just been doing those as $.vars
samcv i know it isn't from the main program
seatek that i do not know
03:13 cibs left
samcv hmm i guess i could abstract more... but 03:13
seatek i'm pretty sure it would be though
samcv wanted my base class to not be specific to what the data was
and only load/unload
seatek you could do that with roles if i'm remembering right
samcv ugh no wait that makes it way too complex 03:14
since the datatypes are all different. at least the structures
03:14 cibs joined
seatek m: role R { has $!var = 5; }; class C does R { method test() { say $!var }; }; my $c - C.new; $c.test; 03:16
camelia rakudo-moar 2efeda: OUTPUT«WARNINGS for <tmp>:␤Useless use of "-" in expression "my $c - C.new" in sink context (line 1)␤Use of uninitialized value of type Any in numeric context␤ in block <unit> at <tmp> line 1␤Cannot resolve caller Numeric(C: ); none of these signatures m…»
seatek i hate this one-line thing 03:17
viki m: role R { has $!var = 5; }; class C does R { method test() { say $!var }; }; my $c = C.new; $c.test;
camelia rakudo-moar 2efeda: OUTPUT«5␤»
seatek yes that's it :)
thanks viki 03:18
BenGoldberg It doesn't have to be one-line. Stuff your code in a gist, and then type m: url
seatek it's like stuff you "inherit" from roles are "copied" into your parent class
viki seatek: the roles get inlined into the class, yeah
seatek so you can get away with makign your base class have private attributes if you use roles
viki: yeah :) i really like that feature. i've been loving the way you can use roles 03:19
samcv: that make sense? if you just throw your base class "holder" vars into role Base::Blah instead of class Base::Blah then you've got what you want 03:21
samcv nice 03:22
03:22 araujo left
seatek i think of roles like templates or macros that get copied in 03:22
03:22 araujo joined, araujo left 03:23 araujo joined
seatek BenGoldberg: ooo! I didn't know we do that :) 03:25
samcv ah yeah seatek that's what i wanted, roles 03:26
want it to inherit like everything. though maybe not submethods tho idk at this point because not using those
since all the objects are using 100% of the functions of the base class
03:27 noganex joined
seatek you can slice those up too if you want to clear them out and organize them a bit, and just bring in multiple roles 03:27
samcv so inheriting objects are actually two objects? and roles are one object that takes the parent and adds to it?
well idk 'two objects' but it doesn't inherit private attributes and state right? 03:28
but multiple layers of it that are seperate
seatek classes pay attention to that attribute scoping
roles just copy the code ;)
samcv cool
03:30 noganex_ left
seatek oh you migth want to watch out for collisions though -- when you use roles if you use the same names in what roles them in... it's fatal 03:30
classes override the same names, roles explode horribly
i have found that particularly vexxing on more than one occasion 03:31
but i'm a slob
samcv m: my role test { has $!thing = 10; }; my class test2 does test { method get { $!thing++; return $!thing; } }; my test2 $classy; $classy.get.say;
camelia rakudo-moar 4663d4: OUTPUT«Cannot look up attributes in a test2 type object␤ in method get at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤» 03:32
samcv hmm this doesn't work
or. uh did i not call .new
m: my role test { has $!thing = 10; }; my class test2 does test { method get { $!thing++; return $!thing; } }; my $class = test2.new; $classy.get.say;
camelia rakudo-moar 4663d4: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable '$classy' is not declared. Did you mean '$class'?␤at <tmp>:1␤------> 3rn $!thing; } }; my $class = test2.new; 7⏏5$classy.get.say;␤»
samcv m: my role test { has $!thing = 10; }; my class test2 does test { method get { $!thing++; return $!thing; } }; my $classy = test2.new; $classy.get.say;
camelia rakudo-moar 4663d4: OUTPUT«11␤»
samcv hmm ok that worked
viki m: role test { has $!thing = 10; }; class :: does test { method get { $!thing++; return $!thing; } }.new.get.say; 03:33
camelia rakudo-moar 4663d4: OUTPUT«11␤»
viki m: role test { has $!thing = 10; }; class :: does test { method get { ++$!thing } }.new.get.say;
camelia rakudo-moar 4663d4: OUTPUT«11␤»
ugexe m: role Foo { method test { say $?CLASS } }; role Bar { method test { say $?CLASS } }; class Impl does Foo does Bar { }; Impl.new.test # is this what you were trying to do?
camelia rakudo-moar 4663d4: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Method 'test' must be resolved by class Impl because it exists in multiple roles (Bar, Foo)␤at <tmp>:1␤» 03:34
seatek no we like to use code that actually works ;)
samcv ugh for some reason the same thing that worked in a simpler example isn't working in my code 03:36
seatek that's just the universe punishing you for something
samcv saying it's not declared in the dependant class 03:37
that `does` that role
yes seatek that is it
though one difference is i don't initialize the attribute like i did with camelia
seatek you defined the role yeah? and use'd it or have it in thet same file yeah? 03:38
samcv yes 03:39
seatek i think it wants to come before the class too if i'm remembering right (like compile order is important)
or i may be hallucinating
samcv gist.github.com/2eea11092a4e2ce092...9da9327f9f here is the relevant code 03:40
well this file should be able to standalone probably
so yeah. it gives that error 03:41
Attribute $!hash not declared in class history-class
seatek what if you take away the my's 03:42
samcv nope seatek. tried that already ;) 03:43
ugexe s/$!hash/%!hash/
samcv omg 03:44
seatek ha!! :)
samcv !!!
seatek ugexe: good eye! :)
samcv magic!
dudz what's a word that means priority but the top of the list is the lowest priority? 03:48
ugexe a stack? 03:49
last in, first out
dudz thats how stacks work? 03:50
so i could call my list a stack list, and that would mean what i typed above?
TEttinger not sure what priority means here
ugexe if you can replace priority with order
dudz yeah its really just in which order i want to run something 03:51
i'm trying to find a nice and tidy comment title for something
ugexe LIFO and FIFO are common terms
TEttinger LIFO and FIFO are two common terms when you are talking about the order things go in and are taken out. LIFO=last in first out, FIFO=first in first out
stacks are LIFO, queues are FIFO 03:52
geekosaur generally priority in this context is for when things are added in indeterminate order but should be taken out in a particular order. (priority queues, priority search queues) 03:53
dudz yeah, i have a list of values, and they can come in any order, but when i find them i want to compare and sort based on the least important so i can treat them with cruelty 03:56
geekosaur I don't think there's a pqueue in the ecosystem yet
oh, hah. github.com/hoelzro/p6-priorityqueue 03:57
04:03 khw left 04:05 labster joined
seatek did i hear someone say that stuff use'd is going to be lexically scoped soon, so that something use'd in one place will need to be use'd again in another if it's needed there too? 04:14
and if i did hear that, was it true?
geekosaur if your'e doing a use inside an inner scope instead of at file scope, I'd expect the result to be restricted to that scope, yes 04:15
ugexe m: { use Test; ok 1; }; { ok 1 }
camelia rakudo-moar 0a874f: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared routine:␤ ok used at line 1␤␤»
geekosaur perl 5, and in fact most other languages that allow imports at scopes inner to a file/compunit, work the same way 04:16
seatek gosh i must already be compensating for it
in perl5 if i used a file in one file -- say some library... and another library needed that file, i'd never get an error, because it was like, that file got interpreted once, and was useable everywhere then... within the same running program. i swear it was like that 04:18
MasterDuke nine is/was working on lexical module loading, don't remember if that got committed already 04:19
geekosaur perl 5 does use in two steps. if you always use qualified names from it, only the load step is needed. for unqualified names, there also needs to be an import step which is scoped
note that use foo; my $x = Foo->new; (perl 5) it's actually a fully qualified name, so no import is needed and it's available everywhere. 04:20
seatek MasterDuke: that will be The Way Things Are when it gets committed then?
MasterDuke believe so
seatek k 04:21
geekosaur (but in perl 6 there are more scopes, so I'm not sure the equivalent would do that)
seatek i'll just keep my lists of use'es very very long 04:22
geekosaur mm, actually I think it will. so if you are looking for classes from the "use"d module, it will work; if you are looking for subs from it (as opposed to methods), you need to "use" in each scope where you need them
seatek yieah
04:24 cibs left, vendethiel left
seatek why would anyone want to do a "unit class ...." rather than a "class ..." 04:24
just to make conversions easier?
MasterDuke the commit messages here have some more info github.com/rakudo/rakudo/tree/lexi...odule_load
04:25 cibs joined
MasterDuke so everything isn't indented one extra level perhaps 04:25
viki seatek: saves an indent level, if that class is all you got in the file: github.com/zoffixznet/perl6-App-No...ste.pm6#L1
seatek hehe - ok. Cuz wherever i see it being used, it looks suspicously like it used to be a "package" with lots of subs 04:28
geekosaur wouldn't surprise me if that were the most common current usage 04:29
04:32 kyclark_ joined 04:38 giraffe left, ugjka joined 04:45 giraffe joined 04:46 adu joined 05:08 Cabanossi left 05:10 Cabanossi joined 05:11 TimToady left, TimToady joined 05:20 perlawhirl left, perlawhirl joined 05:29 aborazmeh left 05:35 eisen74 joined 05:48 adu left 05:57 eisen74 left 06:19 xtt2 joined 06:23 domidumont joined 06:27 domidumont left 06:28 domidumont joined
ZzZombo Do hashes still return things in random order? 06:30
TEttinger why wouldn't they? 06:36
06:37 bjz_ left
TEttinger a hash function returns a seemingly-random value that is determined by the thing being hashed 06:37
ZzZombo because that is stupid not to have a way to disable it on selected hashes. 06:38
TEttinger I don't know enough perl6 to be able to answer about the current implementation. most programming languages have the default hash, dictionary, map, whatever they call it have unpredictable ordering; some use actually random ordering because it can be faster and the order of an unordered data structure shouldn't be relied upon. many languages have extra data structures that combine a list and hash 06:41
this can be called an ordered hash/map/dictionary, insertion-ordered whatever, Java for some reason calls it LinkedHashMap
they aren't the default for memory usage reasons; the list has memory costs in addition to the normal hash's memory usage. they are better in other ways. 06:43
06:43 pierre__ joined, pierre__ left, pierre__ joined
TEttinger (also adding or removing needs changing both the hash and list) 06:43
viki ZzZombo: there is a way to "disable" it.... .sort 06:49
ZzZombo But usually when you have such sort of a problem to solve, it meanss you wanna the order intact.
06:56 kyclark_ left
TEttinger viki, yeah, insertion-ordering and sorting are a weirdly partly-overlapping thing 07:03
07:06 xiaomiao left 07:09 xiaomiao joined 07:22 perlawhirl left 07:24 CIAvash joined
dalek c: 848a53d | gfldex++ | doc/Type/Method.pod6:
tell that the invocant defaults to self and how to call a free floating
 07:26
synopsebot6 Link: doc.perl6.org/type/Method
ZzZombo There is already an example "greeting".$m("hello"); 07:28
right above
is it right? 07:29
07:30 pierre__ left 07:33 solarbunny left
dalek c: d2e6925 | gfldex++ | doc/Type/Method.pod6:
link to Routine
 07:34
synopsebot6 Link: doc.perl6.org/type/Method
c: 554c148 | gfldex++ | doc/Type/Method.pod6:
link to Signature
synopsebot6 Link: doc.perl6.org/type/Method
07:34 solarbunny joined
gfldex ZzZombo: yes, but with an anonymous method and as such differently invoced. 07:35
07:40 travis-ci joined
travis-ci Doc build errored. Wenzel P. P. Peppmeyer 'link to Signature' 07:41
travis-ci.org/perl6/doc/builds/179005566 github.com/perl6/doc/compare/848a5...4c148c089c
07:41 travis-ci left
ZzZombo wut 07:47
dalek c: acc5d53 | gfldex++ | doc/Type/Signature.pod6:
link to Method and Sub
 07:51
synopsebot6 Link: doc.perl6.org/type/Signature
gfldex the SSL module comes with a heisentest
07:54 BenGoldberg left 08:00 geekosaur left 08:01 geekosaur joined 08:02 RabidGravy joined 08:04 skids left 08:08 bjz joined 08:13 travis-ci joined
travis-ci Doc build passed. Wenzel P. P. Peppmeyer 'link to Method and Sub' 08:13
travis-ci.org/perl6/doc/builds/179006641 github.com/perl6/doc/compare/554c1...c5d53626cc
08:13 travis-ci left 08:14 firstdayonthejob joined 08:16 cibs left 08:17 cibs joined
ZzZombo what was the reverse of 'so' again? 08:20
RabidGravy you mean 'not' ? 08:23
haven't had much coffee yet, struggling with what reverse means in this case :) 08:24
08:28 pmurias joined
pmurias ZzZombo: how do you want the hashes to order things? 08:29
08:39 bjz_ joined 08:40 bjz left
ZzZombo in the same order you put them in 08:52
08:53 domidumont left
RabidGravy you would have to make your own Map class to do that 08:53
08:55 Tonik joined
seatek yeah on the very rare occation i've had to worry about that, i use arrays of hashes 08:59
09:02 pierre__ joined 09:09 TEttinger left 09:15 perlawhirl joined
pmurias ZzZombo: you will have to use a module for that 09:26
09:31 pierre__ left 09:33 pierre__ joined 09:35 rindolf joined
pmurias ZzZombo: the random ordering is not something that has been "stupidly" enabled without a way to turn it of, but the ordering you want would have to be implemented at a performance cost 09:36
09:39 Actualeyes left 09:40 Actualeyes joined 09:41 espadrine joined
RabidGravy I'm still struggling to find a way to "discover" the inner packages within something like "class A { class B { } }" 09:41
so e.g. in the compose of A I want to find 'B' (and any random inner classes that might be there) 09:42
I'm sure there is a way to do this, I just can't determine the right pseudo package to monkey with 09:43
09:49 vendethiel joined 09:51 FROGGS joined
FROGGS o/ 09:52
RabidGravy boom
09:53 Actualeyes left
pmurias ZzZombo: hopefully someone will write a module to implement your insertion ordered hashes but I don't think it will be added to the core 09:56
10:00 grondilu joined 10:02 xtt2 left, grondilu left
RabidGravy doing it naively would be trivial, by just sub-classing Hash with an over-ride on AT-KEY and ASSIGN-KEY to keep the insertion order 10:02
and then whichever of the .keys, .values, .pairs are required 10:03
10:04 labster left 10:18 pierre__ left 10:27 wamba joined 10:36 pierre__ joined
perlawhirl ZzZombo: github.com/zostay/perl6-ArrayHash 10:40
from the description: "You can think of this as a Hash that always iterates in insertion order or you can think of this as an Array of Pairs with fast lookups by key." 10:41
10:45 pierre__ left 10:48 bjz_ left 10:51 bjz joined 10:53 canopus left 11:00 CIAvash left
RabidGravy I think my earlier question can be arranged as, is there a way of doing: 11:01
m: class A { class B {}; say A::.values }
camelia rakudo-moar 7c5ea3: OUTPUT«((B))␤»
11:01 canopus joined
RabidGravy but where you don't know A in advance and have been handed it in a variable 11:02
or does it only work with the type name like that 11:03
obviously I can do it in an EVAL but there's nasty
FROGGS m: class A { class B {}; say A::.values }; my $a = A.new; say $a.WHAT::.values 11:04
camelia rakudo-moar 7c5ea3: OUTPUT«((B))␤()␤»
FROGGS m: class A { class B {}; say A::.values }; my $a = A.new; say ::($a.WHAT)::.values
camelia rakudo-moar 7c5ea3: OUTPUT«((B))␤Use of uninitialized value of type A in string context.␤Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.␤ in block <unit> at <tmp> line 1␤No such symbol ''␤ in block <unit> at <tmp> line 1␤␤Ac…»
RabidGravy ah
FROGGS m: class A { class B {}; say A::.values }; my $a = A.new; say ::($a.^name)::.values
camelia rakudo-moar 7c5ea3: OUTPUT«((B))␤()␤»
FROGGS hmmm
RabidGravy duh me
FROGGS m: class A { class B {}; }; my $a = A.new; say $a.WHAT:: 11:05
camelia rakudo-moar 7c5ea3: OUTPUT«(A)␤»
FROGGS no idea why it doesnt print B 11:06
RabidGravy yeah that's what I've been seeing 11:07
FROGGS m: class A { class B {}; }; my $a = A.new; say A::.values; say $a.WHAT.::.values
camelia rakudo-moar 7c5ea3: OUTPUT«===SORRY!===␤MVMArray: Can't pop from an empty array␤»
FROGGS sad that there is no B in that sentence either :P
RabidGravy I think I'll go with the EVAL in the first instance 11:08
in the whole scope of the rabbit hole I'm going down that's fairly sane 11:09
11:09 CIAvash joined 11:10 CIAvash left 11:14 domidumont joined 11:27 xfix left 11:31 xfix joined 11:39 domidumont left 11:41 Actualeyes joined 11:46 pierre_ joined 11:49 Lucas_One joined 11:51 CIAvash joined, pierre_ left 11:54 perlawhirl left
ZzZombo why is there Str.chop() but no something to remove characters from the start? 12:14
nine MasterDuke: plan is that it will behave like this: { { use Foo; Foo.new; #`(Foo found and works) }; Foo.new #`(There's no Foo in this scope, so we throw an error here) }
12:15 CIAvash left
nine seatek: ^^^ 12:16
ZzZombo m: put 'String'.substr(2,*-1) 12:18
camelia rakudo-moar 7c5ea3: OUTPUT«rin␤»
ZzZombo m: put 'String'.substr(2) 12:19
camelia rakudo-moar 7c5ea3: OUTPUT«ring␤»
ZzZombo how does it evalute *-1 into an actual index? 12:20
jnthn *-1 is just a function -> $x { $x - 1 }, and it calls the function with .chars of the string 12:21
seatek nine: thank you for the confirmation. i've been coding for that, so i'm happy i haven't been typing extra for no good reason! :)
ZzZombo is it built-in into the function itself or black compiler magic?
jnthn Part of the method definition
lizmat m: 'string'.substr(2, { .say; 1 } )
camelia rakudo-moar 7c5ea3: OUTPUT«4␤»
jnthn You can pass any code object there, as lizmat showed 12:22
lizmat m: 'string'.substr(0, { .say; 1 } )
camelia rakudo-moar 7c5ea3: OUTPUT«6␤»
ZzZombo hm, the definitions look rather ordinarily. Nothing says they accept Callables.
jnthn m: my $s = * - 1; say $s(4); say $s(43)
camelia rakudo-moar 7c5ea3: OUTPUT«3␤42␤»
ZzZombo unless they are simplified on purpose in the docs, or something.
12:23 firstdayonthejob left
jnthn m: Str.^lookup('substr').candidates>>.signatures.say 12:23
camelia rakudo-moar 7c5ea3: OUTPUT«No such method 'signatures' for invocant of type 'Method'␤ in block <unit> at <tmp> line 1␤␤»
jnthn m: Str.^lookup('substr').candidates>>.signature.say
camelia rakudo-moar 7c5ea3: OUTPUT«((Cool $: $from, $length?, *%_))␤»
jnthn Hm, maybe the magic happens inside of there
oh 12:24
12:24 bjz_ joined
jnthn m: say &substr.candidates>>.signature 12:24
camelia rakudo-moar 7c5ea3: OUTPUT«((Str:D \what, Int:D \start) (Str:D \what, Callable:D \start) (Str:D \what, Int:D \start, Int:D \want) ($what, \start, $want?))␤»
seatek nine: it's a big change. but the sooner the better i say. less carnage to clean up.
12:25 bjz left
ZzZombo The return value of your trait_mod is ignored, isn't it? 12:28
jnthn Yup
lizmat jnthn: fwiw, substr is one of the few cases where the logic is in the sub, not in the method 12:29
this is contrary to most other sub/method duals
jnthn Indeed. Should probably be fixed at some point.
lizmat do you know if there is a specific reason to do it like that?
ok, I'll put it on my list then
jnthn I seem to recall arguing against it being that way before :)
lizmat well, performance would come to mind 12:30
but otoh, that was probably before we had good MMD caching support
jnthn *nod*
But yeah, it'll get in the way of us doing Cat some day
If the sub doesn't delegate to the method 12:31
Probably same story for when we improve Uni and friends
lizmat ok
jnthn So worth fixing
lizmat oki, need to finish up native shaped array opts first 12:32
jnthn :)
lizmat 1+dim .STORE specifically :-)
jnthn Nice 12:35
Lunch, then I'll see if I feel like a SEGV hunt this afternoon...
pmurias jnthn: re mutators blog post, are those even used commonly enough that syntax sugar instead of Proxy is required?
12:36 pierre_ joined
jnthn pmurias: Syntax sugar in that area has been suggested/requested more than once in the past, and I wrote the post to have a good place to point people at as to why we don't provide it and don't plan to. 12:39
RabidGravy is there any way of allowing a parameterized class definition like "class Foo[whatever] { }" assuming I'm already monkeying with the MOP?
seatek roles do that.. different from roles? 12:40
RabidGravy yeah different to roles
actually as part of the definition
seatek are you never satisfied? ;) 12:41
MasterDuke jnthn: your post reminded me of Steve Yegge's post about the kingdom of nouns vs verbs, but i haven't read it recently enough to know if they really are similar. are you familiar with it? and if so, are they related?
RabidGravy nope :) I could achieve the intent with a trait but it would be overly verbose in the circumstance 12:43
jnthn RabidGravy: You can implement method ^parameterize like Array does 12:44
RabidGravy: In short: no, classes aren't parametric, but at the meta level what is parametric or not simply boils down to "does the meta-object respond to .parameterize", so if you're willing to interpret that and make it mean what you'd like it to mean, you're in with a chance :) 12:46
seatek it's unnatural 12:47
RabidGravy right so adding a role to the meta object with parameterize should do the job 12:48
jnthn MasterDuke: It's been a while since I read that post also. :)
RabidGravy let's see
gfldex RabidGravy: please provide code if possible, this may be ENODOC 12:49
jnthn MasterDuke: I think it was getting at a slightly different point than my main point, though my notes on us not stigmatizing procedural programming likely relate.
MasterDuke: Though I think lack of first class functions was also a moan of the kingdom of nouns post too 12:50
Curiously, Java has actually got those at long last :)
MasterDuke yeah, i should read it again and see how much things are better 12:52
FROGGS jnthn++ # blog 12:53
jnthn I guess one of the things I've found a tad frustrating with Java (and C# to a degree too) is that in a well-meaning quest to get people to write good OO code, they made OO code the only code you could write. But OO isn't easy, and badly done OO tends to make things worse rather than better. 12:55
ZzZombo so I'm generating a method dynamically, is it possible to change the invocant part of my method handler(Any:D: $self,|c) from Any:D to a $class:D that I know only from a string? 12:56
jnthn It's like autotune. When every single pop song is using it, it all becomes a bit naff. :P
lunch; bbl :) 12:58
13:01 pierre_ left 13:10 cdg joined, wamba left
nine First draft of lexical module loading docs: gist.github.com/niner/70f7b46eefb7...6bea11efeb 13:14
Comments/suggestions/patches very welcome :)
13:14 AlexDaniel joined 13:21 perlawhirl joined
ZzZombo method eager(Seq:D:) returns Bool:D 13:24
Returns an eagerly evaluated List based on the invocant sequence
wut
perlawhirl m: say 1..10 13:26
camelia rakudo-moar 7c5ea3: OUTPUT«1..10␤»
perlawhirl m: say eager 1..10
camelia rakudo-moar 7c5ea3: OUTPUT«(1 2 3 4 5 6 7 8 9 10)␤»
perlawhirl it eagerly evaluates a lazy list
oh i just saw your first message... are you questioning the Bool thingy 13:30
ZzZombo yes
perlawhirl umm yeh... that looks wrong
timotimo that's why it's in boldface :) 13:31
perlawhirl timotimo: thanks 13:32
ok... text appearance setting in my term emulator had boldface disabled 13:34
i see it now
timotimo oh, interesting
perlawhirl ecureCRT 13:35
i dunno what happened to the S... SecureCRT 13:36
13:40 Tonik left
AlexDaniel ah no 13:41
let's see
ZzZombo if I have sub x($blah,|c), how do I refer to c in the body? 13:43
just |c?
RabidGravy well just c
ZzZombo hm
RabidGravy it's a Capture object
so e.g.
ZzZombo "Inside a Signature, a Capture may be created by prefixing a sigilless parameter with a vertical bar |. This packs the remainder of the argument list into that parameter." <-- why does it have to be sigilless? 13:44
RabidGravy m: sub foo(|c) { say c.list }; foo("ab");
camelia rakudo-moar 7c5ea3: OUTPUT«(ab)␤»
13:44 pierre_ joined
ZzZombo RabidGravy: thanks! 13:44
timotimo what happens if you call it |$foo? 13:45
dalek c: 97836f7 | (Aleks-Daniel Jakimenko-Aleksejev)++ | doc/Type/Seq.pod6:
Seq.eager does not return a Bool

It is a List for sure
synopsebot6 Link: doc.perl6.org/type/Seq
timotimo m: sub x(|$foo) { say $foo.perl }; x(1, 2, 3, :5lol)
camelia rakudo-moar 7c5ea3: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Obsolete use of | or \ with sigil on param $foo␤at <tmp>:1␤------> 3sub x(|$foo7⏏5) { say $foo.perl }; x(1, 2, 3, :5lol)␤ expecting any of:␤ shape declaration␤»
timotimo ah, that's why, i suppose
RabidGravy fairly good reason :) 13:46
AlexDaniel ZzZombo: ok, I think the Bool thingy is wrong indeed, so I fixed it
ZzZombo eh, why is it obsolete?
AlexDaniel ZzZombo: thanks for reporting
ZzZombo doing my part
AlexDaniel well, you probably meant @foo?
without a question mark though :)
commit: pre-glr sub x(|$foo) { say $foo.perl }; x(1, 2, 3, :5lol) 13:50
committable6 AlexDaniel, gist.github.com/cfeb632a353ec6120a...14c8ea6588
AlexDaniel commit: all sub x(|$foo) { say $foo.perl }; x(1, 2, 3, :5lol) 13:51
committable6 AlexDaniel, gist.github.com/fee63950b0aaf947b2...1d0aff82cf
AlexDaniel does anybody have any idea?
perlawhirl about what... the 'Obselete' thing? 13:52
it's coming from a $/.CURSOR.panic() inside the Grammar for token parameter 13:53
AlexDaniel … do we still need it? 13:54
timotimo depends. when was it not obsolete? what did it mean? 13:55
does anybody still remember? :)
AlexDaniel commit: all sub x(|$foo) { say $foo.perl }; x(1, 2, 3) 13:56
committable6 AlexDaniel, gist.github.com/45770fc35b46dc7398...90ce153cc6
timotimo i think it's been there for years
perlawhirl *shrug* generally captures are used when you don't want to unpack anything... typically to punt it elsewhere... maybe assigning a capture to a scalar is bad because you giving it a container/shape
AlexDaniel it is a pre-moar thing
14:00 bjz_ left 14:03 pierre_ left 14:04 andrzejku joined
ZzZombo Quoted method name requires parenthesized arguments. 14:05
return $self."%accessors<setter>"(c.kv.eager); <--here
how to fix?
AlexDaniel hmmm what is $self? 14:06
ZzZombo object
AlexDaniel I'm just wondering why not "self" 14:07
viki m: my $self = 'FF'; my %accessors = setter => 'parse-base'; say $self."%accessors<setter>"(16)
camelia rakudo-moar 7c5ea3: OUTPUT«255␤»
ZzZombo because a Wizard made it so
AlexDaniel m: $self = ‘FF’; my %accessors = setter => ‘parse-base’; say $self."%accessors<setter>": 16
camelia rakudo-moar 7c5ea3: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable '$self' is not declared␤at <tmp>:1␤------> 3<BOL>7⏏5$self = ‘FF’; my %accessors = setter => ␤»
AlexDaniel m: my $self = ‘FF’; my %accessors = setter => ‘parse-base’; say $self."%accessors<setter>": 16 14:08
camelia rakudo-moar 7c5ea3: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Quoted method name requires parenthesized arguments. If you meant to concatenate two strings, use '~'.␤at <tmp>:1␤------> 3se-base’; say $self."%accessors<setter>"7⏏5: 16␤»
viki m: my $self = 'FF'; my %accessors = setter => 'parse-base'; my \c = class { method kv { 16 } }; say $self."%accessors<setter>"(c.kv)
camelia rakudo-moar 7c5ea3: OUTPUT«255␤»
viki ZzZombo: can't reproduce.
AlexDaniel viki: well it's right there ↑ if you omit parens
so it is just unparseable this way, use parens
viki AlexDaniel: yes and "fixing it" would be using parantheses, which ZzZombo's example does. 14:09
AlexDaniel u: shrug
unicodable6 AlexDaniel, U+1F937 SHRUG [So] (🤷)
ZzZombo sorry, turned out there was a stray line indentical to that, guess I accidentally made a copy I didn't notice before. 14:10
AlexDaniel \o/
u: raising 14:12
unicodable6 AlexDaniel, U+1F64B HAPPY PERSON RAISING ONE HAND [So] (🙋)
AlexDaniel, U+1F64C PERSON RAISING BOTH HANDS IN CELEBRATION [So] (🙌)
AlexDaniel ok after a recent update all of these are rendered with a female character instead of a neutral smiley face or genderless human outline… great 14:16
I remember there was a person here trying to figure out which font is used to render a specific character
viki On my box they're rendered as a bald dude with a very sickly skin colour: i.imgur.com/VmLb0tS.png 14:20
AlexDaniel 🤷🙋 14:21
viki: I guess it is better than this: files.progarm.org/2016-11-26-16213..._scrot.png
ZzZombo Any help would be appreciated: 14:23
m: gist.github.com/ZzZombo/5caa51f9b2...556920a0d8
camelia rakudo-moar bdc7a9: OUTPUT«("\$!name" => "name", Attribute.new, ("\$!name",))␤5===SORRY!5=== Error while compiling <tmp>␤Too many positionals passed; expected 4 arguments but got 5␤at <tmp>:58␤»
AlexDaniel viki: and obviously better than this: files.progarm.org/2016-11-26-16231..._scrot.png
14:23 pierre_ joined
ZzZombo that's what I made for Delphi-like attribute accessors. 14:24
Except that error is in my way!
viki ZzZombo: what are :$handled [:$setter=True,:$getter=True] ? is that destructuring? 14:25
ZzZombo Citing Sub.pod6: "Use destructuring to call traits with complex arguments.", so I guess so?
timotimo hack.p6c.org/~timo/mysterious_but_h...ess_binary - feel free to download and run this binary and guess why i'm pretty proud of it :) 14:26
viki, nine, others
ZzZombo timotimo: rusty click of the day?
viki ZzZombo: does it need a comma after $handled?
timotimo ZzZombo: i don't program in rust, though i'm intrigued about it 14:27
ZzZombo timotimo: it was a pun on the word "risky" :)
AlexDaniel 🙌
timotimo thought so 14:28
AlexDaniel nope, didn't work
ZzZombo viki: I don't understand, do you ask if there should be a comma?
viki ZzZombo: yeah
ZzZombo I don't see one in the examples. 14:29
viki doesn't run random executables to guess why someone is proud
AlexDaniel timotimo: soo… that's moar
timotimo AlexDaniel: and? :) 14:30
AlexDaniel timotimo: it's in a single binary file
timotimo well, yeah, moar is always only a single binary file
except it usually has libmoar.so to go with it, which this one don't
but that's not exciting at all
viki you made it relocatable?
AlexDaniel is it small or something? 14:31
viki I'm just being asked if I want to open it with notepad.exe
timotimo viki: you're on windows, then? :\ 14:32
this isn't what i'd call "relocatable"
AlexDaniel :D
viki timotimo: I'll answer with a visual aid: i.imgur.com/GLz8lFK.png 14:33
:)
timotimo and at a whooping 15 megabytes, that's definitely not small
AlexDaniel SO WHAT DID YOU DO? 14:34
timotimo fair enough. so you *can* run it :)
AlexDaniel :)
14:34 pierre_ left
viki I don't want to 14:34
timotimo trusting me or not is your choice, of course
viki I have bad experience with "harmless binaries"
timotimo yeah, okay, i'll spill the beans :(
you're no fun
viki And even if I trust you, what if you messed up :)
timotimo you remember how people kept saying "it can't be done! you're crazy!"
viki no 14:35
timotimo about putting a program into a single binary with the moar binary?
viki and your binary is a hello world packaged into a single binary?
timotimo no
it's nqp packaged into a single binary
viki :o
timotimo unmodified nqp
AlexDaniel timotimo: how do I run it? 14:36
pmurias timotimo: why are you trying to convince people to run binaries downloaded over unencrypted http? ;)
timotimo chmod +x it, ./ it
AlexDaniel ERROR: please be so kind and provide an arbitrary single commandline argument.
timotimo yes, do as it says please
AlexDaniel --help gives moarvm help
viki timotimo++ cool!
timotimo try "lol" or "wat" or "arbitrary"
055af1619093f1812ee62df6c3e6dc17dc1c53b0 mysterious_but_harmless_binary
sha1sum of the binary for pmurias
AlexDaniel Illegal instruction ? 14:37
timotimo uh oh
why is that?
AlexDaniel you remember how people kept saying "it can't be done! you're crazy!" 14:38
timotimo turs out they just meant "it's illegal to do that"?
viki ZzZombo: you can add --ll-exception switch and it'll give you more stuff in the error
AlexDaniel no, but at least this particular binary does not seem to work :)
timotimo it can't give more info with Illegal instruction
pmurias timotimo: the binary should fire up a NQP REPL? 14:39
timotimo that's a "the CPU is upset and kills your process"
pmurias: if you don't give it an nqp script, yes
viki ZzZombo: and at a guess, you don't need $class in add_method
ZzZombo: (from docs) method add_method(Metamodel::MethodContainer: $obj, $name, $code) <-- it has a colon after MetamodleMethodContainer, that indicates it's the invocant 14:40
AlexDaniel timotimo: soo… do I need to have it in a specific place or do I need to have some kind of a file around? 14:41
ZzZombo what is $obj then?
also that switch didn't really tell me anything.
timotimo AlexDaniel: can you gdb it? does the sha1sum match? what exactly are you doing to cause the error? 14:42
AlexDaniel timotimo: sha1sum matches
timotimo you have a 64bit system?
AlexDaniel ./mysterious_but_harmless_binary lol
Illegal instruction
yes
ZzZombo oh wait
viki s: 42.HOW, 'add_method' 14:43
SourceBaby viki, Something's wrong: ␤ERR: X::Multi::NoMatch exception produced no message␤ in sub sourcery at /home/zoffix/services/lib/CoreHackers-Sourcery/lib/CoreHackers/Sourcery.pm6 (CoreHackers::Sourcery) line 33␤ in block <unit> at -e line 6␤␤
ZzZombo if I look at the file names, it says "(C:\rakudo\share\nqp\lib/Perl6/Metamodel.moarvm:add_method)", so probably you are right. But still, the state of add_method is unclear.
14:43 chris2 left
timotimo AlexDaniel: so ... gdb it, please? 14:44
viki ZzZombo: looks like a typo in the docs. Should be add_method(Metamodel::MethodContainer $obj: $name, $code)
14:44 cdg left
viki ZzZombo: oh, no, I see why it's written that way 14:44
ZzZombo I think even better would be to use $class instead of $obj, the latter implies an instance rather than a type. 14:45
14:45 domidumont joined
AlexDaniel timotimo: anything specific? gist.github.com/AlexDaniel/91dd4da...7ad1e0ae99 14:45
timotimo "run lol" please
viki ZzZombo: you see, $foo.^add_method(<a b c>) is the same as $foo.HOW.add_method($foo, <a b c>)
ZzZombo viki: why?
RabidGravy it is an object
viki And in the docs the method is documented without the ^
ZzZombo so the ultimate conclusion on the use of it? 14:46
AlexDaniel oh, this is interesting
ZzZombo get rid of $class in param list?
AlexDaniel timotimo: gist.github.com/AlexDaniel/8b0a685...b2749f9394
viki ZzZombo: yes 14:47
timotimo AlexDaniel: but there's nothing special about that line? :\
oh!
OH!
haha, i'm silly
ZzZombo m: gist.github.com/ZzZombo/5caa51f9b2...556920a0d8 14:48
camelia rakudo-moar bdc7a9: OUTPUT«("\$!name" => "name", Attribute.new, ("\$!name",))␤5===SORRY!5=== Error while compiling <tmp>␤Missing block␤at <tmp>:101␤------> 3<BOL>7⏏5<EOL>␤»
timotimo i configured my moar with -march=native
ZzZombo err
viki ZzZombo: what's with c.kv.eager? 14:49
m: gist.github.com/zoffixznet/21fff13...195ed2e2fe
camelia rakudo-moar bdc7a9: OUTPUT«("\$!name" => "name", Attribute.new, ("\$!name",))␤New name: Bob, old name: unnamed.␤A.new␤»
timotimo AlexDaniel: so since you don't have the exact same cpu as me, it probably uses some optimization-related instruction your cpu doesn't have
AlexDaniel timotimo: ok, fix it? 14:50
timotimo: and I'll run it again
timotimo yes, i'm uploading the new version now
c37e662155612839aceb11a070c5a47a895dcf1e
that's the new sha1sum, it's up at the same address as before
viki zoffix@VirtualBox:/tmp/tmp.HFcxxlx7fR$ ./mysterious_but_harmless_binary "wat" 14:51
> nqp::say("Hello world!")
Hello world!
>
14:51 ugjka is now known as UgJkA
AlexDaniel nope 14:52
ok let's see
ZzZombo viki: thank you
AlexDaniel ah no
it works!
timotimo hooray
AlexDaniel timotimo: IT WORKS!! IT WORKS!!!!
ZzZombo but what does '|' do exactly here?
viki ZzZombo: it's a slip 14:53
viki is in the middle of devouring chicken wings....
ZzZombo bon appetite
timotimo AlexDaniel: i'm glad :) 14:54
14:56 CIAvash joined
ZzZombo viki: what do you think about the whole thing? Any possible improvements? 14:56
14:58 chris2 joined, setty1 left
ZzZombo m: sub x('Str:D' $s){put $s};x('meow') 15:01
camelia rakudo-moar bdc7a9: OUTPUT«Constraint type check failed for parameter '$s'␤ in sub x at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
ZzZombo m: sub x(Str:D $s){put $s};x('meow')
camelia rakudo-moar bdc7a9: OUTPUT«meow␤»
ZzZombo what does it attempt to do in the first case?
timotimo if there was no $s, it'd only match for 'Str:D'
m: sub x('Str:D' $s){put $s};x('Str:D') 15:02
camelia rakudo-moar bdc7a9: OUTPUT«Str:D␤»
timotimo it clearly works the same way with a $s after it
viki ZzZombo: my %accessors=(setter=>$setter,getter=>$getter); can be written as my %accessors= :$setter, $getter; and if c.Bool {...} doesn't need a .Bool. Looks fine other than that, but I know nothing of MOP or traits
timotimo you missed a : before $getter
viki yeah
ZzZombo what line? 15:03
viki 10
Oh, it was *I* who missed the :
but the %accessors is on line 10 15:04
AlexDaniel m: sub foo(25 $n) { dd $n }; foo 25
camelia rakudo-moar bdc7a9: OUTPUT«25␤»
ZzZombo I don't see anything like that... ah
AlexDaniel I have never seen that
titsuki hi #perl6
Does [email@hidden.address] currently work ? I can't send a bug report.
AlexDaniel titsuki: it was working yesterday, I think 15:05
viki titsuki: what happens when you do send something?
titsuki viki: it returns the following message: gist.github.com/titsuki/0f4e610ad5...6c9bfff965 15:06
AlexDaniel m: sub foo(True $n) { dd $n }; foo False 15:07
camelia rakudo-moar bdc7a9: OUTPUT«Bool::False␤»
AlexDaniel bug or not?
viki m: sub foo(True) { dd 'meow' }; foo False
camelia rakudo-moar bdc7a9: OUTPUT«"meow"␤»
viki well, that certainly is :D
AlexDaniel OOPS 15:08
bisect: sub foo(True) { dd 'meow' }; foo False
bisectable6 AlexDaniel, On both starting points (old=2015.12 new=bdc7a91) the exit code is 0 and the output is identical as well
AlexDaniel, Output on both points: "meow"
TimToady notabug
m: say False ~~ True
camelia rakudo-moar bdc7a9: OUTPUT«Potential difficulties:␤ Smartmatch against True always matches; if you mean to test the topic for truthiness, use :so or *.so or ?* instead␤ at <tmp>:1␤ ------> 3say False ~~ 7⏏5True␤True␤»
viki Sure, but it doesn't show that warning when used in sig.
ZzZombo m: sub foo(True $n) { dd $n }; foo 'KAABLAAAAM' 15:09
camelia rakudo-moar bdc7a9: OUTPUT«Type check failed in binding to $n; expected Bool but got Str ("KAABLAAAAM")␤ in sub foo at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
TimToady you're supposed to know the principle by the time you start writing multi-siggies :)
AlexDaniel well, not exactly
viki titsuki: lemme try to send one as a test... 15:10
titsuki viki: thx!
TimToady
.oO(humongous backlog is humongous)
 15:11
ZzZombo uh, I thought to call foo with such signature you would have to call it as foo <expression returning True> <Any>
why does it bind parameters straight to $n? 15:12
viki titsuki: yeah, I got the same error 15:17
titsuki: what was the bug anyway?
15:17 setty1 joined
titsuki viki: Here: gist.github.com/titsuki/11a483c0aa...8491698f51 15:19
viki :o 15:20
titsuki viki: I'll try to send it later. thx!
AlexDaniel viki: so I guess I'll rakudobug it because there is no warning?
viki that looks scary :)
AlexDaniel viki: or should we just document it as a trap 15:21
viki AlexDaniel: well, you can try, but RT is broken
AlexDaniel: well, IMO it should warn, just like it does with regular smartmatch, but you also have TimToady's comments above... and well, Rule 1? :)
AlexDaniel f∞×↓⋆ng RT 15:22
ZzZombo how can I get the default value for an attribute?
TimToady is fine with a warning there too, for the record
viki \o/
TimToady which is why I had a smiley above
viki AlexDaniel: titsuki: I emailed perlbug-admin at perl.org and the email seems to have gone through... so... let's wait 15:23
AlexDaniel ok, Rule 1.1: Rule 1 does not apply if there is a smiley
viki :D
ZzZombo [23:10:55] <ZzZombo> uh, I thought to call foo with such signature you would have to call it as foo <expression returning True> <Any> 15:28
[23:11:54] <ZzZombo> why does it bind parameters straight to $n?
anyone knows that?
viki ZzZombo: well, the True thing is wrong, since the thing uses smartmatch 15:29
ZzZombo: foo('meow' $n) {} ... on the other hand. That lets you specify an exact value $n must be for that sub to be called and, yeah, usually you wouldn't use $n (since you know what it should be), but I guess it works with it specified too 15:30
15:30 araujo left
ZzZombo I meant this: 15:31
m: sub foo(True $n) { dd $n }; foo 'KAABLAAAAM'
camelia rakudo-moar bdc7a9: OUTPUT«Type check failed in binding to $n; expected Bool but got Str ("KAABLAAAAM")␤ in sub foo at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
ZzZombo also, shouldn't there be a comma between the args?
so many questions D:
viki m: multi n(τ) { "tau!" }; multi n(π){ "pi!" }; multi n { "meh, just $^n" }; say n $_ for π, 2*π, 42
camelia rakudo-moar bdc7a9: OUTPUT«pi!␤tau!␤meh, just 42␤»
viki ZzZombo: think of it as sub foo(Bool $n) { dd $n }; foo 'KAABLAAAAM'; it'd blow up the same 15:32
and you don't need a comma
except instead of a Bool you provide an exact value
15:32 khw joined
viki (which also happens to be incorret in this case due to smartmatching) 15:32
ZzZombo m: sub foo('13666' $n) { dd $n }; foo 'KAABLAAAAM' 15:33
camelia rakudo-moar bdc7a9: OUTPUT«Constraint type check failed for parameter '$n'␤ in sub foo at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
ZzZombo m: sub foo('13666' $n) { dd $n }; foo '13666'
camelia rakudo-moar bdc7a9: OUTPUT«"13666"␤»
ZzZombo I see 15:34
AlexDaniel viki: I loved your comment, by the way
viki: about cryogenic chamber
viki heh
AlexDaniel viki: you've missed some numbers in the amount of years though, I think
viki shrugs 15:35
AlexDaniel digits I mean
viki it was 1AM :)
AlexDaniel yeah, you needed some sleep…
viki: by the way, fwiw, commit messages are normally ≤ 72, not 80 15:37
viki I'm a rebel ;)
AlexDaniel that's fine, but the lines wrap around weirdly even on github :)
viki At least until dalek stops cutting off commit messages that are X-lines long :) 15:38
seatek they even go in and pick out all the extra invisible spaces you might have left in your files
viki Seem to wrap fine on my github: i.imgur.com/pB16vKx.png 15:39
seatek i think they're... obsesive ;)
AlexDaniel viki: files.progarm.org/2016-11-26-17395..._scrot.png and files.progarm.org/2016-11-26-17401..._scrot.png 15:40
psch getting frozen inside a cryogenics chamber is probably not particularly conductive to getting unfrozen anyway 15:41
viki I see. 15:42
psch 'cause, y'know, that's the science of achieving extremely low temperatures and the effect of those temperatures :P
15:43 mscha joined
psch granted, only the cryonicists really care about the distinction 15:43
i'm not one, i'm too poor /o\
mscha m: say (2..36).map({108.base($^b)}); # This works, but ... 15:44
camelia rakudo-moar bdc7a9: OUTPUT«(1101100 11000 1230 413 300 213 154 130 108 99 90 84 7A 73 6C 66 60 5D 58 53 4K 4G 4C 48 44 40 3O 3L 3I 3F 3C 39 36 33 30)␤»
mscha m: say (2..36).map(108.base(*)); # ... why doesn't this?
camelia rakudo-moar bdc7a9: OUTPUT«Cannot resolve caller base(Int: Whatever); none of these signatures match:␤ (Int:D $: Int:D $base, *%_)␤ (Int:D $: Cool $base, $digits?, *%_)␤ in block <unit> at <tmp> line 1␤␤»
viki blasts www.youtube.com/watch?v=ndOi66foF7o and goes on another bug-fixing marathon
psch viki: ^^^ that's what i was thinking of wrt Whatever vs Block-closures! 15:45
viki mscha: * when it's a term on its own turns into a Whatever, not WhateverCode
psch mscha: the Whatever binds to the base method call, hence you're calling 108.base(*) 35 times, and the first of those already doesn't work because Whatever isn't meaningful as a base
15:46 cdg joined
mscha Thanks, viki, psch. 15:46
15:50 cdg left 15:57 wamba joined 15:58 kyclark_ joined 16:10 mxco86 left 16:11 araujo joined 16:17 seatek left 16:19 kyclark_ left 16:29 chris2 left 16:39 CIAvash left 16:40 mscha left 16:43 chris2 joined
viki grrr... need a faster box 16:45
I'm fixing bugs faster than spectest can finish...
or, put another, way: I'm breaking too many things in my fixes that constantly fixing them and re-running spectest is annoying ^_^ 16:47
16:49 cibs left 16:50 cibs joined
AlexDaniel viki: what about making rakudo faster? 16:51
:)
viki heh
m: say 579 / 154 16:53
camelia rakudo-moar bdc7a9: OUTPUT«3.759740␤»
viki my 24-core VM runs stresstest almost 4 times faster than my current box... That's a lot of making-rakudo-faster to do :)
But I can't use Atom on that VM ('cause I ssh)... Can't win :( 16:54
gfldex viki: you can tunnel nfs via ssh 16:57
viki What's nfs? Wasn't it a Windows file system or something? 16:59
RabidGravy well a Sun file system
viki I tried sshfs, but Atom just freezes, 'cause I think it tries to read all the files to make things in its tree view or something
AlexDaniel I use sshfs a lot 17:01
but yes, if your editor does that kind of stuff then it is a problem 17:03
viki: oh, another option is to rsync after you save 17:14
viki yah, worth tryingh 17:15
17:24 finanalyst left 17:26 firstdayonthejob joined 17:28 adu joined
viki m: dd Version.new("...|||||||.") 17:31
camelia rakudo-moar 085145: OUTPUT«Version.new␤»
viki m: dd Version.new("...|||||||.") leg v6.c
camelia rakudo-moar 085145: OUTPUT«Order::Less␤»
viki wonders how come bogus version strings in there don't just throw
Like there's a condition for that case: # get "v" highlander␤else { self.new } 17:32
17:32 Actualeyes left 17:33 telex left 17:34 telex joined 17:36 cdg joined 17:40 cdg left
viki Ah, I guess it's 'cause you can generally feed it any old slop and it'll make something sensible: 17:41
m: say Version.new("34\x[308]5:###meow77.|[43")
camelia rakudo-moar 938d4d: OUTPUT«v34̈5.meow.77.43␤»
geekosaur fsvo sensible 17:42
viki m: use v34̈5.meow.77.43 17:43
camelia rakudo-moar 938d4d: OUTPUT«===SORRY!===␤Error encoding ASCII string: could not encode codepoint 776␤»
FROGGS hehe 17:45
good CATch 17:46
viki It was Zefram, actually. I'm debugging 'it :)
vendethiel .u ĐÐ 17:48
yoleaux U+00D0 LATIN CAPITAL LETTER ETH [Lu] (Ð)
U+0110 LATIN CAPITAL LETTER D WITH STROKE [Lu] (Đ)
vendethiel ugh
AlexDaniel .u a􏿽xCC􏿽x88 17:49
yoleaux U+0061 LATIN SMALL LETTER A [Ll] (a)
U+0308 COMBINING DIAERESIS [Mn] (◌̈)
17:49 adu left
AlexDaniel u: a􏿽xCC􏿽x88 17:49
unicodable6 AlexDaniel, U+00E4 LATIN SMALL LETTER A WITH DIAERESIS [Ll] (ä)
AlexDaniel I wonder what kind of the output most people expect 17:50
viki my terminal is all messed up with those 17:51
i.imgur.com/qphXsZF.png
geekosaur doesn't handle utf8 very well, does it? 17:54
17:56 zakharyas joined 18:02 mxco86 joined, lizmat_ joined
tailgate can a multi-dispatch proto in NQP have different number of arguments? 18:03
18:05 lizmat left 18:07 mxco86 left, BenGoldberg joined
tailgate class methods, or subs I mean 18:07
viki FROGGS: is it actually a catch or DIHWIDT? 18:12
use v\d+blah is only for perl versions, innit 18:13
oh 18:14
v34̈5.meow.77.43
dasd
FROGGS would still be nice if it failed differently
viki m: say v34̈5.meow.77.43 v42
camelia rakudo-moar 2f72fa: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Two terms in a row␤at <tmp>:1␤------> 3say v34̈5.meow.77.437⏏5 v42␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement mod…»
viki m: say v34̈5.meow.77.43 cmp v42
camelia rakudo-moar 2f72fa: OUTPUT«Less␤»
18:18 revere joined 18:28 Lucas_One left 18:34 eisen74 joined
RabidGravy ok I think I may have hit the buffers on my MOP hackery du jour 18:35
viki m: say "a\x[308]" leg "b\x[308]" 18:36
camelia rakudo-moar 2f72fa: OUTPUT«More␤»
viki m: say "b\x[308]" leg "k\x[308]" 18:38
camelia rakudo-moar 2f72fa: OUTPUT«More␤»
viki m: say .ords for "b\x[308]", "k\x[308]"
camelia rakudo-moar 2f72fa: OUTPUT«(98 776)␤(107 776)␤»
viki How come it's More?
psch j: say "b\x[308]" leg "k\x[308]" 18:41
camelia rakudo-jvm 76b061: OUTPUT«Less␤»
psch r: use nqp; say nqp::cmp_s("b\x[308]", "k\x[308]") 18:43
camelia rakudo-jvm 76b061: OUTPUT«-1␤»
..rakudo-moar 2f72fa: OUTPUT«1␤»
geekosaur o.O 18:44
psch that's kinda interesting
apparently an MVMGrapheme32 is really just an MVMint32
18:45 cdg joined
psch and on moar the grapheme for "b\x[308]" isn't numericall (i.e. C) less-than than "k\x[308]" 18:45
'cause that's what cmp_s does, in the end
...well, minus "maybe my MVM src is out of date" :) 18:46
github.com/MoarVM/MoarVM/blob/mast...ps.c#L1397 18:47
RabidGravy I've got a weird that is a bit too complicated to post a snippet, but it involves two different EXPORTHOW::DECLARE thingies over-ridding ClassHOW in different ways
now if I nest them in a declaration, if there is a compose() over-ride on the inner one then it blows up with 18:48
"Representation must be composed before it can be serialized"
psch RabidGravy: shouldn't you export a new declarator instead of overriding the ClassHOW to get class to behave differently?
RabidGravy works fine without the compose() method
18:48 cdg left
psch i mean, i assume you have your reason, but maybe that could solve it? 18:49
RabidGravy psch that's what I'm doing
psch ah, okay, i guess i misinterpreted "overriding ClassHOW"
18:49 cdg joined
RabidGravy class EXPORTHOW::DECLARE::foo is Metamodel::ClassHOW does Somerole { } 18:49
18:50 andrzejku left
mst ohhh, EXPORT HOW, my brain was trying to parse it as <something> THROW 18:54
RabidGravy :)
psch well, the only thing i can think of is that you're not setting is_composed 18:55
or that some stuff happens out of order for some reason (which probably helps even less as a guess or advice) vOv
RabidGravy it certainly seems like something happening out of order in some way 18:56
all the compose do callsame 18:57
18:57 araujo left 19:02 eisen74 left 19:06 kyclark_ joined 19:10 araujo joined, araujo left, araujo joined 19:12 zakharyas left
psch RabidGravy: can't you golf it by removing most of the new stuff in the ClassHOW subclasses? 19:12
RabidGravy: maybe that already gives insight too
RabidGravy yeah just what I was thinking, it's tricky in the sense that you need three separate files 19:14
psch fwiw, that's a feature i wished for a while camelia would have 19:15
as in, gists with multiple files etc
not sure how easy that actually is, i think she creates tmp file names for gists
19:18 kyclark_ left
RabidGravy yep, it "works" in the degenerate case too, 'ang on I'll just them 19:18
19:19 zakharyas joined
AlexDaniel psch: yeah I know github.com/perl6/whateverable/issues/36 19:22
psch AlexDaniel: well, i suppose if whateverable does it camelia doesn't have to 19:23
AlexDaniel sure. Not yet though :(
RabidGravy psch, gist.github.com/jonathanstowe/9db3...794be0b19b - simply tested with "perl6 -I. -MBar"
psch but yeah, camelia does create 'evalbot-file-XXXXXXXXXX' which means module loading is kinda not doable
RabidGravy remove the compose from "inner" and it doesn't blow up 19:24
psch i suppose shifting everything into tmpdirs per eval call would be possible, but... idk
AlexDaniel once we actually start running stuff in containers for whateverable this won't be an issue at all
psch AlexDaniel: well, if whateverable annexes 'm:' at least... :)
19:25 andrzejku joined
psch RabidGravy: gotta build r-m first, but i'll try and see if i can figure something out 19:25
AlexDaniel if everything planned for whateverable gets implemented, I think we won't need camelia at all… :) 19:26
evalable6: .say for ^1000
evalable6 AlexDaniel, rakudo-moar 2f72fa0: OUTPUT«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␤31␤32␤33␤34…»
AlexDaniel, Full output: gist.github.com/1d5bdf24d4dc69e6c0...f0cd0e64be
AlexDaniel but that'll need some work. You can't even run something on jvm yet 19:27
psch fwiw, the whole EvalServer thing is actually pretty straight forward 19:28
although i suppose it strongly depends on what you're doing in whateverable and how you're doing it
AlexDaniel what is the startup time, by the way? 19:29
I'm wondering how long would it take to do one bisectable query…
psch well that's the thing
we don't do startup time on j: 19:30
because Eval*Server* ;P
AlexDaniel yea but we can't run 4 thousand evalservers
s
… or can we…
19:30 cdg left
timotimo how does java feel about getting swapped out completely to disk while it's waiting for a socket to receive connections? 19:30
psch i'm pretty sure the EvalServer starts a thread per eval request 19:31
mst timotimo: betrayed and unloved
timotimo yo mst
can we do something to make it easier for you to bring your makefile and configure improvements up stream? :)
RabidGravy so I had occassion to look at Scala again during the week, and I am still confirmed in my opinion of it 19:32
psch RabidGravy: huh, *running* Bar works, i.e. './perl6 -I. Bar.pm' 19:34
timotimo github.com/rakudo/rakudo/pull/324 - even though it has been resting for a long time, it'd be cool if this could make it into rakudo 19:36
19:37 stux|RC-only left
RabidGravy Clip Arser!! 19:37
19:40 stux|RC-only joined
AlexDaniel ohhh 19:40
so that's for proper command line argument parsing?
finally?
timotimo parsing arguments to rakudo itself 19:41
AlexDaniel ahmmm…
2 years ago, oops
psch yeah the bit about "join the whole command line invocation with \0 to parse it with <ws=\0>" wasn't too well received vOv
timotimo++ suggested to use some private unicode character for that, no idea if that alleviates those concerns 19:42
viki Clip arser? :) 19:43
timotimo that's what RabidGravy came up with 19:44
viki Glad I'm not the only one who read it that way. 19:45
psch RabidGravy: i don't think i can make anything of your MOP woes :| 19:46
RabidGravy yeah, it's way weird, I'm not quite willing to call bug on it as it's on the hairy side anyway 19:48
psch m: class A::B {...}; class A { class B { } }
camelia ( no output )
psch the curious thing is, stubbing like that ^^^ doesn't work for the inner
which is kinda weird
i had thought maybe that does something, but it just complains about "stubbed but not defined" 19:49
19:51 zakharyas left 19:57 ___crlz___ joined
RabidGravy the thing is that it appears to be the presence of the compose there, it seems to crap out without calling the provided compose 20:01
20:04 rpburkholder joined
dalek c: c72b2e9 | (Daniel Mita)++ | doc/Language/testing.pod6:
Remove duplicate done-testing text
 20:06
synopsebot6 Link: doc.perl6.org/language/testing
c: 66070d3 | (Aleks-Daniel Jakimenko-Aleksejev)++ | doc/Language/testing.pod6:
Merge pull request #1020 from mienaikage/remove-duplicate-text

Remove duplicate done-testing text
synopsebot6 Link: doc.perl6.org/language/testing
psch RabidGravy: well, the stub not working as it should might mean that it doesn't find the package and thus cannot compose it 20:09
RabidGravy: but that's a more or less completely blind guess, i'm really not too deep into our MOP, honestly 20:10
m: role R { }; class A does R { }; class B is A { }; B.^roles.say 20:11
camelia rakudo-moar 2f72fa: OUTPUT«((R))␤»
psch not sure why that would happen, anyway
i mean, missing something related to {Pseudo,}Stash
20:12 domidumont left 20:17 labster joined
RabidGravy yeah, maybe something is getting confused by it it not being a direct descendent of PackageHOW or summat 20:18
let's try that
20:20 ___crlz___ left
RabidGravy nope 20:21
20:22 ___crlz___ joined
RabidGravy this only came about because I was thinking about expressing the relationship between a schema and the tables in an ORM-like thing 20:22
vendethiel
.oO( He said it! The Pattern That Shall Not Be Named )
 20:26
RabidGravy :) 20:28
conceptually I don't find it difficult, it's finding a nice way to express things in a way that plays to Perl 6 strengths that is difficult 20:29
20:30 B3AN joined 20:31 B3AN left 20:37 adu joined 20:42 chris2 left 20:44 TEttinger joined
RabidGravy but doing it nicely with like "schema Foo { table Bar {}; table Baz {} };" seems off the cards for the problem stated above 20:45
20:52 rindolf left 20:58 chris2 joined, lizmat_ is now known as lizmat 21:00 rindolf joined
moritz perlpunks.de/paste/show/5839f868.76a3.3ad why does this die? 21:02
$.left seems to be a BinaryTree[Int], not an Int, no? 21:03
21:04 ranguard joined
geekosaur why are the type objects claimed to be T instead of Int by .perl? 21:05
moritz good question 21:06
is the exception an error in rakudo, or in me? 21:07
moritz too tired to tell
geekosaur code looks reasonable to me 21:08
geekosaur was wondering if maybe that stray T was actually a hint as to what was going wrong...
21:09 ranguard left
moritz I'll submit a rakudobug 21:11
AlexDaniel moritz: please tell if that works
ah, there was a bug reported 2 hours ago 21:12
so the issue is probably gone
moritz just checking if the bug still exists with today's rakudo: 21:16
m: gist.github.com/moritz/5fe8be21269...10b3f1ec58
camelia rakudo-moar 2f72fa: OUTPUT«LL[Int].new(item => 5, next => LL[Int].new(item => 4, next => LL[T]))␤5␤No such method 'visit' for invocant of type 'Int'␤ in method visit at <tmp> line 6␤ in block <unit> at <tmp> line 15␤␤»
21:16 adu left
geekosaur specifically wonders: if the type object ended up using T instead of Int, is something else confused and trying to use T instead of Int and complaining about *that* mismatch incorrectly? 21:19
moritz ok, here is the weird thing
the backtrace says it's in the inner invocation of that method 21:20
why didn't it call say() in there before recursing?
geekosaur it did. that was the 5. it's dying on the next call (since only one additional visit in the backtrace) 21:21
which is why I wonder if that one is somehow looking for, in the gist, LL[T] and throwing on an LL[Int] (and then reporting that badly) 21:22
(although that's a weird way to misreport that mismatch... then again iirc some similarly odd mismatch reports of that variety have happened) 21:23
moritz huh, when I change $.next to $!next, it works 21:28
21:30 adu joined
moritz things are going south here 21:33
I'm trying to prepare some slides for a training
first I get caught up in a rakudobug 21:34
now vim refuses to export syntax-hilighted HTML, which it did nicely a few hours ago
21:35 ranguard joined 21:41 rindolf left 21:44 espadrine left 21:48 trnh joined 21:50 cdg joined
tojo hmm, just installed rakudo with rakudobrew to debian jessie, but can't get IO::Socket::Async examples to work, but with osx precompiled rakudo they work 21:54
any hints about this or is it just me?
i don't get any error messages, but it doesn't open listening port 21:55
RabidGravy well they work on *Linux* because I wrote some of them
it's not so clever when it can't open the port to listen sometimes 21:56
21:56 andrzejku left
RabidGravy try a really high random port number 21:56
jnthn Could also check if it shows up in `neststat -l` 21:57
uh, netstat
tojo ah 21:58
tcp6 0 0 localhost:10033 [::]:* LISTEN
ipv6
how i can change to listening ipv4?
timotimo hm, do we allow to actually give the bind argument? 21:59
then it'd probably just be 0.0.0.0?
not sure if that throws out ipv6 reliabley?
tojo hm, with that it listens on every interface? 22:00
geekosaur it should... but I thought linux used the 6to4 stuff 22:01
so listening on [::] with a port < 65536 should also listen on 0.0.0.0 same port
timotimo what, does ipv6 have more ports? 22:02
(well, port is really a tcp-or-udp-related thing
but you get the idea)
geekosaur it's ulong instead of ushort for port number, yes 22:03
dalek c: f381605 | coke++ | doc/Type/Method.pod6:
fix typo
synopsebot6 Link: doc.perl6.org/type/Method
c: f94db38 | coke++ | xt/code.pws:
learn new code snippets
tojo but yeah, that 0.0.0.0 did the trick, thanks!
geekosaur you just don't see that much since it can't really be reflected into ipv4
RabidGravy for clients it's a pain as libuv seems to prefer v6 so if it can resolve an AAAA it will try that first
geekosaur and my machine isn't a good test as I only have ipv4 locally 22:04
hm, no, listed ip6 sockets and the autoconfigure local one is set. so I guess debianoids do distinguish them 22:05
22:06 dmaestro joined
RabidGravy ooh does localport work for clients? that would fix a longstanding issue on HUA regarding the tests on a v6 enabled network 22:08
localhost rather 22:09
geekosaur depends on what's in the hosts file
22:10 adu left, kurahaupo is now known as kurahaupo_, kurahaupo_ is now known as kurahaupo__ 22:17 FROGGS left
RabidGravy no, specifying the localhost to the client as 0.0.0.0 rather than letting it pick one (possibly v6) 22:20
geekosaur that should work if it lets you specify, yes 22:21
RabidGravy looks at the code
perlawhirl Does anyone wanna help me bike shed an option name before i push this module to ecosystem
github.com/0racle/p6-Pod-To-Markdown-Fenced
should the config option be 'info' instead of 'lang', given that Commonmark refers to that thing as the 'info string' 22:22
RabidGravy geekosaur, the code would indicate that the IO::Socket::INET will bind to a specfied localhost for both client and servers 22:24
22:31 adu joined 22:34 ___crlz___ left 22:36 ___crlz___ joined 22:37 wamba left 22:38 eisen74 joined 22:45 kyclark_ joined 22:48 pmurias left 22:53 seatek joined 22:56 kyclark_ left 22:57 bjz joined 22:58 canopus left 23:03 canopus joined 23:05 wamba joined 23:09 wamba left 23:21 RabidGravy left 23:23 Ven joined 23:25 kurahaupo__ left 23:30 Actualeyes joined, trnh left 23:33 chris2 left 23:34 mscha joined 23:35 mscha left 23:36 firstdayonthejob left 23:38 adu left 23:48 chris2 joined, cdg left 23:49 cdg joined 23:53 cdg left
viki titsuki: got reply from the admin: "Please try again. There was a configuration error that has been fixed." 23:53
23:58 canopus left

Powered by Cro and the Raku® Programming Language.

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

Thank you!