»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.
00:02 espadrine left 00:04 canopus left 00:12 canopus joined 00:26 cibs left
mj41 moritz: perlgeek.de/blog-en/perl-6/2016-bo...akudo.html ++ ... My Docker image is not official. You should recommend hub.docker.com/_/rakudo-star/ 00:26
00:28 cibs joined 00:34 lukaramu left 00:37 mj41 left, mr_ron left
seatek timotimo: checked and the adverb stacking actually is mentioned already 00:41
00:46 pierre_ joined 00:49 pierre_ left 00:50 pierre_ joined 00:54 pierre_ left 00:55 pierre_ joined, pierre_ left 01:07 ZzZombo_ left 01:19 BenGoldberg joined 01:28 kyclark_ joined, aborazmeh joined, aborazmeh left, aborazmeh joined 01:37 samcv joined 01:42 Tonik left 01:55 pierre_ joined 01:57 kyclark_ left, cdg joined 02:00 pierre_ left 02:10 xinming left 02:11 chinarulezzz left, xinming joined 02:13 kyclark_ joined 02:17 pierre_ joined 02:20 cdg left 02:22 kalkin- left, kalkin-_ joined, kyclark_ left 02:32 ZzZombo joined, ZzZombo is now known as Guest36651, Guest36651 left, Guest36651 joined, Guest36651 is now known as ZzZombo 02:37 zakharyas joined 02:40 pierre_ left 02:45 ilbot3 left 02:47 ilbot3 joined 02:53 kyclark_ joined 03:24 UgJkA left, noganex joined 03:26 ugjka joined 03:27 noganex_ left
[Coke] day 1 article is up and ready to go for 12/1. 03:30
03:31 ponbiki joined
[Coke] right now it says the url will be /11/28, but I am hoping that will be correct when the publish happens. 03:31
03:32 cognominal joined
dalek : ed0d227 | (Will Coleda)++ | misc/perl6advent-2016/schedule:
Note that this one is ready to go.
03:39
AlexDaniel the list is kinda… empty 03:46
I wish I had more time :-/
03:49 kyclark_ left 03:56 rpburkholder left 03:58 cibs left 03:59 cibs joined 04:32 kurahaupo joined 04:51 cdg joined 04:52 cdg left 05:00 kurahaupo left 05:01 kurahaupo joined 05:03 khw left 05:06 Cabanossi left 05:07 Cabanossi joined, kurahaupo left 05:17 pierre_ joined 05:41 xinming left 05:51 skids left 06:01 daph left 06:04 aborazmeh left 06:08 pierre_ left 06:25 AlexDaniel left 06:27 _28_ria left 06:30 _28_ria joined 06:39 domidumont joined 06:41 xfix left 06:45 bjz joined 06:46 domidumont left, domidumont joined 06:50 xinming joined 06:51 RabidGravy joined 06:52 rpburkholder joined 06:57 domidumont left 07:02 bjz left, eisen74 left 07:07 pierre_ joined, xinming left 07:08 xinming joined 07:14 pierre_ left 07:15 bjz joined 07:19 xinming left 07:24 domidumont joined 07:26 xfix joined 07:28 xinming joined 07:33 wamba joined 07:43 Actualeyes left 07:51 CIAvash joined 07:56 brillenfux joined 08:00 _28_ria left, firstdayonthejob joined 08:01 _28_ria joined 08:03 ugjka left 08:08 abraxxa joined
moritz mj41++ 08:13
08:13 BenGoldberg left
timotimo i'm not sure what to write about :S 08:13
gfldex i may have an adventpost unless it's to short. Is a short post better then none? 08:18
timotimo yes, a lot 08:19
08:22 CIAvash left 08:24 setty1 left
gfldex disengages the breaks 08:24
timotimo brakes? :)
don't forget people in here are willing to provide assistance 08:25
gfldex those too
i may need some input of jnthn, if he happens to be available
moritz timotimo: do a git log in all the repos you contributed to, find a commit that looks interesting, and talk a bit about it :-)
gfldex moritz: that's what I did :) 08:26
ZzZombo would it possible to provide a descriptive message when a parameter constraint fails?
be*
psch m: sub f($ where { $_ eq "foo" or die "can only do things with 'foo'!"}) { }; f "bar" 08:27
camelia rakudo-moar 054aca: OUTPUT«can only do things with 'foo'!␤ in sub f at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
moritz m: sub f($ where { $_ eq "foo" or fail "can only do things with 'foo'!"}) { }; f "bar" 08:29
camelia rakudo-moar 054aca: OUTPUT«can only do things with 'foo'!␤ in any at /home/camelia/rakudo-m-inst-2/share/perl6/runtime/CORE.setting.moarvm line 1␤ in sub f at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
psch i think there's a branch somewhere that stashes post constraints for more detail error messages too 08:30
seatek You cannot create an instance of this type -- that error message should be changed to "class not found" ;) 08:31
well, assuming... etc., etc., ;)
08:31 mj41 joined 08:32 CIAvash joined
moritz seatek: "not found" error messages are bad 08:34
seatek: why was it not found? did the runtime not try hard enough?
if it wasn't found because it doesn't exist, just say so 08:35
08:36 cibs left
seatek yeah i had spelled the class definition with an extra 't'... and took forever trying to figure out what it meany by that error message 08:36
meant 08:37
like a meany
a simple not found, or not defined, or doesn't exist or anything like that would have been a great clue
psch m: class Aa { }; A.new
camelia rakudo-moar 054aca: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared name:␤ A used at line 1␤»
seatek but it led me to thinking about types 08:38
08:38 cibs joined
seatek and then whether classes inheriting from classes that use roles could cause a problem... 08:38
etc., etc
undeclared name would be awesome :) 08:39
psch seatek: well, your error message apparently comes from something with a Lexotic REPR vOV 08:40
s/V/v/
seatek it was called from a test
so completely different scope
and the use name This::thing was the "file" that contained further definitions of This::thing::like:this 08:41
gfldex m: multi sub trait_mod:<is>(Label $l, :$foo){}; FOO is foo: for 1,2,3 { next FOO }
camelia ( no output )
seatek what's nice is that when i spell it right it works 08:48
timotimo does it actually run, gfldex? o_O 08:50
gfldex m: multi sub trait_mod:<is>(Label $l, :$foo){ say 'it does run or does it?' }; FOO is foo: for 1,2,3 { next FOO } 08:52
camelia rakudo-moar 054aca: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unexpected block in infix position (missing statement control word before the expression?)␤at <tmp>:1␤------> 3un or does it?' }; FOO is foo: for 1,2,37⏏5 { next FOO }␤ expecting any of:␤ …»
gfldex m: multi sub trait_mod:<is>(Label $l, :$foo){}; FOO is foo: for 1,2,3 { next FOO } 08:54
camelia rakudo-moar 054aca: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unexpected block in infix position (missing statement control word before the expression?)␤at <tmp>:1␤------> 3abel $l, :$foo){}; FOO is foo: for 1,2,37⏏5 { next FOO }␤ expecting any of:␤ …»
gfldex looks like it just stopped working o.0
DrForr Oh dear god, someone is actually *using* Slang::Roman.
timotimo :D :D 08:56
maybe Slang::Roman is a good topic for an advent post?
DrForr Well, at least submitted an issue.
08:56 rpburkholder left
DrForr And yes, I do want to do something, haven't gotten inspiration yet. 08:57
arnsholt Haha. Took me a second to realize that 0r was not standard Perl 6 notation
Too much SmallTalk recently =D
timotimo what does smalltalk do with 0r? 08:58
arnsholt Not 0r actually
But <base>r<digit> is how you write arbitrary-base numbers 08:59
So 16rDEADBEEF for hexadecimal for example
gfldex m: multi sub trait_mod:<is>(Label $l, :$foo){ say 'cally cally' }; FOO: for 1,2,3 { next FOO }; BEGIN &trait_mod:<is>(FOO, :foo); 09:00
camelia rakudo-moar 054aca: OUTPUT«cally cally␤»
gfldex timotimo: ^^^
timotimo pff :) 09:01
gfldex m: multi sub trait_mod:<is>(Label $l, :$foo){ say 'cally cally' }; FOO: for 1,2,3 { say 'but is it called in the right order?'; next FOO }; BEGIN &trait_mod:<is>(FOO, :foo); 09:02
camelia rakudo-moar 054aca: OUTPUT«cally cally␤but is it called in the right order?␤but is it called in the right order?␤but is it called in the right order?␤»
gfldex it is \o/
Xliff Is it better to use // in this situation: my $a = $this-might-not-be-defined // $default-val. 09:04
Or should I be using || 09:05
psch m: say 0 // 1
camelia rakudo-moar 054aca: OUTPUT«0␤»
psch m: say 0 || 1
camelia rakudo-moar 054aca: OUTPUT«1␤»
psch m: say Int // 1
camelia rakudo-moar 054aca: OUTPUT«1␤»
Xliff Yes. I am doing an if-not-defined-then....
timotimo for defined you need //
that's the defining characteristic :)
psch Xliff: infix:<//> is "$LHS.defined ?? $LHS !! $RHS"
Xliff: infix:<||> is "so $LHS ?? $LHS !! $RHS" 09:06
Xliff Then I am using it correctly.
Rather, I am doing what I thought I was supposed to be doing.
psch ...except i think so has lower precedence than ternary..?
Xliff And I do not have to rewrite a lot of code.
psch m: say so 0 ?? 0 !! 1
camelia rakudo-moar 054aca: OUTPUT«True␤»
psch yeah, looks like
Xliff Why the use of so? 09:07
timotimo to check precedence, i think?
Xliff (so $LHS) ?? $LHS !! $RHS
psch Xliff: just for being explicit, you're right it's not needed (and actually wrong because precedence)
Xliff ok
So I can go back into lurk mode.
Xliff lurks.
09:07 dakkar joined
dalek : 5f1feb2 | gfldex++ | misc/perl6advent-2016/schedule:
Update schedule
09:10
09:22 giraffe left 09:25 rindolf joined 09:26 zakharyas left 09:27 zakharyas joined 09:29 bjz_ joined 09:31 giraffe joined, bjz left, zakharyas left 09:33 jonas2 joined
dalek c: ecc15e2 | gfldex++ | doc/Language/glossary.pod6:
since we link to Multi-Dispatch, we might actually put a text there
09:34
synopsebot6 Link: doc.perl6.org/language/glossary
09:41 pierre_ joined 09:42 travis-ci joined
travis-ci Doc build errored. Wenzel P. P. Peppmeyer 'since we link to Multi-Dispatch, we might actually put a text there' 09:42
travis-ci.org/perl6/doc/builds/179377409 github.com/perl6/doc/compare/39046...c15e2983a4
09:42 travis-ci left 09:49 andrzeju joined, brillenfux left 09:50 andrzeju is now known as andrzejku 09:52 pierre_ left, pierre_ joined 09:56 gregf_ left 09:59 girafe left 10:02 Actualeyes joined 10:06 pierre_ left
seatek omg, omg, omg - TWEAK really does exist now! um.. i mean... yeah, nice evening we're having... 10:08
timotimo yeah, it does
i'm glad you noticed :)
seatek it's like early christmas! 10:09
timotimo i'm glad i could be santa for someone
seatek oh you did it! thanks timotimo++ :)
timotimo :3
seatek i'm not going to go back and undo my crazy BUILD gyrations from before, but from this point on... so much nicer 10:10
10:10 nowan_ left
timotimo cool 10:10
seatek was secretly holding out for TWERK
timotimo the mechanism that makes TWERK work is accessible from perl6 code, and overridable from a method you put into your own things 10:12
so you can have TWERK
seatek hehe :)
timotimo really, you can make anything happen if you want 10:13
randomize the order of initialization, for example
run a piece of code after every step of the buildplan, why not
seatek i would. but i've been having to deal with those Crust guys, and they're more anal retentive than a 50's propaganda housewife with cleaning chemicals and gloves 10:14
timotimo um, wat?
seatek absolute properness, cleanliness and conformity 10:15
that's my motto now
i'm gonna buy a uniform
10:15 CIAvash left 10:16 pierre_ joined, pierre_ left
timotimo it looks like crust hasn't been committed to for almost a year 10:16
10:16 pierre_ joined
timotimo so what keeps you from just forking and doing what you need? 10:16
seatek yes i'm tempted to fork it away
timotimo i thought these words you used were a bit excessively harsh
but what do i know :) 10:17
seatek i have a couple stupid severs that leave the environment with 2 environment variables in it HTTP_CONTENT_TYPE is one of them
it must not be there
it's called CONTENT_TYPE!!!!
if HTTP_CONTENT_TYPE is there as well, Crust kills itself 10:18
i just wanted to be able to turn off that check
10:18 nowan joined
timotimo having a clean environment is pretty important, have you not seen the recent thing where you could tell most servers on the 'net to use your privately supplied proxy server for all their own http requests? 10:18
seatek cuz the CGI spec says nothing about it nto being there... but no, they conform to PSGI whose document says that it must not be there. it doesn't say you have to die though
yes! but it's not even enabled (that check) unless you use their Lint filter -- if you enable that, which they suggest 10:19
Xliff Just FORK IT, man!
FORK IT!!!
10:19 brillenfux joined
Xliff throws up his hands in mock frustration. 10:20
seatek well, i submitted another pull request to the SCGI crust handler middlewear guy, to check for and delete those environment variables if they exist
but i hate that, becuase you're going through the environment YET AGAIN in the whole twisty, turny Crusty process
anyway, if i were to name something TWERK and those guys saw it, they'd have a anuryism. o 10:21
suppose you're a parent 10:22
you have a baby
and you don't want that baby to see something weird and crazy out there, cuz you think it might be dangerous
and some people from the world come knocking on your door.
what do you do? send them away? 10:23
no, you shoot your baby so it won't see anything weird
how does that make sense?
psch ...what
seatek they would not be good parents
Xliff Oh, that is some really twisted logic there.
seatek ;) i'm good at that :)
Xliff I've noticed.
8-)
seatek if the crust handler scgi guy doesn't take the pull request to delete those 2 stupid environment variables if they exist, i'll just copy the crust lint middleware and pull out the murderous fascism from it 10:25
Xliff FORK IT!!!!! 10:26
Fork the code fascists!!!
seatek Crust::Middleware::Lint::Dangerous
Xliff Fork 'em!
seatek Crust::Middleware::Lint::Egalitarian 10:27
masak TIL Perl 5 allows printing to a string by calling `open()` on a reference to the string. (since 5.8.0)
seatek that's bizarre. i've been writing so many tests lately that it sounds interesting too 10:28
did you know they used to use the moon as a communications satellite? 10:31
timotimo bounce things off of it? 10:32
seatek yup
10:32 ufobat joined
timotimo makes some sense, the moon is visible from a broad chunk of the earth at the same time 10:32
seatek i've been reading this book on the formation of Iridium that satellite company 10:33
10:33 pierre_ left
DrForr And creator of Iridium flares :) 10:33
seatek i never understood why they were called that either :)
they chose the name Iridium because it was the atomic number of the number of satellites they needed 10:34
Xliff screams in frustration as he realizes he needs to refactor a bunch of code he just finished writing.
seatek has a brief twinge of empathy for Xliff, then has more pie 10:35
Xliff Hope your pie-hole doesn't choke on that pie!
moritz Xliff: better refactor it now, when you're still familiar with it, than in half a year
Xliff moritz: Yes. However refactoring is best done after sleep.
Which I am lacking. 10:36
seatek you'll forget everything by morning
Xliff Insomnia is a pain in the ass.
Xliff glares at seatek.
I'll have you know I make copious comments. I'll be fine. :P
10:37 darutoko joined
seatek i started using a blue filter thingy for nighttime work, and i get tired quicker now 10:37
Xliff blue. filter. hunh? 10:38
timotimo throws out blueish colors
think of it as going into submarine mode
Xliff Filter on what?
timotimo displays
Xliff Oh.
timotimo you may know it as f.lux or redshift
seatek yeah it makes it all subdued
timotimo yesterday i learned from a friend that you can turn the color temperature even lower than the default
Xliff www.google.com/url?sa=t&rct=j&...l47fM9ofVA 10:39
???
Oh dammit, google.
timotimo i have it set to 2500 during the night now, and it's really, really, really red
10:39 labster left
Xliff www.redshift3d.com/ ??? 10:39
timotimo no
seatek have you tried turning it off timotimo ? i do that sometimes and it's like burning my eyes
timotimo jonls.dk/redshift/
yes
Xliff Which is better?
seatek non-burt eyes 10:40
Xliff Correction... which one are you using?
timotimo they are not comparable, Xliff
seatek what OS? i'm using ubuntu on on it....
redshift
on windows i use f.whatever that was 10:41
f.lux?
yes f.lux
i like redshift better though 10:42
oo thanks for that link timotimo i'm gonna play with the config file
timotimo i'm sad that i can't set brightness above 1.0 :P 10:43
Xliff hacks registry.
seatek i never tried reshift on windows 10:44
it's experimental...
moritz fwiw on android I use twilight 10:45
seatek that might not be a bad idea either! 10:46
moritz though in the morning I sometimes have to switch it off because it causes some jumps in display brightness
Xliff What is a good blue-shifted temp for redshift? 10:49
seatek mine's at 3500 right now and it's nice and dreamy 10:51
nine just avoids looking at a screen before going to bed...
seatek i lack that kind of willpower
Xliff Wow! 10:52
nine Proper sleep does improve willpower btw ;)
Xliff That is easier on the eyes.
Now I'm getting all.s..zzzzzzzzzzzz
tadzik I'm at 5500 10:53
seatek nine - is your scope-tightening stuff, with the need to put more use's in, going to effect compile times at all? or is the compiler going to be smart enough about them?
Xliff: nice isn't it! :)
Xliff is sleeping. Do not disturb!
10:54 Xliff is now known as Xliff_ZZZzzzz, rindolf left
ZzZombo m: class A{has $!a;method m{$!a}};say A.new.m 10:55
camelia rakudo-moar 054aca: OUTPUT«(Any)␤»
10:55 RabidGravy left
seatek i'm starting to get paranoid about compile times now, with the size of the thing i'm working on, and have resorted to developing by just using tests.. but that' even getting long now... so i've resorted to simulating the use of some classes when testing even... 10:57
i mean, if it comes to the point where i actually have to think about what i'm typing before i type it!... 10:58
i kinda like compiling though. it's got a certain... zen about it 10:59
10:59 rindolf joined
ZzZombo uh, how do I check a value is specifically not Any? 10:59
seatek .defined
ZzZombo no
only Any, not Nil or anything else
DrForr $a !~~ Any # ? 11:00
psch m: say 1 ~~ Any
camelia rakudo-moar 054aca: OUTPUT«True␤»
DrForr True, Any is the root of too many things.
seatek i think that if anything is something it's not Any
11:01 brillenfux left
psch if you're sure you never get Any.new you can do !=:= Any 11:01
m: say Any =:= Any; say 1 =:= Any; say Nil =:= Any
camelia rakudo-moar 054aca: OUTPUT«True␤False␤False␤»
jnthn $value !=== Any
psch m: say Any.new === Any
camelia rakudo-moar 054aca: OUTPUT«False␤»
seatek it's like the infinities again, only crazier 11:02
ZzZombo eh, that works! Thanks.
11:02 ufobat left
seatek how do you set a variable to Any? 11:02
ZzZombo my $a=Any? 11:03
seatek m: my $i = 1; $i = Any;
camelia ( no output )
lizmat my $a
ZzZombo or leave it andefined
undefined*
lizmat m: dd my $a
camelia rakudo-moar 054aca: OUTPUT«Any $a = Any␤»
gfldex m: my Any $a; # if you want to be explicit
camelia ( no output )
nine seatek: I don't think it will effect compile times. Using an already loaded module doesn't cost much. 11:08
seatek nine: nice :) thank you :)
m: my $i = 3; say $i.WHAT 11:09
camelia rakudo-moar 054aca: OUTPUT«(Int)␤»
seatek m: my $i = 3; say $i.WHAT; $i = Any; say $i.WHAT
camelia rakudo-moar 054aca: OUTPUT«(Int)␤(Any)␤»
seatek m: my $i = 3; say $i.WHAT; $i = Any; say $i.WHAT; my Int $c = 2; $c = Any;
camelia rakudo-moar 054aca: OUTPUT«(Int)␤(Any)␤Type check failed in assignment to $c; expected Int but got Any (Any)␤ in block <unit> at <tmp> line 1␤␤»
seatek that's how i get confused about Any's
psch why? 11:10
Any is the default type constraint
m: sub f($) { }; say &f.signature.params[0].type
camelia rakudo-moar 054aca: OUTPUT«(Any)␤»
seatek but $i is an Int
can ints be set to Any?
psch yeah, but liskov only works downwards not upwards
gfldex m: say Any.WHO, Any.new.WHO
camelia rakudo-moar 054aca: OUTPUT«AnyAny␤»
psch seatek: you assign an Int to a variable that is constrained to Any 11:11
m: say Int.^mro
camelia rakudo-moar 054aca: OUTPUT«((Int) (Cool) (Any) (Mu))␤»
psch seatek: Int isa Any
gfldex ZzZombo: you have to check .WHO
psch seatek: but you can't assign an Any to a variable constrained to Int
seatek ah.. .ok got it
yeah that makes perfect sense now
ZzZombo why?
gfldex because a test against .WHO does not care about inheritance 11:12
seatek that's actually kinda neat more than confusing now :)
ZzZombo like this: 11:13
if $value.WHO!===Any.WHO
?
gfldex yes 11:14
11:15 pierre_ joined
gfldex m: sub Any(){Int}; dd Any, Any() 11:16
camelia rakudo-moar 054aca: OUTPUT«Any␤Any(Any)␤»
gfldex m: sub Any(){'oi‽'}; dd Any, Any()
camelia rakudo-moar 054aca: OUTPUT«Any␤Any(Any)␤»
psch m: dd Any(StR) 11:17
camelia rakudo-moar 054aca: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared name:␤ StR used at line 1. Did you mean 'str', 'Str'?␤␤»
psch m: dd Any(Str)
camelia rakudo-moar 054aca: OUTPUT«Any(Str)␤»
psch can't overwrite coercers, they're hiding deep in the guts
jnthn It's a parse-time distinction.
gfldex it should complain when I try to define that sub
jnthn Why?
psch m: sub Any() { say "not really Any" }; say &Any() 11:18
camelia rakudo-moar 054aca: OUTPUT«not really Any␤True␤»
11:18 wamba left
gfldex because I'm stupid and Rakudo is not 11:18
anyway that's a ENODOC that I shall act upon
11:20 chinarulezzz joined
dalek c: 53ef347 | gfldex++ | doc/Type/Sub.pod6:
show how to call subs that (don't) overload coercers
11:24
synopsebot6 Link: doc.perl6.org/type/Sub
11:32 travis-ci joined
travis-ci Doc build errored. Wenzel P. P. Peppmeyer 'show how to call subs that (don't) overload coercers' 11:32
travis-ci.org/perl6/doc/builds/179404506 github.com/perl6/doc/compare/ecc15...ef347a7508
11:32 travis-ci left
seatek traavisssss.... 11:33
jnthn TIL you can use $!foo in a signature anywhere inside a method, not just in the method signature itself
For example 11:34
method foo($bar) {
dalek c: 8087f17 | gfldex++ | doc/Language/typesystem.pod6:
we can use the MOP to add coercers
synopsebot6 Link: doc.perl6.org/language/typesystem
jnthn with resolve($bar) -> $!bar { }
psch ah, the 'just' works as an 'only' there
cause i had some trouble parsing that 11:35
jnthn Hm :)
Yes, "not only in the method sig" is also a valid way to put it :)
It saved me a $!bar = $_; inside of the block, anyway
psch neat
jnthn Funny thing being, I never deliberately made that work. :) 11:36
It just naturally falls out of how attributive parameters are implemented.
Only just now was writing something and was like "hmm, I wonder if I could just..." :) 11:37
psch r: class A { has $.foo; method bar($baz) { with $baz -> $!foo { } } }; A.new.bar("quux").foo # aw :|
camelia ( no output )
..rakudo-jvm 76b061: OUTPUT«Unable to bind attributive parameter '$!foo' - could not find self␤ in method bar at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
jnthn Aww
Wonder why.
psch probably because my impl of attributive binding isn't quite up to snuff :S 11:38
jnthn Maybe it's not looking in outer scopes for the self...
(Which this relies upon)
psch seems like this warrants a test, too
anyway, lunch time for me :)
jnthn ooh, lunch ain't a bad idea 11:39
masak psch: you want me to rakudobug that_
?
viki masak: I was waiting for you to appear :) I have a Q
masak go ahead 11:40
viki masak: what's your opinion on printf having another candidate: (IO::Handle $fh, Cool $format, *@args)... that's adding to the current (Cool $format, *@args)... I recall you were against such variable args
Currently there's no functional way to printf to a filehandle and Perl 5 does use the same scheme I'm proposing. 11:41
(by functional I mean programming style; there is an OO way) 11:42
jnthn m: print $*OUT, 'hi'
camelia rakudo-moar 054aca: OUTPUT«This type cannot unbox to a native string: P6opaque, IO::Special␤ in block <unit> at <tmp> line 1␤␤»
jnthn Note it'd be incosistent with other I/O functions
I'd say it should just me a method printf
m: print $*OUT: 'hi'
camelia rakudo-moar 054aca: OUTPUT«hi»
viki OK
jnthn m: printf $*OUT: 'hi'
camelia rakudo-moar 054aca: OUTPUT«Use of uninitialized value of type List in numeric context␤␤Use of uninitialized value of type List in numeric context␤␤hi»
jnthn ... 11:43
And you can always do that, but what are the warnings about?
m: $*OUT.printf('hi')
camelia rakudo-moar 054aca: OUTPUT«Use of uninitialized value of type List in numeric context␤␤Use of uninitialized value of type List in numeric context␤␤hi»
viki unsure. I only added that last night
jnthn m: $*OUT.printf("hi %d", 42)
camelia rakudo-moar 054aca: OUTPUT«hi 42»
viki m: printf 'meow' 11:44
camelia rakudo-moar 054aca: OUTPUT«meow»
viki weird. I'll take a look into that today
s: $*OUT, 'printf', \('x')
SourceBaby viki, Sauce is at github.com/rakudo/rakudo/blob/7c1f...le.pm#L656
jnthn really lunch & 11:46
viki m: use nqp; sub foo (*@args) { nqp::sprintf('x', nqp::clone(nqp::getattr(@args, List, '$!reified'))) }; foo 11:47
camelia rakudo-moar 054aca: OUTPUT«Use of uninitialized value of type List in numeric context␤␤Use of uninitialized value of type List in numeric context␤␤»
viki m: use nqp; sub foo (*@args) { nqp::sprintf('x', nqp::clone(nqp::getattr(@args, List, '$!reified'))) }; foo []
camelia ( no output )
viki m: use nqp; sub foo (*@args) { nqp::sprintf('x', nqp::clone(nqp::getattr(@args||[], List, '$!reified'))) }; foo 11:49
camelia ( no output )
11:51 grondilu joined
viki m: sprintf 'x' 11:52
camelia rakudo-moar 054aca: OUTPUT«Use of uninitialized value of type List in numeric context␤␤Use of uninitialized value of type List in numeric context␤␤»
viki hehehe :)
11:56 travis-ci joined
travis-ci Doc build passed. Wenzel P. P. Peppmeyer 'we can use the MOP to add coercers' 11:56
travis-ci.org/perl6/doc/builds/179406783 github.com/perl6/doc/compare/53ef3...87f1782f97
11:56 travis-ci left
ZzZombo There is way to look what package something comes from, isn't there? 11:58
viki .WHO 12:00
m: package Meows; { sub foos is export {} }; import Meows; say &foos.WHO
camelia rakudo-moar 054aca: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤This appears to be Perl 5 code. If you intended it to be Perl 6 code, please use a Perl 6 style declaration like "unit package Foo;" or "unit module Foo;", or use the block form instead of the semicolon form…»
viki m: package Meows { sub foos is export {} }; import Meows; say &foos.WHO 12:01
camelia rakudo-moar 054aca: OUTPUT«Sub␤»
viki heh
m: package Meows { class foos {} }; import Meows; say Meows::foos.WHO 12:02
camelia rakudo-moar 054aca: OUTPUT«Meows::foos␤»
viki m: package Meows { class foos {} }; import Meows; say Meows.WHO
camelia rakudo-moar 054aca: OUTPUT«Meows␤»
viki cues in "I've no idea what I'm doing" meme 12:03
12:04 canopus left
ZzZombo Attribute::Handling is a builtin type, not an external module 12:07
wut
m: dd Attribute::Handling 12:09
camelia rakudo-moar 054aca: OUTPUT«Could not find symbol '&Handling'␤ in block <unit> at <tmp> line 1␤␤Actually thrown at:␤ in block <unit> at <tmp> line 1␤␤»
ZzZombo m: use Attribute::Handling 12:10
camelia rakudo-moar 054aca: OUTPUT«===SORRY!===␤Could not find Attribute::Handling at line 1 in:␤ /home/camelia/.perl6␤ /home/camelia/rakudo-m-inst-2/share/perl6/site␤ /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤ /home/camelia/rakudo-m-inst-2/share/perl6␤ Com…»
ZzZombo W. T. F.
moritz who says that Attribute::Handling is a built-in type? 12:11
ZzZombo perl6
12:12 wamba joined
ZzZombo dayum 12:12
changed 'use' to 'import' and it worked.
WTF!!!
12:13 canopus joined 12:14 gregf_ joined, CIAvash joined
viki ZzZombo, a bug 12:17
ZzZombo Dear God... 12:18
viki There is no god 12:19
Does Attribute::Lazy suffer the same issue?
ZzZombo ah, you were supposed to answer with "There is more"
:(
Could not find Attribute::Lazy 12:20
I mean
"Could not find Attribute::Lazy"
viki I recall there being a bug with stuff that has core namespace in it, like IO::Whatever, so maybe this is the same
*shrug* 12:21
m: use Attribute::Handling
camelia rakudo-moar 054aca: OUTPUT«===SORRY!===␤Could not find Attribute::Handling at line 1 in:␤ /home/camelia/.perl6␤ /home/camelia/rakudo-m-inst-2/share/perl6/site␤ /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤ /home/camelia/rakudo-m-inst-2/share/perl6␤ Com…»
viki :/
Looks fine on camelia
ZzZombo That package is defined in the main Perl6 file.
That I use'd in the same file. 12:22
psch m: package Attribute::Foo { }; use Attribute::Foo
camelia rakudo-moar 054aca: OUTPUT«===SORRY!===␤Attribute::Foo is a builtin type, not an external module␤»
viki m: package Meows {}; use Meows
camelia rakudo-moar 054aca: OUTPUT«===SORRY!===␤Could not find Meows at line 1 in:␤ /home/camelia/.perl6␤ /home/camelia/rakudo-m-inst-2/share/perl6/site␤ /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤ /home/camelia/rakudo-m-inst-2/share/perl6␤ CompUnit::Reposit…»
viki aha
ZzZombo, would you report that bug? 12:23
ZzZombo how?
viki huggable, rakudobug
huggable viki, [email@hidden.address] or use perl6 query on rt.perl.org ; see github.com/rakudo/rakudo/#reporting-bugs
viki email to ^ that address
ZzZombo ye
viki Thanks
ZzZombo BTW, why you can't do 12:24
sub x(Type self: |blah)
as opposed to
sub x(Type $self: |blah)
?
psch m: sub x(Int $self: ) { } 12:25
camelia rakudo-moar 054aca: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Can only use the : invocant marker in the signature for a method␤at <tmp>:1␤------> 3sub x(Int $self: 7⏏5) { }␤»
psch you can't do either?
m: sub x(Int self: ) { }
camelia rakudo-moar 054aca: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Malformed parameter␤at <tmp>:1␤------> 3sub x(Int7⏏5 self: ) { }␤ expecting any of:␤ constraint␤ formal parameter␤»
ZzZombo eh, I meant method, sorry
psch 'self' is a special name that automatically exists lexically inside method bodies
m: class A { method f(A:D: ) { say self.perl } }; A.new.f 12:26
camelia rakudo-moar 054aca: OUTPUT«A.new␤»
psch hence you don't even need a name for the invocant
ZzZombo well, I thought for the same of explictioness, you should be able to do that.
psch well, you'd have to do it the normal way for sigilless parameters 12:27
m: class A { method f(A:D \self: ) { say self.perl } }; A.new.f
camelia rakudo-moar 054aca: OUTPUT«===SORRY!===␤Lexical 'self' already declared␤»
viki m: method (Int $self:) {}
camelia ( no output )
psch but it's already there, sooo 12:28
viki works fine?
m: class { method (Int $self:) {} }.new
camelia ( no output )
psch i suppose the \self case might be reasonable to make work, but it feels a bit weird honestly
CIAvash .tw CIAvash 12:29
yoleaux Added Perl 6 to alternativeTo #Perl6 alternativeto.net/software/perl-6/ (@CIAvash)
12:30 lucasb_ joined
masak viki: backlogging. I agree with jnthn; it should be a method. 12:31
viki: I don't remember being against such an extra candidate -- but consider me against it now :> 12:32
viki masak, it was when we were bikesheding comb 1, '...'
and you said something along the lines that variadics before normal params were a code smell 12:34
ZzZombo Dammit 12:35
viki .tw Zoffix
yoleaux Rakudo Star #Perl 6 Distro version 2016.11 is out 🎊🍻🎈🍕😀 rakudo.org/2016/11/27/announce-raku...e-2016-11/ #Programming #Perl (@zoffix)
viki neat
ZzZombo I forgot to fill in the topic of the letter
lucasb_ what is .tw?
ZzZombo will it get through?
viki twitter
lucasb_ ahh, interesting 12:36
viki ZzZombo, should, yeah
lucasb_ last tweet...
ZzZombo sweet
lucasb_ will TimToady be doing a blog post this year?
viki: I see you alreadry grabbed the 24th day. maybe he posts something on the 25th 12:37
ZzZombo <psch> i suppose the \self case might be reasonable to make work, but it feels a bit weird honestly
Almost everywhere I can see Perl allows to express yourself explicitly, however weird it may be.
psch ZzZombo: patches welcome :) it's probably somewhere around github.com/rakudo/rakudo/blob/nom/...P.nqp#L471 12:38
viki ZzZombo, it works tho, see my examples above
psch viki: it's explicitly about explicitly introducing a sigilless self in the method lexpad
viki: ZzZombo's point, that 12:39
viki ah
psch m: method (\self:) { } # as this forbids currently 12:40
camelia rakudo-moar 054aca: OUTPUT«===SORRY!===␤Lexical 'self' already declared␤»
viki hard to read code on this tiny thing :) imgur.com/D3TqoqI
psch ...that's really just your own fault :P 12:41
viki :)
psch ZzZombo: it might well be that you just gotta add a check in the if near the linked line that checks if there already is a lexical 'self'
viki w00t! I found a secret way to create RT tickets without using email :) 12:43
( mobile ) 12:44
12:45 wamba left
viki ZzZombo, so you're from Russia? 12:45
ZzZombo Yes, why? 12:46
viki What city?
ZzZombo psch: maybe later, when I finish with this thing.
Ulan-Ude.
viki Ah. I lived in Novosibirsk for 15 years 12:47
ZzZombo Wow, how's that been?
viki It was ok
jnthn visited Ulan Ude years ago 12:49
It's hard the unsee that huge Lenin head on the square :P
12:49 user9 left
ZzZombo :) 12:51
ZzZombo throws a snowball at viki and jnthn
From Russia with love :)
jnthn :)
Yay, some snow. I'm still waiting to get that here this winter... :) 12:52
dalek href="https://perl6.org:">perl6.org: 7cbce3d | (Steve Mynott)++ | source/ (2 files):
refs to Rakudo Star 2016.11
12:54
12:55 pmurias joined 12:56 grondilu left 12:57 chinarulezzz left 12:58 eliasr joined, user9 joined 13:01 pierre_ left 13:02 dmaestro left
ZzZombo m: put Bool::True 13:04
camelia rakudo-moar 054aca: OUTPUT«True␤»
13:05 TEttinger left, brillenfux joined
viki m: put Bool::True.perl 13:06
camelia rakudo-moar 054aca: OUTPUT«Bool::True␤»
13:09 pierre_ joined
ZzZombo m: my %x;%x<a>=Nil;%x<a>.get_value(); 13:09
camelia rakudo-moar 054aca: OUTPUT«No such method 'get_value' for invocant of type 'Any'␤ in block <unit> at <tmp> line 1␤␤»
ZzZombo why "Any"? It's Nil. 13:10
It got me confused.
psch m: my $x = Nil; say $x
camelia rakudo-moar 054aca: OUTPUT«(Any)␤»
psch Nil is literally no value, which means assigning Nil to something is resetting it to the default
13:10 AlexDaniel joined
psch m: my $x is default(42) = 10; say $x; $x = Nil; say $x 13:10
camelia rakudo-moar 054aca: OUTPUT«10␤42␤»
ZzZombo ugh
psch ZzZombo: you seem to be X/Y-ing :) 13:11
ZzZombo: or, well, maybe exploring semantics vOv
jnthn Nil means you don't have to worry what type something was declared with in order to clear it 13:12
ZzZombo But then how do I know for sure for example, .first() just didn't found anything instead of finding a value that happened to be Any. I always thought of Nil like 'there is nothing to return, but everything is totally okay', 13:15
find*
dalek c: 29fd20f | gfldex++ | doc/Language/functions.pod6:
link to Method and Sub
synopsebot6 Link: doc.perl6.org/language/functions
c: 35b4489 | gfldex++ | doc/Language/functions.pod6:
methods can exist independently of objects
synopsebot6 Link: doc.perl6.org/language/functions
psch m: say <a b c>.first: { $_ eq .uc } 13:18
camelia rakudo-moar 054aca: OUTPUT«Nil␤»
psch m: say [Int, Str, Any].first: { $_ === Any }
camelia rakudo-moar 054aca: OUTPUT«(Any)␤»
psch doesn't understand the question
ZzZombo I dunno, it's what I got. 13:19
jnthn I think the question is "if I assign it into a variable, the Nil vanishes, so what do I do?"
viki If you store the result in a variable you won't be able to tell the difference
jnthn m: my $a = [Int, Str, Any].first: { $_ === Any }; say $a
camelia rakudo-moar 054aca: OUTPUT«(Any)␤»
psch oh, right, on assignment it gets weird
jnthn m: my $a = [Int, Str].first: { $_ === Any }; say $a
camelia rakudo-moar 054aca: OUTPUT«(Any)␤»
jnthn Like that for example 13:20
m: my $a := [Int, Str].first: { $_ === Any }; say $a
camelia rakudo-moar 054aca: OUTPUT«Nil␤»
jnthn m: my $a := [Int, Str, Any].first: { $_ === Any }; say $a
camelia rakudo-moar 054aca: OUTPUT«(Any)␤»
jnthn But the behavior is a property of assignment, so just don't assign. :)
psch clearly we need a with-derivate that distinguishes only on === Nil
(we don't actually need that i think)
jnthn Not really.
How often are you actually working with type objects?
psch well, they usually come as SixModelObject when i'm working with them :) 13:21
jnthn Note that if you were doing .first({ $_ === Nil }) you're back to square one again :)
gfldex m: my @a = <a b c>.first: { $_ eq .uc }; say @a ~~ (), @a ~~ Empty; 13:22
camelia rakudo-moar 054aca: OUTPUT«FalseFalse␤»
gfldex i would have goess that Empty smart matches against an empty Array 13:23
viki Well, it's a Slip
s: Slip, 'ACCEPTS', \([]) 13:24
SourceBaby viki, Sauce is at github.com/rakudo/rakudo/blob/843a...Any.pm#L19
viki s: Slip.new, 'ACCEPTS', \([])
SourceBaby viki, Sauce is at github.com/rakudo/rakudo/blob/843a...st.pm#L760
viki :o
m: say Slip.^mro
camelia rakudo-moar 054aca: OUTPUT«((Slip) (List) (Cool) (Any) (Mu))␤»
gfldex i would also expect the empty list to match against an empty Array
viki s: &infix:<~~>, \([], Empty) 13:25
SourceBaby viki, Sauce is at github.com/rakudo/rakudo/blob/843a...Mu.pm#L818
gfldex m: say List.new ~~ (), List.new ~~ Empty;
camelia rakudo-moar 054aca: OUTPUT«TrueTrue␤»
viki I'm guessing this is the same issue as with andthen + orelse
The Empty vanishes when passed to infix:<~~> 13:26
viki has no idea
gfldex can it be special cased?
13:27 aborazmeh joined, aborazmeh left, aborazmeh joined
viki The output from SourceBaby would suggest it doesn't vanish 13:27
gfldex m: my @a; say @a.list ~~ (), @a.list ~~ Empty;
camelia rakudo-moar 054aca: OUTPUT«TrueTrue␤»
viki m: say [] ~~ Empty; 13:28
camelia rakudo-moar 054aca: OUTPUT«True␤»
viki m: my @a = <a b c>.first: { $_ eq .uc }; dd @a
camelia rakudo-moar 054aca: OUTPUT«Array @a = [Any]␤»
viki Well, there's your problem
:D
gfldex m: my @a := <a b c>.first: { $_ eq .uc }; dd @a 13:29
camelia rakudo-moar 054aca: OUTPUT«Type check failed in binding; expected Positional but got Nil (Nil)␤ in block <unit> at <tmp> line 1␤␤»
13:29 cpage left
ZzZombo why do you use arrays for stroing .first() results? 13:29
storing*
gfldex because Array is to go-to-class for things that have more then one value 13:30
viki Can you get more than one value with .first?
gfldex if it's a LoL i might
viki But it'd still be 1 value 13:31
m: my @a = [<a b c>, <f g e>].first: { $_ eqv <a b c> }; dd @a
camelia rakudo-moar 054aca: OUTPUT«Array @a = [("a", "b", "c"),]␤»
timotimo yeah, you'll need to grep + [^2] or something
gfldex in that case i'm just confused
still I will rakudobug @a ~~ () to get a fix or a docable answer 13:32
no, i wont :) 13:33
m: say Array.new ~~ (), Array.new ~~ Empty;
camelia ( no output )
viki erm why is that no output?
gfldex that just worked in a /query to camelia
viki m: say Array.new ~~ (), Array.new ~~ Empty;
camelia rakudo-moar 054aca: OUTPUT«TrueTrue␤»
viki chalks it up to camelia bug 13:34
gfldex m: say Seq.new ~~ (), Seq.new ~~ Empty;
camelia rakudo-moar 054aca: OUTPUT«Too few positionals passed; expected 2 arguments but got 1␤ in block <unit> at <tmp> line 1␤␤»
gfldex m: my @a; say @a.Seq ~~ (), @a.Seq ~~ Empty;
camelia rakudo-moar 054aca: OUTPUT«TrueTrue␤»
ZzZombo Alright, I hope you will catch that pesky bug, if it's not a phantom. Gotta catch some sleep instead. 13:35
gfldex i somehow like @a ~~ () better then @a.elems == 0
viki m: package Attribute::Foo { }; use Attribute::Foo 13:36
camelia ( no output )
viki m: package IO::Foo { }; use IO::Foo
camelia ( no output )
viki m: package IO::Foo { }; use IO::Foo
camelia rakudo-moar 054aca: OUTPUT«===SORRY!===␤»
gfldex m: my @a; say so @a.Seq ~~ ()|Empty; 13:37
camelia ( no output )
gfldex m: my @a; say so @a.Seq ~~ () | Empty;
camelia ( no output )
13:37 camelia left, camelia joined
gfldex i broke camelia :( 13:37
viki I restarted her 13:38
13:39 ChanServ sets mode: +v camelia
lucasb_ m: package IO::Foo { }; use IO::Foo 13:39
camelia rakudo-moar 054aca: OUTPUT«===SORRY!===␤IO::Foo is a builtin type, not an external module␤»
13:39 pierre_ left
viki ah crap. it was just a camelia bug :( 13:40
lucasb_ how come this snippet changed from previous evaluation?
viki already commented on RT ticket
13:40 pierre_ joined
viki lucasb_: camelia was sick. 13:40
lucasb_ ah
the ticket about get and prompt in REPL... 13:43
if you call get/prompt, then it'll wait for you to type something and press enter, even in the REPL 13:44
viki That's a dupe, there's already a ticket from 2016.04 or something
lucasb_ or maybe it's something broken on the windows system of the user
viki: what? is it a valid issue?
I thougth it was just a misunderstanding 13:45
or mis-expectation
viki lucasb_: yeah, it just hangs
I just tried in 2016.04, and I'm guessing that's still present in 2016.11
I pressed Enter, CTRL+D, CTRL+C nothing does anything 13:46
but it works fine outside the REPL
dalek c: 2cf9b86 | gfldex++ | doc/Type/List.pod6:
we should list () as a way to create List, also show ~~ against ()
synopsebot6 Link: doc.perl6.org/type/List
lucasb_ well, in my environment it's different
viki This is on Win7 Professional 13:47
lucasb_ ah, right. sorry
viki prefers just evaulaing @a in bool context instead of either @a ~~ () or @a.elems == 0 :) 13:48
DrForr Win7 "professional"... Heh.
13:48 pierre_ left 13:49 seatek left
viki In fact, @a.Bool can be faster than @a.elems, 'cause .elems reifies everything, but .Bool just 1 item 13:49
DrForr: ? what
gfldex does ~~ () call .Bool ?
viki s: &infix:<~~>, \([], ()) 13:50
SourceBaby viki, Sauce is at github.com/rakudo/rakudo/blob/843a...Mu.pm#L818
viki s: &infix:<~~>, \([], (),)
SourceBaby viki, Sauce is at github.com/rakudo/rakudo/blob/843a...Mu.pm#L818
viki :(
lucasb_ how do I use SourceBaby? I want to search for the List.ACCEPTS method 13:51
viki s: (), 'ACCEPTS', \([])
SourceBaby viki, Sauce is at github.com/rakudo/rakudo/blob/843a...st.pm#L760
viki SourceBaby: help
SourceBaby viki, Use s: trigger with args to give to sourcery sub. e.g. s: Int, 'base'. See modules.perl6.org/dist/CoreHackers::Sourcery
lucasb_ viki: thanks
viki hm, I thought smartmatch against a List just matched by elements 13:53
by number of elements that is
m: say so (42, 42) ~~ (Match.new) 13:54
camelia rakudo-moar 054aca: OUTPUT«True␤»
viki m: say (42, 42, 42) ~~ (Match.new, Match.new) 13:55
camelia rakudo-moar 054aca: OUTPUT«False␤»
viki m: say (42) ~~ (Match.new, Match.new) 13:56
camelia rakudo-moar 054aca: OUTPUT«(「」 「」)␤»
viki m: say (42,) ~~ (Match.new, Match.new)
camelia rakudo-moar 054aca: OUTPUT«False␤»
viki AH
m: say so (42, 42) ~~ (Match.new,)
camelia rakudo-moar 054aca: OUTPUT«False␤»
psch m: say "foo" ~~ m:g/(.)/ 13:57
camelia rakudo-moar 054aca: OUTPUT«(「f」␤ 0 => 「f」 「o」␤ 0 => 「o」 「o」␤ 0 => 「o」)␤»
viki Ah
psch ^^^ that is also List on the RHS
viki m: say Match.^mro
camelia rakudo-moar 054aca: OUTPUT«((Match) (Capture) (Cool) (Any) (Mu))␤»
psch m: say ("foo" ~~ m:g/(.)/).WHAT
camelia rakudo-moar 054aca: OUTPUT«(List)␤»
viki m: say "foo" ~~ m:g/(.)/; say $/.^mro
camelia rakudo-moar 054aca: OUTPUT«(「f」␤ 0 => 「f」 「o」␤ 0 => 「o」 「o」␤ 0 => 「o」)␤((List) (Cool) (Any) (Mu))␤»
viki It also has some sort of magic for HyperWhatevers: github.com/rakudo/rakudo/blob/843a...#L773-L788 13:58
14:00 wamba joined
viki Here's the other prompt-hangs-on-windws ticket rt.perl.org/Ticket/Display.html?id...et-history 14:00
14:04 aborazmeh left 14:05 bjz_ left
viki gfldex: no, it doesn't. The, .elems gets called on () right away, and .elems gets called on the LHS array here, when `<` is called: github.com/rakudo/rakudo/blob/843a...st.pm#L772 14:09
`<` -> .Real -> .Numeric -> .elems
m: my $seq = (1 ... 1e100); say so $seq 14:11
camelia rakudo-moar 054aca: OUTPUT«True␤»
viki m: my $seq = (1 ... 1e100); say $seq ~~ () 14:12
camelia rakudo-moar 054aca: OUTPUT«(timeout)»
14:12 kyclark_ joined
viki m: my $seq = (1 ... 1e100); say $seq.elems == 0 14:12
:)
camelia rakudo-moar 054aca: OUTPUT«(timeout)»
14:13 kyclark_ left
gfldex this may provide lizmat with the reason to special case 14:13
14:16 Khisanth left
viki s/The, .elems gets called on () right away//; # that only happens when LHS is not an Iterable 14:21
# but that's not the part that makes the above timeout
14:23 kyclark_ joined, tyil left 14:24 kyclark_ left 14:27 kyclark_ joined, cdg joined 14:29 Khisanth joined 14:30 cdg left 14:31 cdg joined
moritz star-m: use JSON::Tiny; say to-json(42); 14:32
camelia star-m 2016.10: OUTPUT«42␤»
14:33 jfuller joined 14:35 jfuller_ joined, jfuller left 14:36 kyclark left, kyclark_ is now known as kyclark, jfuller_ left
viki star-m: use JSON::Fast; say to-json 42 14:37
camelia star-m 2016.10: OUTPUT«42␤»
viki star-m: use JSON::Fast; say from-json '42'
camelia star-m 2016.10: OUTPUT«a JSON string ought to be a list or an object␤ in sub from-json at /home/camelia/star-2016.10/share/perl6/site/sources/11ED005DCC03AA42CAA87A439556DB9B6B2C9E59 (JSON::Fast) line 305␤ in block <unit> at <tmp> line 1␤␤»
viki ehehe
I love that it generates what it thinks is invalid JSON :) 14:38
timotimo oh, right, it ought to allow just literals, right? 14:40
14:40 wamba left
timotimo i forgot what json now actually is and what it isn't 14:40
viki timotimo: yeah: github.com/timo/json_fast/issues/9
timotimo why didn't i fix that yet? 14:45
14:46 kyclark left 14:49 espadrine_ joined 14:51 nicq20 joined
nicq20 Hello o/ 14:51
viki \o
14:59 cdg_ joined
timotimo done, btw 14:59
dalek c: b800bd2 | (Zoffix Znet)++ | README.md:
Trigger doc rebuild
viki gfldex: should rebuild now. I recall someone updated perl6 on hack, but they never ran `rakudobrew rehash` to update the shims for executables, so the doc build was failing, because pod2bigpage wasn't found 15:00
15:00 pmurias left 15:01 cdg left
dalek c: 3c2d876 | coke++ | doc/Type/List.pod6:
fix typos
15:05
synopsebot6 Link: doc.perl6.org/type/List
c: 8f6f8d0 | coke++ | xt/code.pws:
learn new code
15:06 brillenfux left 15:07 jonas2 left
[Coke] viki: if there are more instructions to upgrading rakudobrew, please add them to "github.com/perl6/infrastructure-do...akudobrew" (I'm the one that just updated it, based on those directions) 15:08
apologies for any disruption. 15:09
viki adds `rakudobrew rehash` to ./upgrade-rakudobrew script 15:11
[Coke] viki++
given that the directions blow away .rakudobrew ... what does the rehash do on a fresh install? 15:12
viki You need to run it after installing any module that installs binaries
[Coke] ew. Ok.
viki :) 15:13
15:13 lucasb_ left
DrForr Aren't we pushing people that aren't core devs away from rakudobrew? 15:14
viki Yes.
DrForr (not saying that you aren't, just curious.)
[Coke] mst was, I think. I am a core dev, so it's hard to argue. :)
(coreish?)
Down to 13 gaps in the calendar. first gap is 8 days out. 15:15
15:15 pmurias joined
viki Is #1 spot now chizzeled in stone? 15:16
viki was hoping for some sort of State of The Butterfly recap for the year
huggable: advent
huggable viki, github.com/perl6/mu/blob/master/mi...6/schedule
15:17 espadrine_ left
viki nice to have bdfoy pitching in a post... 15:19
maybe worth mentioning to him that MAIN args get the same treatment as <...> he opened a ticket about a few days ago...
[Coke] viki: it's set to be published on Dec 1st; there's a paragraph or so about how it's been year, isn't that grand? 15:20
dalek : 3b2bf4e | moritz++ | misc/perl6advent-2016/schedule:
[p6advent]: Move my slot further to the front, talk about other Perl 6 books too
15:21
: eee14b7 | jnthn++ | misc/perl6advent-2016/schedule:
Update schedule
15:22
viki [Coke]: sounds good
jnthn finally thought up a topic in the shower this morning, then forgot to add it to the schedule :)
[Coke] Down to 12 gaps in the calendar. First gap is 9 days out. 15:23
viki: I'm still happy to switch if someone wants, though.
just need to change the schedule and make a slight update to the first paragraph.
note: since day 1 is already up there in preview form, please feel free to read it and give feedback. 15:24
ah, I did'nt actually hit schedule, only set the time. 15:25
I hit schedule.
and now the date in the URL looks right. 15:26
when you schedule wordpress, does it make the article available at the published URL immediately, but not link to it anywhere but the admin screen? 15:27
'cause I have a URL to share if someone wants to read it.
15:28 tyil joined
moritz [Coke]: it's only available to logged-in folks in that blog 15:28
with sufficient access (editor or administrator, iirc)
[Coke] so I can share it here without concern?
moritz yes
viki [Coke]: well, I'd love to do a recap of the year. I imagine there are some folks who weren't impressed with our "first production release" and maybe they decided to wait a bit.
[Coke]: but if it's already set to publish... *shrug* don't wanna be a bother :) 15:29
[Coke] viki: sure. This was intended to be one that could go anywhere, but I went first to fill the scary gap on the 1st.
perl6advent.wordpress.com/2016/12/...b-webhooks
15:29 khw joined
mspo a little early? 15:30
[Coke] mspo: read the last 5 minutes.
mspo got it
that url doesn't work for me, fwiw ;)
so I think you have to login 15:31
also, dogfood a blog
[Coke] mspo (doesn't work) excellent. :)
moritz [Coke]: "whipuptitude" might need an explanation (or a link) for the not so hardcore perl folks :-) 15:32
viki: it twould be nice if you could increase the text contrast a bit on the advent blog 15:34
viki agreed... though that would require raking for a new theme ('cause we can't make just small adjustments)
moritz :( 15:35
viki I'll do it tonight, when I'm at a location with faster Internet
moritz thanks, ++viki
15:40 kyclark joined 15:42 mad__ joined, mohae_ joined 15:45 mohae left, cooper_ left 15:47 wamba joined 15:48 espadrine_ joined
timotimo jnthn: i was in the shower today, too, but i didn't come up with a topic :( something must be broken in mine that works in yours 15:50
jnthn Maybe you have to sleep badly first also :P 15:51
timotimo i got that box ticked
AlexDaniel timotimo: how long have you been in the shower?
timotimo i find it surprising that people do their/they're but never yours/your's 15:52
AlexDaniel: hm, like 20 minutes?
15:52 mohae joined 15:54 CIAvash left
AlexDaniel timotimo: perhaps try more? You know, it takes some time for water to wear away rock 15:54
15:55 mohae_ left
timotimo i'm a bonehead, not a rockhead 15:56
15:56 kyclark left, skids joined 15:59 CIAvash joined 16:01 nowan left, andrzejku left, xinming left, ZzZombo left, solarbunny left, b2gills left, kaare_ left, beatdown left, dalek left, synopsebot6 left, emdashcomma left, amalia_ left, effbiai left, jferrero left, Undercover left, SourceBaby left, ugexe left, buggable left, Juerd left, parisba left, hobbs left, bhm left, ggherdov left, Kaffe left, apejens joined, hobbs joined, hobbs left, hobbs joined, bhm joined, effbiai joined, effbiai left, effbiai joined 16:02 nowan joined, hahainternet joined, andrzejku joined, emdashcomma joined, kaare_ joined, ZzZombo joined, amalia_ joined, jferrero joined
viki weeee 16:02
16:02 Juerd joined, beatdown joined, xinming joined, buggable joined, ugexe joined, ZzZombo is now known as Guest76904 16:03 b2gills joined
[Coke] moritz: got a good link for whipuptitude? 16:03
16:04 parisba joined
rindolf [Coke]: there's perl.org.il/presentations/larry-wal...cript.html 16:04
[Coke]: also www.shlomifish.org/lecture/Perl/New...slides.pdf 16:05
16:06 ggherdov joined
[Coke] Both of those (like most of the others) have a bunch of other stuff. I'll just put it inline. :) 16:06
16:06 Kaffe joined 16:07 eisen74 joined
rindolf [Coke]: and www.perl.com/pub/2007/12/06/soto-11.html 16:07
16:07 xinming left 16:09 mj41 left 16:13 mj41 joined 16:17 dalek joined, ChanServ sets mode: +v dalek, Undercover joined, ChanServ sets mode: +v Undercover, SourceBaby joined 16:18 xinming joined 16:19 eisen74 left 16:20 Actualeyes left 16:21 bazzaar joined
bazzaar o/ perl6 16:21
nicq20 \o
viki \o\
16:21 solarbunny joined
dalek c: 2fb08fe | coke++ | doc/Language/quoting.pod6:
Probably meant Pi here?

  briandfoy++
16:23
viki aye 16:24
bazzaar I'm having a problem installing rakudo-star-2016.11 on linux, panda install fails with File::Find error "Required named parameter 'name' not passed" 16:27
viki :o
m: dd q:w { [ ] { } } 16:29
camelia rakudo-moar 054aca: OUTPUT«("[", "]", "\{}")␤»
bazzaar anyone else reported that? maybe it's me being stupid
viki bazzaar: first time I hear it, but it got released just yesterday...
m: dd q:w / [ ] { } / 16:30
camelia rakudo-moar 054aca: OUTPUT«("[", "]", "\{", "}")␤»
viki I don't get what that { } variant is doing...
m: dd q:w [ [ ] { } ] 16:31
camelia rakudo-moar 054aca: OUTPUT«("[]", "\{", "}")␤»
viki m: dd q:w [ [ ] { } ]
camelia rakudo-moar 054aca: OUTPUT«("[]", "\{", "}")␤»
viki m: dd q:w [ [ lulzwat ] { } ]
camelia rakudo-moar 054aca: OUTPUT«("[lulzwat]", "\{", "}")␤»
nicq20 viki: Kind of odd it does not grab the '[]' chars the same way. 16:33
16:33 kyclark joined
dalek c: 37849ba | (Douglas L. Schrag)++ | doc/Language/quoting.pod6:
Fix example that mixed up q:w and Q:w

The original example had very strange behavior; not sure how to document this:
   Q:w { [ ] \{ \} } eqv ('[', ']', '{', '}') # False
   > Q:w { [ ] \{ \} }.perl
   ("[", "]", "\\\{\\}")
16:36
c: eed4ffb | (Zoffix Znet)++ | doc/Language/quoting.pod6:
Merge pull request #1022 from dmaestro/quoting_qw_fix_example

Fix example that mixed up q:w and Q:w
viki Feels like a bug 16:38
m: q:w “ 12 “ lulzwat ” 42 ” 16:41
camelia ( no output )
viki m: dd q:w “ 12 “ lulzwat ” 42 ” 16:42
camelia rakudo-moar 054aca: OUTPUT«("12", "“lulzwat”", "42")␤»
viki m: dd q:w “ 12 “ lulz wat ” 42 ”
camelia rakudo-moar 054aca: OUTPUT«("12", "“lulz", "wat”", "42")␤»
viki Rakudobugged: rt.perl.org/Ticket/Display.html?id=130205 16:48
BTW, the One Weird Trick to report tickets without having to email. is to append `m/` to the homepage URL to get mobile interface: rt.perl.org/m/ 16:49
It's pretty amusing it also lets you create a ticket in Spam queue :D 16:50
tony-o is there any need in the ecosystem for something like npmjs with private (a la company private) repositories? 16:51
16:51 domidumont left
viki has no idea what nqpjs is ... 16:51
... or npmjs 16:52
tony-o npmjs provides a service where you company can keep modules private to your company (or an arbitrary group) and makes them installable via npm. so it'd be equivalent to zef being able to search a 'private' ecosystem that only your company or friends or whoever has access to 16:53
viki so like stratopan.com 16:54
tony-o looks like that, yea
viki Yeah, it's handy. 16:55
(that was the answer to original Q) :)
tony-o is stratopan being maintained?
half of the links don't work 16:56
dalek c: 2e980b2 | gfldex++ | doc/ (2 files):
move index entry for empty list to List
16:57
c: ced7c9f | gfldex++ | doc/Language/operators.pod6:
link to empty list
viki no idea
But I remember it from years ago and I see it's still in "Beta" 16:58
tony-o the original intent of modules.zef.pm was to go in that direction. right now it's just a landing space for searching modules and browsing the code
all of the login and management functions are turned off 16:59
viki Heh. For a site purported to be for private repos, they sure aren't shy about disclosing all of their users :) stratopan.com/explore/users
tony-o yea i was just looking at that, weird
[Coke] gah. didn't I just fix both emtpy and absense !? 17:00
viki hehe
dalek c: 56fc629 | coke++ | doc/Type/List.pod6:
fix typos
17:02
[Coke] must have missed some instances, ah well.
17:03 CIAvash left
gfldex [Coke]: it's nice to know that at least one person is reading the gibberish i prodoce :) 17:04
typing around a big got cup of tea does not improve spelling
[Coke] gfldex: reading is too strong a term :) 17:11
17:12 kurahaupo joined
[Coke] (that is, I'm just making the tests pass) 17:14
17:15 lolo78 joined 17:17 kurahaupo left, kurahaupo joined 17:21 kyclark left, eisen74 joined 17:23 alimon1 joined
pmurias viki: nqpjs is the the nqp-compiling-code-to-js 17:26
17:26 kurahaupo__ joined
viki pmurias: nah, I mistyped npm-js 17:28
17:28 dakkar left
pmurias tony-o: I can assume the demand for having a 'private' ecosystem will increase with more people having a lot of 'private' modules 17:29
17:33 cibs left 17:35 cibs joined 17:36 chris2 left 17:37 bazzaar left 17:44 bazzaar joined
bazzaar regarding earlier post, on problem installing Rakudo Star 2016.11 (specifically installing panda) on linux, pls see : gist.github.com/bazzaar/bb3eed14f7...88535f68ae 17:48
17:50 chris2 joined
timotimo bazzaar: that seems like the invocation of the written-in-perl6 prove command is wrong, or something like that 17:52
um, actually
it's apparently using last month's prove6 binary for your this month's rakudo star
bazzaar timotimo: I didn't nuke my previous perl6 R* install, prior to trying the latest R* 17:54
17:56 eisen74 left 17:59 andrzejku_ joined 18:03 nicq20 left 18:04 mj41 left 18:06 labster joined 18:09 labster left, labster joined
tony-o pmurias: that's kind of what i figured, i know some folks are doing day to day stuff in p6 and was mostly curious if anyone was starting to develop things internally that might call for private ecos 18:09
[Coke] we have github enterprise here; I'd probably tie into that before a local private cpan. 18:10
18:17 committable6 left
bazzaar timotimo: thanks for your help, the install now works after I moved the previous perl6 to a backwater. :) 18:18
18:19 setty1 joined
tony-o [Coke]: using github enterprise would be fine for the backend - i see the way that tying in is through a list of p6 modules available in that github/org and being searchable/installable via zef or panda or whatever 18:21
18:24 eliasr left
AlexDaniel huggable: advent 18:25
huggable AlexDaniel, github.com/perl6/mu/blob/master/mi...6/schedule
AlexDaniel hmmmm 18:26
viki You look like you would like to write an Advent article! 18:28
18:31 wamba left
AlexDaniel are there any special dates? 18:32
viki What sort of special?
AlexDaniel or anyone can write about anything on any day?
viki Anything on any day
AlexDaniel okay, hmmm
viki AlexDaniel: oh, and I murdered committable... 18:33
AlexDaniel yea it's time to put them in a loop. They are now known to be consistently unstable 18:34
“Unhandled exception in code scheduled on thread 6” and stuff
18:35 committable6 joined, ChanServ sets mode: +v committable6 18:37 bazzaar left 18:40 domidumont joined
dalek : d233c03 | (Aleks-Daniel Jakimenko-Aleksejev)++ | misc/perl6advent-2016/schedule:
23rd

Conveniently before “Perl 6 Core Hacking”
18:47
viki Haha. That's a pretty depressing title :) 18:52
Perfection can only be persued, never attained, eh? :) 18:53
pursued too
jeek Perused?
18:56 thowe joined, ZzZombo joined 18:57 ZzZombo is now known as Guest96738, Guest76904 left
AlexDaniel it will be a post about interesting gems from the ticket history and about bug reporting in general 18:58
viki sweet
AlexDaniel viki: so I'm hoping that it will end up being very positive :)
viki AlexDaniel: did you see I found a way to create a new ticket without an email? 18:59
AlexDaniel viki: oh, no
how? Using some direct link?
viki rt.perl.org/m/
Using mobile version of the site 19:00
AlexDaniel oooooooooooooooooooooooooooooooooooooh
did not know it existed
viki I assuem the person who disabled the 'Create Ticket' button didn't either :)
AlexDaniel good. Very good 19:01
[Coke] I would assume that now tha tyou've posted it here, it'll get spam killed in the next 3 months.
DrForr Web dev Barbie says responsive design is *hard*!
AlexDaniel hmm does not seem to support html 19:02
viki :(
19:02 girafe joined
AlexDaniel but that's still better than nothing, I guess? 19:02
viki: thank you very much!
19:05 kyclark joined 19:12 kyclark left
timotimo bumped JSON::Fast and even uploaded a tag 19:14
19:14 domidumont left
AlexDaniel timotimo: hey, any progress on MoarVM panic thingy? 19:16
19:17 cdg_ left, cdg joined
timotimo no, sorry :( 19:18
19:19 darutoko left
AlexDaniel timotimo: is there any way we can create a meaningful ticket out of this problem? 19:20
19:22 cdg left
timotimo i'm not sure, it's very far from golfed, sadly 19:22
AlexDaniel timotimo: is there anything in particular you are stuck on, or is it just lack of time? 19:23
19:25 geekosaur left 19:26 RabidGravy joined 19:27 geekosaur joined
AlexDaniel I really want to make this thing move forward 19:28
timotimo sorry, it's just lack of time and energy 19:34
[Coke] Is there a ticket?
AlexDaniel [Coke]: no. Wanna try to golf it? 19:35
19:43 labster left 19:50 eisen74 joined, domidumont joined, mj41 joined
[Coke] Don't NEED a golfed version for a ticket. 19:52
need a ticket to fix it, though.
I can try to do a golf off a ticketed version.
19:53 ugjka joined
viki wonders what this ugjka is... 19:55
19:56 kyclark joined
viki Has been lurking in #zofbot for a few days and I've been getting weird /msg from random people asking how I am 19:56
Do I sound paranoid yet? :) 19:57
AlexDaniel viki: github.com/ugjka ? 19:58
19:58 kyclark left
AlexDaniel some irc-related projects 19:58
viki Yeah and no perl related projects :D 19:59
oh
github.com/zoffixznet/perl6-IRC-Cl...Y/issues/1
heh
I *am* paranoid
viki clears kickban from #zofbot
20:05 Tonik joined, kyclark joined 20:06 lukaramu joined
jnthn AlexDaniel: Is the panic in question the "zeroed target" one? 20:07
AlexDaniel jnthn: yes 20:08
jnthn OK...got a console tab with a GDB session where I managed to catch an example of that a few steps further back from where it actually blows up. Didn't get to the bottom of it. 20:09
I've got a private project that produces it about 1 in 10 times, but it shows up quickly if it's going to show up at all. 20:10
Unfortunately the error really just means "memory got corrupted and we've caught it in a slightly better way than SEGVing"
So hard to know if there's one root cause or multiple 20:11
Will continue hunting it later this week, anyways.
20:11 labster joined
AlexDaniel hmmm I should try something 20:11
jnthn (Most likely Wednesday) 20:12
AlexDaniel unicodable6: { say ‘hello’; False }
unicodable6 AlexDaniel, Oops, something went wrong! 20:13
AlexDaniel unicodable6: { say ‘hello’; False }
20:13 unicodable6 left
diakopter lol 20:13
dogbert17 jnthn timotimo: is ther a switch to force a GC as often as possible?
20:14 unicodable6 joined
AlexDaniel unicodable6: { say ‘hello’; False } 20:14
dogbert17 AlexDaniel: and viki managed to kill commitable
with this: committable6: 2015.07 say 1.0000001 ** (10 ** 8) 20:15
20:15 unicodable6 left
dogbert17 hmm, did I do that ? 20:15
jnthn dogbert17: You can tweak the GC thresholds in a header file
AlexDaniel no
20:16 unicodable6 joined
AlexDaniel unicodable6: { say ‘hello’; False } 20:16
unicodable6 AlexDaniel, Oops, something went wrong!
AlexDaniel unicodable6: { say ‘hello’; False }
unicodable6 AlexDaniel, Oops, something went wrong!
20:16 unicodable6 left
dogbert17 jnthn: you know which one? 20:16
jnthn collect.h iirc
If not, allocation.h
AlexDaniel ok, I thought I had an idea of what could be causing it in particular…
dogbert17 jnthn thx
20:17 unicodable6 joined
AlexDaniel but no, I guess that's not it 20:17
dogbert17 timotimo and I have been looking at a MoarVM panic a few days ago
AlexDaniel committable6: 2015.07 say 1.0000001 ** (10 ** 8)
20:17 committable6 left 20:18 domidumont left
AlexDaniel Unhandled exception in code scheduled on thread 15 bla-bla 20:18
20:18 committable6 joined, ChanServ sets mode: +v committable6
AlexDaniel let's try again 20:18
committable6: 2015.07 say 1.0000001 ** (10 ** 8)
diakopter dogbert17: lol, a collectable6 could run a rakudo with those thresholds at minimum
20:18 committable6 left
viki !!! 20:19
dogbert17 diakopter: collectable6, how many bots do we have :)
[Coke] why are people putting [BUG] on their subject in bug reports? (didn't I remove the hint we had that suggested it?) 20:20
viki
.oO( we had a hint? )
pmurias tony-o: isn't the big part of having a private custom cpan pinning down a set of working versions? 20:21
viki I put it in when I'm reasonably sure something is an actual bug
20:21 lolo78 left 20:22 committable6 joined, ChanServ sets mode: +v committable6, wamba joined
[Coke] viki: there were docs in the rakudo repo, yes. 20:22
20:23 lolo78 joined
viki And vis-avis RT#130207, we *do* have /latest 20:24
All 4 icons link to latest*whatever rakudo.org/how-to-get-rakudo/
[Coke] sure, but that's not what he's asking for. 20:25
viki It is, in a way.
[Coke] (which is, I think, a latest in the downloads directory.)
viki You can automate anything you want with the current latest links, they're just done with .htaccess rather than with a symlink 20:26
Like rakudo.org/downloads/star/rakudo-st...est.tar.gz is a working link 20:27
hm, he wants a rakudo one
viki tries something
20:29 lolo78_ joined 20:30 TEttinger joined, lolo78 left, lolo78_ is now known as lolo78
viki Hm. If I add a symlink, it does show up in the folder, rakudo.org/downloads/rakudo/rakudo-...st.tar.gz, but when you click it, the file name that's downloaded is named just that, but I want it to be named with the version. 20:31
lizmat vaguely remembers there was a HTTP header for that 20:32
Content-Name ?
viki Yeah, there is Content-Disposition IIRC
Oh, wait, yeah, disposition affects whether it's displayed inline. 20:33
jonadab Yes, that works for MIME attachments as well.
[Coke] <cfheader name="Content-Disposition" value='attachment;filename="#Replace(file_name,",","")#"'>
(there's some coldfusion for you that works. :)
viki While, we're at it, maybe it's worth changing the look of downloads pages? 20:34
maybe not :D 20:35
viki `Take`s that ticket. 20:36
Will think about it on the way home.
20:40 Vynce joined 20:42 mattp_ joined, mattp__ left
viki a symlink so something shows up in the directory itself, but use the current mod_rewrite solution to review it to proper name.... and a script to automate updating this stuff 20:44
s/review/rewrite/;
a perl script with a fake extention that dynamically looks up what's the latest version and serves it (no need for any updating scripts) 20:50
??? 20:51
Profit!
viki relocates
AlexDaniel huggable: test
huggable AlexDaniel, Test failed
AlexDaniel huh
ok
huggable: foo
huggable AlexDaniel, nothing found
AlexDaniel huggable: foo :is: bar 20:52
huggable AlexDaniel, Added foo as bar
AlexDaniel huggable: foo
huggable AlexDaniel, bar
AlexDaniel great
are there any other public logs of this channel besides irclog.perlgeek.de ? 20:53
lizmat AlexDaniel: not that I know of 20:54
lizmat starts working on the P6W
AlexDaniel huggable: new ticket :is: rt.perl.org/m/ticket/create?Queue=16
huggable AlexDaniel, Added new ticket as rt.perl.org/m/ticket/create?Queue=16
[Coke] AlexDaniel: I would probably not advertise that link. 20:55
20:55 bjz_ joined
AlexDaniel [Coke]: why not? 20:55
[Coke] let me rephrase that. I would recommend not advertising that link. 20:56
because the rt admin team deliberately removed the one they knew about.
AlexDaniel and?
[Coke] Seems like the friendly thing for us to do would be to let them know they missed one, instead of going behind their back, as it were. 20:57
AlexDaniel they had a spam issue. If there is no spam issue with m/ link, there is no problem
timotimo dogbert17: you can also nqp::force_gc()
AlexDaniel if there is a problem, then they will probably figure it out themselves
lizmat AlexDaniel: you probably just created the spam issue, long term :-( 20:58
[Coke] also, the lines with the inverted A's are not being logged on irclog.perlgeek.de... bug or feature?
AlexDaniel lizmat: this link does not appear in clog
dogbert17 timitimo: thx
s/timitimo/timotimo/
AlexDaniel and if anybody decides to share that link again, ask huggable instead of pasting it
[Coke] AlexDaniel: everyone involved in rt.perl.org is a volunteer. we should be working with them, not against them. 20:59
I'm going to mention the link to the admins, in case they are not aware.
AlexDaniel /o\ 21:00
I mean, feel free to mention it, sure. But why are you pretending that there is an issue with that link when there is actually none at the moment
21:02 seatek joined
AlexDaniel one good thing we can do is ask moritz to delete the mentions of /m/ link 21:02
viki AlexDaniel, nice going, lol :)
I should've kept that link my private secret :) 21:03
AlexDaniel XD
mst AlexDaniel: ... *dude*
AlexDaniel what?
mst publishing a URL that will help spammers circumvent a security measure is *not* constructive
viki [Coke], ask them what sort of help is needed to restore the create ticket button. I can write code 21:04
AlexDaniel spammers can just spam the email address
which is what they do constantly
mst yeah, which has a filtering system at the *email* level 21:05
AlexDaniel so what is the actual problem with the button? Where are the details?
[Coke] mst was, as I recall, working on a potential solution with the admins. doing an end run around them doesn't make us look good.
AlexDaniel don't you have to create an account to use that button?
[Coke] AlexDaniel: the original button was abused to create a spam storm on teh p5 queue, and was disabled. 21:06
mst AlexDaniel emailed them. I've completely failed to chase it up yet. but I do have a plan when I remember.
AlexDaniel: the problem is that the anti-spam measures are implemented in the mail submission system. which create ticket doesn't go via. hence 'circumvent'.
21:08 obfusk_ left
AlexDaniel and there are no anti-spam measures for creating an account? 21:08
timotimo dogbert17: i think the bug you recently posted about the wrong kind of free was the thing we (you? someone?) just fixed 21:09
AlexDaniel was it a manual attack or we just got unlucky?
viki Clearly someone from this channel needs to join the RT admin group :) 21:10
geekosaur saw that but also saw it had been in holding for 6 hours before being remailed so figured it got addressed before anyone else saw it
mst I don't know exactly what happened
if you could remind me to follow it up (harass, even, I'm ok with this when I'm failing to get to something) instead of making it into a bigger trainwreck maybe we'd find out :) 21:11
AlexDaniel .in 2d harass mst
yoleaux AlexDaniel: I'll remind you on 30 Nov 2016 21:11Z
dogbert17 timotimo: yes, I actually managed to run 'make spectest' with FSA_SIZE_DEBUG = 1 having applied the fix
thowe waves to mst
AlexDaniel .in 4d harass mst
yoleaux AlexDaniel: I'll remind you on 2 Dec 2016 21:11Z
AlexDaniel .in 6d harass mst 21:12
yoleaux AlexDaniel: I'll remind you on 4 Dec 2016 21:12Z
AlexDaniel .in 8d harass mst
yoleaux AlexDaniel: I'll remind you on 6 Dec 2016 21:12Z
viki :/
mst that's totally reasonable.
viki \o/
mst notice he's set the bot to spam -him- so if I get to it sooner he gets spammed and not me
viki :D 21:13
AlexDaniel that's ok
mst AlexDaniel++ # most people don't take me seriously when I say I'm ok with being harassed when I keep forgetting something like this. your response was actually less hassle for me.
21:19 mj41 left 21:23 Vynce left 21:28 kyclark left, FROGGS joined 21:34 bjz_ left 21:42 cdg joined 21:50 labster left 21:54 labster joined 22:00 RabidGravy left
seatek weird - i've got an error message: method 'blah' must be implemented by My::Role because it is required by a role. But My::Role is the role defining that abstract requirement (and itself does no other role nor inherits) 22:01
[Coke] seatek: code?
seatek i don't think it will be of much use -- it's a role that wasn't changed that just started showing that error after i've been modifying in class in the same file (which does that role) 22:04
viki You're punning the role? 22:05
m: role Foo { method meows { … } }.new 22:06
camelia rakudo-moar 054aca: OUTPUT«Method 'meows' must be implemented by Foo because it is required by a role␤ in any compose_method_table at gen/moar/Metamodel.nqp line 2824␤ in any apply at gen/moar/Metamodel.nqp line 2834␤ in any compose at gen/moar/Metamodel.nqp line 3006␤ i…»
22:06 pecastro left
seatek the role defined 4 normal attributes, and defined 4 abstract method requirements 22:06
but it's saying that it must implement those method requirements now itself
viki I don't believe you
Prove it :)
seatek hehee :)
that will not be easy - i'll try to think of a way 22:07
[Coke] ... you could show the code. :) 22:08
22:09 bjz joined, FROGGS left
seatek gist.github.com/adaptiveoptics/6c5...9c282f9214 22:10
Error while compiling 22:11
Method 'connect' must be implemented by HG::Antenna because it is required by a role
and the line is that role def line
22:12 skids left
AlexDaniel e: gist.githubusercontent.com/adaptiv...520problem 22:13
eval: gist.githubusercontent.com/adaptiv...520problem
evalable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, rakudo-moar 4dffef7: OUTPUT«(exit code 1) ===SORRY!===␤Type 'HG::Antenna::Receiver' is not declared␤at /tmp/iaOC2LObeb:3␤-…»
AlexDaniel, Full output: gist.github.com/a5438c84dcb1016c73...f7c5e43df6
AlexDaniel ah, right
22:14 andrzejku_ left
AlexDaniel cannot read even a single line before actually running the code 22:14
eval: gist.githubusercontent.com/AlexDan...tfile1.txt 22:16
evalable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, rakudo-moar 4dffef7: OUTPUT«»
AlexDaniel seatek: so what is required to reproduce it? 22:17
perlpilot seatek: smells like an accidental pun to me.
22:17 rindolf left
seatek what is a pun? 22:17
viki seatek: a role used as a class
seatek ah .. yeah, none of that :)
viki (which is legal and is called punning)
seatek just classes that does that role ;) 22:18
perlpilot seatek: show *all* the code :-)
seatek it's about 4,000 lines now
AlexDaniel that's better than 15 lines that don't reproduce the issue 22:19
22:19 pierre_ joined
viki seatek: maybe you wrote `is` instead of `does` at some place 22:20
22:20 kyclark joined
viki m: role Foo { method connect { ... } }; class Bar is Foo {} 22:20
perlpilot yeah, that's easy enough to do
camelia rakudo-moar 054aca: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Method 'connect' must be implemented by Foo because it is required by a role␤at <tmp>:1␤»
seatek i'll try to determine what's causing it and will let you know. it's just been a couple method additions in classes in that file that do that role
i'll double check
22:21 pecastro joined
viki m: role Foo { COMPOSE { BEGIN say callframe(0).file }; method connect { ... } }; class Bar is Foo {} 22:21
camelia rakudo-moar 054aca: OUTPUT«<tmp>␤5===SORRY!5=== Error while compiling <tmp>␤Method 'connect' must be implemented by Foo because it is required by a role␤at <tmp>:1␤»
viki Maybe see if that ^ incantation gives you the file with the issue? (maybe it won't)
22:21 bjz left
viki there's also .line method that gives the line number 22:21
(that probably gives some bogus file tho :P) 22:22
seatek yeah no mistaken is instead of does -- though one class IS a class of another class that DOES... ;) thought that might be it, commented it out, but it wasn't 22:24
sometimes i really hate oop
perlpilot seatek: I bet if you show the actual code someone here could find the problem
22:25 pierre_ left, zilti left
AlexDaniel for sure 22:27
seatek figured it out. sortof 22:28
it's those damn multi-line comments that i love
they screw up sometimes
viki Oh, and I thought of my plan for the RT#130207 while on the bus: add a script that generates those download pages which will be just a purtier version of the current page and the top it'll list the "latest" links in more obvious way. So we get prettier pages and no need to update any links after release 22:29
AlexDaniel seatek: are you sure that there is nothing we can do in this case?
seatek: e.g. perhaps some kind of a warning could be issued or something?
seatek multiline comments were somehow "hiding" an error about a stupid variable not having a my declared on it
viki seatek: get an editor that supports commenting. Select &_Ctrl+D in atom does all the commenting for me
seatek totally unrelated to the role
viki (I forget what the default shortcut is tho)
seatek no, the syntax was right -- i checked and double checked 22:30
viki Ah
seatek and teh syntax highlighting shows me too
TimToady
.oO(We must...be...protocoligorically correct.)
seatek i'm going to "undo" back to the comments being there so i can try to find anything tha tmight stand out 22:33
nothing wrong with the commenting. checked and checked. all bounded by #|{ } 22:35
lizmat and another Perl6 Weekly hits the net: p6weekly.wordpress.com/2016/11/28/...ing-along/ 22:37
seatek here is the comment that causes the problem. if i remove it, then i'm not told i must implement connect() in that role: gist.github.com/adaptiveoptics/6c5...9c282f9214
crazy! 22:38
but i've seen other weirdness with multiline comments
you can verify this even by using just that code in the file (if you remove the class types so that it doesn't error) 22:43
i just tried it
it's some kind of magic combination that results in lies and death
TimToady does it still error if you change #|{} to #'{} ? 22:44
er, #`{}
if not, it's somethign to do with attaching semantic comments
22:45 alimon1 left
TimToady maybe the semantic comment attacher is accidentally punning? 22:45
22:45 alimon1 joined
seatek with that i get "unable to parse role definition 22:46
expecting any of generic role
oh wait 22:47
no that works! :)
22:47 Actualeyes joined
TimToady are you wanting semantic comments, or just using #|{} by accident? 22:47
22:47 alimon1 left
seatek using #`{ } makes it work 22:48
i like using #|{} because it creates docs
TimToady right, and it probably should work, but maybe it's not generic-proof yet
seatek if #`{} is safer, i'll take that over auto-generated docs any day 22:49
22:49 erblite joined 22:50 _28_ria left 22:51 chinarulezzz joined, _28_ria joined 22:53 firstdayonthejob left
TimToady m: #|{ a Fooish thingy }; role Fooish { method fooey {...} } 22:53
camelia rakudo-moar 054aca: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Method 'fooey' must be implemented by Fooish because it is required by a role␤at <tmp>:1␤»
TimToady m: #`{ a Fooish thingy }; role Fooish { method fooey {...} }
camelia ( no output )
TimToady yeah, it's role punning somehow by accident, seems
viki "o
:o
seatek i step on all the landmines always 22:54
i love that people were thinking punning from the get-go. they were right, but it just wasn't me doing it 22:57
well, not directly
22:58 vendethiel left
seatek is punning a real thing? or is it like golfing? well, i guess golfing is a real thing now... ? 22:58
viki lizmat++ good weekly
Yes, it's a real thing. 22:59
22:59 trnh joined, trnh left, trnh joined
lizmat m: role A {}; dd A.new # punning 22:59
camelia rakudo-moar 054aca: OUTPUT«A.new␤»
TimToady m: role Fooish { method fooey {...} }.WHY.say 23:00
camelia rakudo-moar 054aca: OUTPUT«Method 'fooey' must be implemented by Fooish because it is required by a role␤ in any compose_method_table at gen/moar/Metamodel.nqp line 2824␤ in any apply at gen/moar/Metamodel.nqp line 2834␤ in any compose at gen/moar/Metamodel.nqp line 3006␤…»
viki ohhh
seatek k :)
viki ohhhh
TimToady m: role Fooish #={a Fooish thingy}␤ { method fooey {...} }
camelia ( no output ) 23:01
viki I even have a good guess of what's 'causing it
I merged a commit that made the pod parser care about .WHY
recently
bisect: role Fooish { method fooey {...} }.WHY.say
bisectable6 viki, Bisecting by output (old=2015.12 new=5163e8a) because on both starting points the exit code is 1
viki, bisect log: gist.github.com/88d4ed08a3a0085a6c...96a3ab46de
viki, (2016-03-02) github.com/rakudo/rakudo/commit/fd...7e379b5038
viki oh, I guess it ain't it :)
TimToady bisect: #|{ a Fooish thingy }; role Fooish { method fooey {...} } 23:02
bisectable6 TimToady, On both starting points (old=2015.12 new=5163e8a) the exit code is 1 and the output is identical as well
TimToady, Output on both points: ===SORRY!=== Error while compiling /tmp/ZxO1QFG55c␤Method 'fooey' must be implemented by Fooish because it is required by a role␤at /tmp/ZxO1QFG55c:1
23:04 trnh left 23:05 trnh joined, trnh left, trnh joined 23:07 lolo78 left 23:12 erblite left
viki seatek: don't forget to report that bug, please :) 23:12
23:13 lolo78 joined
seatek that's via email to something, yeah? i've never created a bug here. the whole thought of it make me feel guilty. 23:14
viki yeah, just email to [email@hidden.address] don't worry about feeling guilty... Tell 'em I sent you ;)
And you can include the link to the chat: irclog.perlgeek.de/perl6/2016-11-28#i_13644451
seatek k 23:15
23:16 labster left 23:19 lolo78_ joined, lolo78 left, eisen74 left, lolo78_ is now known as lolo78
seatek viki: done 23:20
viki seatek++ thanks a lot
TimToady seatek++
jeek (SETEC Astronomy)++ 23:32
23:33 pmurias left 23:37 labster joined 23:52 lolo78 left
mspo does moar/rakudo compile everything to NQP as a first step? 23:57
TimToady no, there is no intermediate language, just trees 23:58
seatek what's really surprised me as i've been going along is just how handy multi methods can be. they've saved me a lot of time, not having to re-write stuff, and helps keep the code tidy. never thought i'd have a use for them. 23:59
23:59 lukaramu left