»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
atroxaper Hello, #perl6 ! 04:07
yoleaux 10 Jul 2014 21:09Z <Mouq> atroxaper: Do you expect Pod::SAX to supercede the current Pod::To::HTML? I ask simply because I don't want to be doing work on something else when someone is working on a better solution. Right now my focus with Pod::To::HTML is readibility and encapsulation/thread safety. We should co-ordinate efforts :)
atroxaper .tell Mouq Actually I hope my solution would be better than current Pod::To::HTML. I would like you to look at my Pod::SAX and current Pod::SAX::Goes::HTML at github.com/atroxaper/Pod-SAX 04:13
yoleaux atroxaper: I'll pass your message to Mouq.
sergot o/ 06:47
atroxaper sergot: o/ 07:00
sergot m: "a".encode('iso-8859-1'); 07:23
camelia ( no output )
sergot m: "a".encode('iso-8859-2');
camelia rakudo-moar 2c957f: OUTPUT«Unknown string encoding: 'iso-8859-2'␤ in method encode at src/gen/m-CORE.setting:6528␤ in block at /tmp/5p7E_sqsej:1␤␤»
sergot Why do we not have iso-8859-2? :) 07:24
zengargoyle #perl6 should start taking on www.reddit.com/r/dailyprogrammer/ challenges... 08:31
masak_grr good antenoon, #perl6 09:09
atroxaper masak_grr: hello! 09:10
FROGGS hello Guest29091, welcome to #perl6 :P 09:11
masak_grr thank you. so, when will Perl6 be released? 09:12
atroxaper masak_grr: xDD
nwc10 when it's ready. And not before, in case people get food poisoning.
masak_grr I mean, when is it good enough so that people like me who just do drive-by questions on a channel and then never show up again would want to use it?
no, I mean, when is it ready for *production*? that rigorously defined term that's the same for everybody and that I obviously don't need to define any further? 09:13
when are you going to make an actual release?
jack_rabbit whenever you deem it stable enough to use in production. 09:17
I use it sometimes. It's fun.
Let your business people do their "risk management and mitigation" 09:18
zengargoyle a.k.a. "get rid of that crackpot!" 09:25
masak_grr jack_rabbit: nonono, I don't mean a subjective notion of "ready for production". I mean the obvious *objective* notion. you know, when it's ready. 09:27
masak_grr the point (if it ever comes) where the Perl6 peeps step back and say "OK, we're done. it's ready now" -- and they are *right*, for once. 09:27
I've been waiting for 27 years for this. in the meantime, I use Ruby, which is practically Perl 6 anyway. 09:28
nwc10 apart from the scoping rules? 09:29
masak_grr Ruby operates under the principle of "least surprise", which means that its scoping rules are, by definition, perfect. 09:30
masak ok. _grr suffix gone. I'll stop trolling now. ;) 09:31
jack_rabbit I guess I'd defer to nwc10's "when it's ready. And not before, in case people get food poisoning." 09:34
masak (my actual views on the matter are the same as always. there's important work to do, and jnthn++ FROGGS++ timotimo++ moritz++ sergot++ japhb++ colomon++ [Coke]++ lizmat++ woolfy++ others++ for doing it. much remains. when we get macros and slangs in place, we'll have something really nice.)
I'm sure I missed more than I included. TimToady++, for example.
tadzik++
brrt++ 09:35
atroxaper If my code has 'say @arr.perl' then all works well. If 'say @arr' or without say, then code 139 on Rukudo Parrot...
nwc10 r: my @arr = <like this>; say @arr 09:36
camelia rakudo-jvm 2c957f: OUTPUT«(timeout)»
..rakudo-{parrot,moar} 2c957f: OUTPUT«like this␤»
atroxaper No. I mean in case of my code) 09:37
nwc10 mmm, you've cut it down as far as it will go?
r: 1
camelia ( no output )
nwc10 r: say "Hi"
camelia rakudo-{parrot,jvm,moar} 2c957f: OUTPUT«Hi␤»
nwc10 r: my @arr = <like this>;
camelia ( no output )
nwc10 r: my @arr = <like this>; say @arr 09:38
camelia rakudo-{parrot,jvm,moar} 2c957f: OUTPUT«like this␤»
nwc10 JVM is now warm?
jack_rabbit :) looks like it. 09:40
it's amazing how much of a difference a warm VM makes. 09:41
lizmat
.oO( happiness is a warm VM )
 09:53
jack_rabbit :) 09:56
dalek kudo/nom: 8a0a8f9 | (Elizabeth Mattijsen)++ | src/core/Supply.pm:
Seems removing the paused functionality regresses

Specifically the classify/categorize tests. Guess I need to fix those first before removing this functionality. On the other hand, I don't really see how I can do that just yet. So reverting.
This reverts commit 592a2edf4c4bf6196a73715382a95999ebeafced.		 10:11
zengargoyle can you get a junction to unjunction? like getting: max 1, 2|3, 2; to be 3 and not any(2,3) 10:21
woolfy1 never did anything for Perl 6 other than marketing and some money, I never uploaded a patch or whatever... 10:22
woolfy1 wrote a couple of lines of code of Perl 6 Rakudo... but that was at the NQP-course by jnthn in Frankfurt last year...
masak++ for trolling though 10:23
zengargoyle or turning [1|2, 2] into [1,2],[2,2] 10:25
timotimo zengargoyle: if you want that, you shouldn't be using junctions 10:27
zengargoyle was afraid of that. :( trying to do Ace => 1|11 10:28
lizmat technically, it would be trivial to create a method that would return all the constituent values of a Junction 10:31
timotimo yes
dalek kudo/nom: 67ad387 | (Andrew Egeler)++ | src/core/SupplyOperations.pm:
Attempt to fix SupplyOperation wrongness (take 2)

Each source_tap now only passes messages on to that one client tap, not all of them.
kudo/nom: 87dcd11 | lizmat++ | src/core/SupplyOperations.pm:
Merge pull request #287 from retupmoca/nom

Attempt to fix SupplyOperation wrongness (take 2)
timotimo zengargoyle: what does your use case look like in a more big-picture way?
zengargoyle www.reddit.com/r/dailyprogrammer/co...k_checker/ 10:33
Alice => 1|11, 10
masak woolfy1: I won't let you undersell your contribution just because it's on the community side of things. woolfy++! 10:34
zengargoyle Bob => 3, 6, 7 10:34
max == 21
blah, blah, blah 10:35
find the max hand <=21, then find hands that have total==max to find winners.
timotimo i don't know if this explanation will help you, but basically passing a junction as a parameter to any sub or method that is restricted Any (or more specific) will cause that sub or method to be called once for each value in the junction and the result values will be collected back into a junction 10:37
masak (and Any is the default for parameters) 10:52
timotimo that's right. except for pointy blocks iirc 11:09
m: (-> $a { }).signature
camelia ( no output )
timotimo m: (-> $a { }).signature.say
camelia rakudo-moar 2c957f: OUTPUT«:($a)␤»
timotimo m: (-> $a { }).signature.perl.say
camelia rakudo-moar 2c957f: OUTPUT«:($a)␤»
timotimo m: (-> $a { }).signature.parameters[0].perl.say
camelia rakudo-moar 2c957f: OUTPUT«No such method 'parameters' for invocant of type 'Signature'␤ in block at /tmp/SXmZKvtDE0:1␤␤»
timotimo m: (-> $a { }).signature.arguments[0].perl.say
camelia rakudo-moar 2c957f: OUTPUT«No such method 'arguments' for invocant of type 'Signature'␤ in block at /tmp/11_NFrcR92:1␤␤»
timotimo m: (-> $a { }).signature.^methods.say 11:10
camelia rakudo-moar 2c957f: OUTPUT«<anon> <anon> <anon> <anon> arity count params returns ACCEPTS perl␤»
timotimo m: (-> $a { }).signature.params[0].perl.say
camelia rakudo-moar 2c957f: OUTPUT«$a␤»
timotimo m: (-> $a { }).signature.params[0].^methods.say
camelia rakudo-moar 2c957f: OUTPUT«<anon> <anon> <anon> <anon> <anon> <anon> <anon> name constraint_list constraints type named named_names positional slurpy optional parcel capture rw copy readonly invocant default type_captures sub_signature ACCEPTS perl␤»
timotimo m: (-> $a { }).signature.params[0].constraint_list.perl.say 11:11
camelia rakudo-moar 2c957f: OUTPUT«()␤»
timotimo m: (-> $a { }).signature.params[0].constraints.perl.say 11:11
camelia rakudo-moar 2c957f: OUTPUT«all()␤»
timotimo ...
how do i even ...
m: (-> Any $a { }).signature.perl.say
camelia rakudo-moar 2c957f: OUTPUT«:(Any $a)␤»
timotimo m: (-> Mu $a { }).signature.perl.say
camelia rakudo-moar 2c957f: OUTPUT«:($a)␤» 11:12
timotimo i suppose there we go.
m: (-> $a { say "executed" }(1|2))
camelia rakudo-moar 2c957f: OUTPUT«executed␤»
timotimo m: (-> Any $a { say "executed" }(1|2))
camelia rakudo-moar 2c957f: OUTPUT«executed␤executed␤»
timotimo m: (*.say)(1|2)
camelia rakudo-moar 2c957f: OUTPUT«any(1, 2)␤»
timotimo WhateverCode takes Mu, too
hm, actually ... 11:13
m: (* + 5)(1|2).say
camelia rakudo-moar 2c957f: OUTPUT«any(6, 7)␤»
timotimo m: (* + 5).signature.say
camelia ( no output )
timotimo m: my Mu &c = (* + 5); &c.signature.say 11:14
camelia rakudo-moar 2c957f: OUTPUT«Type check failed in assignment to '&c'; expected 'Callable[Mu]' but got 'WhateverCode'␤ in block at /tmp/dLIEGk7q30:1␤␤»
timotimo m: my &c = (* + 5); &c.signature.say
camelia rakudo-moar 2c957f: OUTPUT«:($whatevercode_arg_1 is parcel)␤»
timotimo that's Mu, too
masak if I want to augment a class and *override* a method, how do I avoid the error message "Package 'X' already has a Method 'y' (did you mean to declare a multi-method?)" ? 11:17
Ven masak: does :P ?
m: class A { method foo { 1 } }; A does role { method foo { 2 } }; say A.new.foo
timotimo masak: i think you can put a new proto in
camelia rakudo-moar 2c957f: OUTPUT«Cannot use 'does' operator with a type object.␤ in sub infix:<does> at src/gen/m-CORE.setting:16985␤ in block at /tmp/WnesD667Q4:1␤␤»
Ven :(
zengargoyle isn't there some sort of 'supercedes' thing? 11:18
masak zengargoyle: yes, but that replaces the whole class. I only want to replace a method. 11:19
suggestion: 'augment' complaining about this is a bug. (it wasn't always thus, and it was nicer when the error didn't happen)
Ven you want to monkey patch the class, basically.
masak Ven: yes, that's what 'augment' and MONKEY_TYPING are for. 11:19
Ven can you not use augment w/o monkey typing ?
Ven I guess 11:19
timotimo i don't think you can - or ought to be allowed to 11:20
zengargoyle can you just wrap the method?
masak no, you can't.
timotimo replacing a class in a lexical context is another thing, though
Ven m: augment Str {} # testing error msg
camelia rakudo-moar 2c957f: OUTPUT«===SORRY!===␤Type 'Str' is not declared. Did you mean 'str'?␤at /tmp/p4HHtCKZRW:1␤------> augment Str ⏏{} # testing error msg␤Malformed augment␤at /tmp/p4HHtCKZRW:1␤------> augment Str ⏏{}…»
masak zengargoyle: interesting suggestion -- I'll try.
Ven oh?
timotimo oh, whoopsie
the parser runs into a fail with Str and assumes the fail was due to Str not existing ... 11:21
is what i think it did
Ven m: augment class Str {} # testing error msg
camelia rakudo-moar 2c957f: OUTPUT«===SORRY!=== Error while compiling /tmp/I_esLQB7U1␤augment not allowed without 'use MONKEY_TYPING'␤at /tmp/I_esLQB7U1:1␤------> augment class Str ⏏{} # testing error msg␤ expecting any of:␤ scoped …»
Ven timotimo: think so
timotimo er ... huh?
Ven huh huh ?
timotimo why is the error message so different if there's a "class" in the middle? 11:22
Ven because of what you said ?
timotimo oh?
but should it even try to match a <typo_typename> there?
Ven w/o class the parser fails to parse it and thinks there's no such class, when really it just couldn't parse w/o the class ?
the "class" (literal, I mean) 11:23
jnthn masak: If we supported it, you would put supercede on the method. 11:23
yoleaux 10 Jul 2014 23:38Z <retupmoca> jnthn: I tried to kill the SupplyOperation bug again: github.com/rakudo/rakudo/pull/287
timotimo well, the way this works is there's a token "typo_typename" that matches something that looks like a typename and if it matches suggests things that have similar names
it's put after a || that usually has a <typename> on the lhs
jnthn masak: It's not a bug that it complains about redefinition inside of an augment.
timotimo jnthn: my suggestion with putting a new proto into the role is wrong, then? 11:24
masak jnthn: I'd be fine with putting supersede (sic) on the method.
jnthn Well, if I understand what masak wants, it's to augment a class and *replace* a method
masak *nod* 11:25
jnthn masak: bah, I hate that word because I can never remember how to spell it. :/
masak it's spelled wrong, period :P
what's that trait that makes a candidate win?
like !important in CSS
jnthn is default
Ven is important :P ?
masak thanks
using that made things work. 11:26
Ven my $masak is important
timotimo oh? is default? seems useful, i should remember that
masak but of course using that is also *WRONG*
jnthn "Supercede has occurred as a spelling variant of supersede since the 17th century, and it is common in current published writing."
masak I mean, beyond the fact that MONKEY_TYPING and augment is suspect, too.
jnthn Bah. :)
timotimo right; because there isn't really a second candidate that you want to have in the candidate list ...
Ven is it actually supersede or supercede ?
masak Ven: supersede 11:27
timotimo how does "is default" work, semantically? is it there to have a winner if two candidates have identical signatures?
Ven There Is More Than One Way To Spell It.
masak timotimo: something like that
Ven but what if ... they both are default ?
timotimo i wish "something like that" would work in programming :P
Ven m: sub a { 5 }; sub a is default { 6} ; say a;
camelia rakudo-moar 2c957f: OUTPUT«===SORRY!=== Error while compiling /tmp/Qiqs_MM2Ld␤Redeclaration of routine a␤at /tmp/Qiqs_MM2Ld:1␤------> sub a { 5 }; sub a is default { 6}⏏ ; say a;␤ expecting any of:␤ postfix␤ stateme…» 11:28
timotimo though i suppose that's how we got to PHP ...
masak timotimo: see S12:1221
synopsebot Link: perlcabal.org/syn/S12.html#line_1221
Ven m: multi a { 5 }; multi a is default { 6} ; say a;
camelia rakudo-moar 2c957f: OUTPUT«6␤»
Ven m: multi a is default { 5 }; multi a is default { 6} ; say a;
camelia rakudo-moar 2c957f: OUTPUT«Ambiguous call to 'a'; these signatures all match:␤:()␤:()␤ in sub a at /tmp/rn2aZM0_jF:1␤ in block at /tmp/rn2aZM0_jF:1␤␤»
jnthn Ven: en.wiktionary.org/wiki/supercede has some info on the history of supercede vs supersede
Ven Should the "is default" be noted in the signature ? 11:29
jnthn No, it's nothing to do with the signature, it's about the sub.
Ven True. 11:30
the same that a multi isn't seen in the signature
jnthn Implementation wise, "is default" is handled right at the end of the multi-dispatch logic. It says "oh dear, I have more than one candidate", and before it reports an error it checks if one of them is marked "is default" 11:31
Ven (and only one of them :P)
jnthn I'm not sure it takes that much care :)
Ven jnthn: if it didn't, it'd probably just take the first is default variant, which it doesn't 11:32
well, I have absolutely no clue, soo :)
jnthn oh, it does care :) 11:33
github.com/rakudo/rakudo/blob/nom/....nqp#L2079
heh, well, sort of :) 11:34
A well-placed "last" in there would probably make it more robust :)
masak +1 11:36
(and a spectest)
dalek osystem: e4c678a | masak++ | META.list:
add Data::Pretty
 11:54
masak ladies and gentlebots.
I present you with Data::Pretty.
github.com/masak/data-pretty
masak by the way, there are two failing tests right now in the module. I don't feel savvy enough to fix them. it's probably not that difficult, but help from someone who knows Rakudo/nqp internals better than I do would be greatly appreciated. 11:55
Ven just has been invited to "ladies who code paris" 11:56
I appreciate the sentiment, but ...
masak :)
Ven I don't code :x
masak Ven: if I may ask, what do you do? 12:00
masak pushed v0.0.2 of Data::Pretty, which also handles regexes :) 12:01
Ven masak: I joke around 12:04
masak ah.
sounds like a nice line of work :P
woolfy1 fwoolfy
Ven but I'd really love to do something instead of coding. Which really shows that, even in my mind, developer is not a nice job to have.
woolfy1 oops 12:05
Ven (or maybe it's just because I don't think very highly of fellows web developers ...)
lizmat Ven: writing a book, is that coding?
Ven lizmat: speaking on IRC, is that philosophing :) ? 12:06
lizmat wishful thinking? 12:07
Ven lizmat: desilusional one, I'd say :P 12:08
dalek ast: 6dfefc4 | (Elizabeth Mattijsen)++ | S22-package-format/local.t:
Some more CompUnit tests
ast: 669ec1f | (Elizabeth Mattijsen)++ | S17-supply/ (2 files):
Some more S17-supply tests, more to come
lizmat :-(
Ven thinks these words have a pretty nice ring to it, but they're probably barely english
lizmat franglais will also work here, maybe 12:09
teodozjan_ hi, i have a little question about phasers, they are documented that INIT is being executed on runtime but this is true only for one file project. If i have multiple file project ale imported classess are executed during compilation? This ends in funny results if I forget about this 12:10
lizmat teodozjan_: could you supply a test case ? 12:11
masak yes, an example, please?
teodozjan_ the compiling one or general?
jnthn They are run at module load time, after the module has been compiled, but before the module's mainline has been run 12:12
and modules are loaded immediately upon a "use" statement
lizmat cycling& 12:13
teodozjan_ gist.github.com/teodozjan/97f5322b601f760ccea8
teodozjan_ even though i added INIT 12:16
github.com/teodozjan/lacuna-cookbu.../Empire.pm
[Coke] masak++ #data::purty 12:17
jnthn teodozjan_: Yes, that's running at the point I'd expect, though I'm a little disappointed the debugger didn't place you inside the INIT phaser...
Ven
.oO( Data::Purr )
 12:18
gtodd did --> my %h = { … } work once upon a time? 12:19
I did not know it well and so will not be able to miss it :) 12:20
teodozjan_ jnthn: this means that initializing attributes in place is not good practice if it leaves some side effects like creating directory
Timbus gtodd, it.. did. does it not anymore? 12:22
m: my %h = { a => 1, b => 2 }; say %h.perl
gtodd just reading p6weekly.org
camelia rakudo-moar 87dcd1: OUTPUT«("a" => 1, "b" => 2).hash␤Saw 1 call to deprecated code during execution.␤================================================================================␤%h = itemized hash called at:␤ /tmp/7gjBYxsAc_, line 1␤Please use %h = %(itemized hash) in…»
gtodd I mean p6weekly.wordpress.com/
Timbus ooo
jnthn The problem was that my @a = [1,2,3] clearly treats the array on the LHS as an item, giving an array of array... And { ... } flattening away was rather inconsistent with that. 12:23
[Coke] rakudo has < 900 fudges that are not "RT, NYI, unspecced". whee. 12:24
Timbus yeah i noticed that a week or so ago, because i thought @a = [] was valid
then when i was corrected, the %h = {} thing was a bit odd.
jnthn Well, it is valid, it just doesn't mean @a = () :)
Timbus i'd kinda prefer if it was alloweed. is there any reason @a = [] behaves the way it does? too much special casing? 12:25
gtodd in pumpkin perl one has to learn about %hash = ( 'thing1' => '1' , 'thing2' => '2') versus %hash = { 'thing1' => '1' , 'thing2' => '2'} :-)
my %h = { … } would've scared me if I had known ...
Timbus it only seems scary from a perl 5 standpoint, imo 12:26
gtodd @a = [] is an AoA with one element? 12:27
jnthn Timbus: Because [...] and {...} should simply never flatten in an assignment.
gtodd: yes
gtodd %h = { } is a hash with errmm .... ok stop now :-)
Ven 2,9s/^#/ <- I can see why people like vim, tbh. 12:29
jnthn teodozjan_: Having modules do side-effects merely by loading them is probably best avoided in general, yes. I'd much rather a design where the top level instructs the module "do X"...
Ven [Coke]: oh, really ? That's great :) 12:30
m: my @a = [1, 2]; say @a.perl;
camelia rakudo-moar 87dcd1: OUTPUT«Array.new([1, 2])␤»
teodozjan_ jnthn: Maybe I spend too much time in java but I didn't expect that compiler will execute method in attribute initilizer during compilation :) 12:31
Timbus hm. i don't see it as flattening when you are assigning it. i guess the plot thickens when you are assigning to elements within other elements. idk
[Coke] m: subset MyInt of Int where True; my MyInt $x = 5; $x = Nil; say $x.WHAT; 12:32
camelia rakudo-moar 87dcd1: OUTPUT«(MyInt)␤»
[Coke] There is a test that checking that $x === Int there.
I could see ~~ Int... 12:33
teodozjan_ jnthn: especially that this didn't cause cannot invoke null object on parrot but on moar smells somewhat related 12:34
jnthn [Coke]: Yes, that test seems very dubious to me.
gtodd " < 900 fudges " wow ... they are disappearing at a faster rate?
jnthn I'd expect it to hold MyInt. It would be crazy for $x to hold Int there, because an Int type object is never true.
So it'd mean the variable holding something that doesn't meet the type constraint on it... 12:35
Should be === MyInt I guess.
timotimo jnthn: would we be putting a spesh function pointer into MVMContainerSpec that would work like the speshes for REPRs? 12:37
like, emit a sp_trivialdecont operation that simply accesses a pointer at an offset or something like that?
oh, this ought to have been #moarvm 12:38
jnthn teodozjan_: Module load time - and thus running the body of the module - is triggered the first time it's used. It does allow various powerful things, but yes, it'll be a surprise if you have a strict compile-time/runtime separation view of the world, as one has in Java. 12:40
teodozjan_ jnthn: After a month shootin myslef in foot I start understanding that :) but anyway parrot takes it totally different than moar 12:43
jnthn teodozjan_: Yes, that discrepancy is a bug 12:49
dalek osystem: 57451ae | sergot++ | META.list:
OpenSSL bindings module added
 12:50
masak jnthn: a reported one? 13:07
jnthn Not sure.
masak is there a minimal example? 13:08
Ven Only a month of shooting yourself in the foot ? You're quite good ;) 13:13
do we have any mail libraries ? *goes to check our lil' cpany* 13:16
timotimo i think we have something that handles mail headers? 13:19
but i could be wrong about that
Ven yeah. I was more looking forward something that could *receive*, but I guess I'll first try to do it in ruby because I need it in ruby :P 13:20
timotimo receive mails? oh god, the pain ... 13:21
or did you mean retrieve? like POP3 or IMAP?
Ven timotimo: first retrieve, then at some point receive. 13:22
timotimo you know receiving mails is basically the worst thing imaginable? 13:22
well, at least configuring mail servers is said to be a terrible pain 13:23
teodozjan_ Ven: I didn't know how to load gun earlier ;) 13:25
Ven what's gun ?
masak &mkdir doesn't seem to complain for directories that already exist. bug? 13:27
Ven: the tool with which to shoot yourself in the foot.
Ven masak: oh. How do you use it ?
masak Ven: howto-pages.org/shootfoot.php 13:28
Ven thankies. 13:29
There should be a Perl 6 one "you wait for a compiler to be advanced enough to do it, and retire before you can actually shoot yourself in the foot" :D 13:30
masak (proudly) shoots himself in the foot daily with an advanced Perl 6 compiler 13:32
so speak for yourself :)
Ven can't handle a little irony, eh ? :)
Ven shoots itself often enough with his perl 6 in prod, sooo 13:33
masak clearly the same person who trolled his own channel earlier today pretending to be a clueless outsider troll can't handle a little irony, noooo :P
nwc10 masak: are you in urgent need of a weekend? 13:34
Ven aren't we all. Hey, 3 days weekend !
nwc10 3? your mileage seems to be varying
masak nwc10: oh, do I have the appearance of needing one? :/ I just have more time today to fool around than usual...
masak is having a Perl 6 day today
nwc10 aha
masak is actually writing Perl 6 right now 13:35
nwc10 oooh
Ven It's france, we don't work very much.
masak .oO( that "strikes" me as plausible )
nwc10 OK, I can't quite work out a dual language pun 13:37
but in the process of failing to, why is en.wiktionary.org/wiki/gr%C3%A8ve a grave accent, but en.wiktionary.org/wiki/gr%C3%A9viste an accute accent?
masak .oO( the strike itself is resisiting, but all the participants are active? ) :P 13:38
Ven nwc10: because french. happens all the time ...
nwc10 ah OK.
masak Ven: I thought French was the ultimate paragon of logic and clarity! 13:39
Ven masak: I bet you did ..
Ven really hates french as a lang
timotimo masak: luckily, writing "I'm actually writing Perl 6 right now" is an act of "writing Perl 6" :3
masak timotimo: oh, you noticed.
colomon_ I think most of the new module failures from the last two days are because eval finally went away. 13:46
Ven now wants to see EVAL away :P 13:47
retupmoca Ven: I have IMAP and POP3 modules written. SMTP too, but only client-side 13:55
Ven: as well as Email::MIME
masak wow, :i is really amazingly slow. 14:04
I'm measuring a ~35x slowdown with it.
colomon_ :( 14:05
colomon_ PerlJam: Benchmark and File::Temp are failing because of eval -> EVAL. 14:07
masak: ditto for Yarn 14:08
masak removes Yarn from ecosystem 14:10
dalek osystem: fef9a4e | masak++ | META.list:
remove yarn

Not maintained, and not very interesting anymore.
See zblog instead. :)		 14:12
PerlJam pushes some updates
[Coke] lizmat: so apparently, no, six months was not enough time for, say, PerlJam. :P :) 14:13
Ven has a lot of cool ideas of projects using perl 6, but ... they all involve web 14:17
masak wonders how far pmurias++ got with his JS backend, and if something from that project can be salvaged and resumed 14:18
Ven masak: it has a lot of PIR, sadly 14:19
masak aww :/
Ven (I didn't mean "web" in "perl 6 as an alt-js" -- just in case)
masak no, I get that.
masak but many of *my* plans would involve generating JS from Perl 6. 14:19
Ven lots of stuff we can't quite compile :P 14:20
and lots of stuff we can just cheat about :D
jnthn masak: (:i) any chance of a small benchmark? :)
masak as a matter of fact, yes.
[Coke] well, jnthn could probably do rakudo -js in about 2 weeks, given previous performance and improved porting tools! 14:21
masak m: say +("flubb" x 1000).comb(/u/); say now - BEGIN now
camelia rakudo-moar 87dcd1: OUTPUT«1000␤0.16337649␤»
masak m: say +("flubb" x 1000).comb(/:i u/); say now - BEGIN now
camelia rakudo-moar 87dcd1: OUTPUT«1000␤0.81209080␤»
masak it gets worse for larger values of 1000.
btyler Ven: nqp-js has a lot of pir? are you sure that isn't just the pir that's in nqp to begin with? he forked perl6/nqp to start with
Ven btyler: erm, I checked pretty early on in my p6-learning-experience, will recheck 14:22
btyler Ven: I don't see any PIR in this subdir: github.com/pmurias/nqp-js/tree/master/src/vm/js :)
Ven btyler: github.com/pmurias/nqp-js/blob/mas...nd.nqp#L72 :P 14:24
PerlJam shouldn't that just be nqp::istype or something anyway? :) 14:25
Ven there are some 'pir' here and there
btyler Ven: ok, fair, missed a few lines. still, hardly such an amount as to present an obstacle from working on it
an obstacle to, even 14:26
Ven Yeah, I probably misread some of these files as being of the rakudo-js project :) 14:26
masak insight: the dark side of the coin of memoization is memory leakage -- and doing memoization without considering life cycles is some degree of irresponsible.
Ven masak: that's specialization in general :P 14:27
jnthn May also simply be that at the time there was no nqp:: alternative, and that nowadays there is...
colomon_ [Coke]: me neither, I had to fix two of my modules this morning. 14:29
[Coke]: also supernovus, it appears 14:31
hmm: /Users/colomon/.rakudobrew/bin/perl6: line 2: 40745 Segmentation fault: 11 /Users/colomon/.rakudobrew/moar-HEAD/install/bin/perl6 "$@"
hoelzro morning #perl6 14:45
masak hoelzro! \o/ 14:46
hoelzro o/ masak!
gtodd hmm would it be accurate to say there are "Java based" languages and there are languages that "run on the JVM" and that they are very different ... 14:50
masak gtodd: I think so. 14:51
gtodd scala seems "like java" but jruby and rakudo-j (and clojure !) are not like java
well I think scala is like java ... :-)
masak gtodd: I mean, to some extent C# is "Java-based", but it doesn't run on the JVM as far as I know.
gtodd maybe I just mean strongly typed or something
jnthn C# also evolved a lot faster :) 14:52
gtodd haha
more money and fear involved
:)
masak well, I admire the direction in which C# has evolved.
I think Microsoft Research is largely to thank for that. 14:53
jnthn *nod*
They got a lot of things right.
gtodd I guess there's even more languages the run on .NET ...
gtodd always mixes up misuses C# .NET CLR etc 14:54
dalek kudo/spesh_decont_contspec: e9c4921 | (Timo Paulssen)++ | src/vm/moar/ops/container.c:
teach Rakudo_Scalar how to spesh.
 14:55
cognominal how far behind is mono?
gtodd anyway ... a thing a perl5 person might notice or see as different better new is that perl6 can haz "Types" ... (whee!) but someone using scala might just see it as a way of using/accessing java/jvm with cleaner syntax 14:56
TimToady gets to have a second emergency eye surgery today; well, not the second one today, the second one this week 14:57
sjn hopes it goes well 14:58
TimToady sprouted another retinal detachment night before last, and another last night, but I was able to use the existing bubble (of sulfur hexafloride!) to stabilize my eye temporarily
retupmoca TimToady: best of luck
TimToady they're going to do a vitrectomy this time, to take away my jelly and replace it with something they can see through :) 14:59
masak best wishes, TimToady.
Ven disagrees that java is "strongly typed" 15:00
gtodd only asking because sometimes people list all the languages they know and I see java scala groovy ... and I wonder about adding [ ] around parts of the list :)
[Coke] hopes TimToady can see through it also!
[Coke] wonders if not-strongly-typed java is a 7 or 8 thing.
jnthn TimToady: Hope all goes to plan...and your retina gets better at staying attached... 15:01
TimToady depending on whether they put in the two-week gas or the eight-week gas, I may not be able to fly to Bulgaria and Japan
Ven notices that TimToady is jelly
:(
gtodd TimToady: 8-] take care of yourself! Maybe use skype :) 15:02
hoelzro best of luck TimToady 15:04
gtodd [Coke]: I don't know java or java history that well but it does seem like a more recent thing 15:05
nwc10 TimToady: are you going to OSCON or not? (I assume that that's driving, not flying, but it's not clear if even "being driven" is compatible with your current "interesting times")
and yes, I hope it goes well 15:06
TimToady OSCON is up in the air too, though I can potentially at least drive there, if I avoid mountain ranges
but it's also not clear that I should go that far from my eye doctors at this point 15:07
we'll see how it goes
gtodd is it an ai rpressure thing or a don't move your eyes around too much thing?
TimToady air pressure, or more precisely, sulfur hexafloride pressure :)
or whatever gas they use next time 15:08
gtodd ahah that sums it up .. heh "don't go that far from my eye doctor" thing
wow
TimToady it's up to doctors by now :) 15:08
liztormato TimToady: Best wishes from woolfy, elbeho and me 15:09
TimToady so far I've had six doctors looking in my eye since Sunday
three at UCLA, and three up here at home
liztormato That's quite a set of eyes to look at your eyes 15:10
TimToady this all started with a PVD a week ago
which is benign unless it tears your retina
apparently mine tore in three places 15:11
TimToady unfortunately a PVD tends to fill up your eye with all kind of floaters large and small, which make it hard to see the tears 15:11
liztormato www.rnib.org.uk/eye-health-eye-cond...detachment 15:12
masak found en.wikipedia.org/wiki/Posterior_vi...detachment
TimToady I had one a year or so ago, but it was not a problem
well, except for the floaters... 15:13
after today I should not have that problem again, at least in that eye 15:14
TimToady s/floride/fluoride/ I guess 15:28
timotimo TimToady: i hope all goes according to plan from now on and you'll recover swiftly! 15:36
hoelzro this is probably a stupid question, but is it possible to get a variable's container (so that I can ask it what type it is) 16:05
atroxaper Evening, #perl6 ! 16:06
I cant imagin why Parrot throws SIGSEGV... If I push say "-" in random place then all is ok. 16:09
masak hoelzro: .VAR is supposed to do that. 16:10
m: my $a = 42; say $a.VAR
camelia rakudo-moar 87dcd1: OUTPUT«42␤»
masak m: my $a = 42; say $a.VAR.^name 16:11
camelia rakudo-moar 87dcd1: OUTPUT«Int␤»
masak but, hm.
hoelzro ahhhh 16:11
masak it doesn't seem to do what I expect.
japhb_ TimToady: Best wishes on a good and speedy recovery. Also, it's hard for me to think of 'sulfur hexafluoride' without thinking 'poison hexapoison'. But then sodium chloride, so ...
hoelzro that's kind of odd 16:12
but it's working for me
m: my $a = 42; say $a.VAR.of.WHAT 16:13
camelia rakudo-moar 87dcd1: OUTPUT«(Mu)␤»
[Coke] belatedly wishes TimToady g'luck. 16:13
if you can reliably segfault parrot: 1) make sure you're using the latest version of rakudo; 2) run it through gdb or something to get a stacktrace 3) open a ticket, please. 16:14
er, atroxaper ^^
jnthn m: my $a = 42; say $a.VAR.WHAT 16:15
camelia rakudo-moar 87dcd1: OUTPUT«(Scalar)␤»
jnthn m: my $a = 42; say $a.VAR.name
camelia rakudo-moar 87dcd1: OUTPUT«$a␤»
atroxaper [Coke]: I'm interested in point 2. How? :)
jnthn m: my Int $a = 42; say $a.VAR.of # the type 16:16
camelia rakudo-moar 87dcd1: OUTPUT«(Int)␤»
[Coke] atroxaper: what OS? 16:17
if not-windows, do you have gdb installed?
atroxaper [Coke]: os x
[Coke] try "$ gdb ./perl6"
"> run /path/to/exploding/script.p6" 16:18
wait until it explodes.
"> bt"
gist that so we can see if you have enough info there. Might have to rebuilding r-p with debugging symbols enabled.
atroxaper [Coke]: ok. I'll try 16:19
[Coke] atroxaper++
dalek kudo-star-daily: 8e8376c | coke++ | log/ (14 files):
today (automated commit)
rl6-roast-data: 4cc469a | coke++ | / (4 files):
today (automated commit)
[Coke] lots of %(itemized hash) deprecations in the R*. 16:20
modules/jsonrpc failing on parrot
; ... this is all in the daily runs of the other thing now, yes?
masak the other thing? do you mean emmabot? 16:24
[Coke] there is a website that was showing all build issues/test failures, etc... 16:25
masak oh! 16:25
[Coke] mumble:8080
[Coke] .seen au 16:25
yoleaux I haven't seen au around. 16:25
[Coke] ^^ I'm going to shut off Pugs on the daily runs, I think. 16:26
It's just too crufty, and no one wants to own it.
masak +1
[Coke] niecza I may make another effort to fudge with a bunch of generic "todo" fudges. 16:27
masak Niecza is probably also not a moving target these days -- but of course, getting spectest results from it might still be interesting. 16:29
[Coke] it is more likely someone might do a small commit here or there to track a change, I think. 16:31
dalek rl6-roast-data: 6c41218 | coke++ | bin/ (2 files):
Stop running Pugs every day.

No one is maintaining it, no real change in months.
Thanks for getting us started! au++ (lots of others++)
Adios. (We'll leave the last log and the data in the daily tracker, but remove the runner)		 16:33
[Coke] Step 2 of that, which is more permanent, is to remove all the pugs fudges from roast. 16:34
masak even that is fairly reversible thanks to Git. 16:45
[Coke] Troo. 16:47
atroxaper With one Rakudo-Parrot version is the last? 16:53
This is perl6 version 2014.06-12-ga7b4cba built on parrot 6.5.0 revision RELEASE_6_5_0
FROGGS atroxaper: you see the revision of the latest commit on the right of the topmost entry: github.com/rakudo/rakudo/commits/nom 16:55
raiph m: slurp.subst( / \N*? (\w+) \n { say $0} /, $0, :g) # this seems to work
camelia rakudo-moar 87dcd1: OUTPUT«｢nGall｣␤␤｢luí｣␤␤｢chúl｣␤␤｢féin｣␤␤｢gnéill｣␤␤｢stór｣␤␤｢mór｣␤␤»
raiph m: slurp.subst( / \N*? (\w+) \n /, $0, :g).say # but this doesn't; what am I doing wrong? 16:56
camelia rakudo-moar 87dcd1: OUTPUT«Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤I mo dhiaidh bhí gleanntáin ghlas’ Ghaoth Dobhair, is beag nár bhris mo chroí.␤␤Ní miste dom ’ rá le brón ’s le crá gur frasach a shíl mise siúl;␤Go Meiriceá…»
FROGGS so when you see 87dcd11386 there, it would be represented as YYYY-MM-nr-g<commit id>
m: slurp.subst( / \N*? (\w+) \n /, { $0 }, :g).say
camelia rakudo-moar 87dcd1: OUTPUT«nGallAgus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤luíI mo dhiaidh bhí gleanntáin ghlas’ Ghaoth Dobhair, is beag nár bhris mo chroí.␤␤chúlNí miste dom ’ rá le brón ’s le crá gur frasach a shíl mise siúl;…»
FROGGS ahh 16:57
raiph: I think :g in the subst method for is broken
it will work in the non method version
raiph m: subst(slurp, / \N*? (\w+) \n { say $0} /, $0, :g).say 16:58
camelia rakudo-moar 87dcd1: OUTPUT«===SORRY!=== Error while compiling /tmp/TDwM6EQepj␤Undeclared routine:␤ subst used at line 1. Did you mean 'substr'?␤␤»
atroxaper FROGGS: it is funny that i can't find my current's version commit :) git log ga7b4cba -> nothing 17:00
FROGGS atroxaper: strip the leading g 17:01
m: my $t = slurp; ($t ~~ s:g/ \N*? (\w+) \n { say $0} /$0/).say # raiph
camelia rakudo-moar 87dcd1: OUTPUT«｢nGall｣␤␤｢luí｣␤␤｢chúl｣␤␤｢féin｣␤␤｢gnéill｣␤␤｢stór｣␤␤｢mór｣␤␤True␤»
dalek ast/S26-WHY: 709b077 | (Rob Hoelz)++ | S26-TODO:
Add note about indent for declarative blocks
 17:02
atroxaper FROGGS: oh.. Thanks! 17:03
FROGGS m: my $t = slurp; ($t ~~ s:g[ \N*? (\w+) \n { say $0} ] = 42).say; say $t
camelia rakudo-moar 87dcd1: OUTPUT«｢nGall｣␤␤｢luí｣␤␤｢chúl｣␤␤｢féin｣␤␤｢gnéill｣␤␤｢stór｣␤␤｢mór｣␤␤True␤42Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤42I mo dhiaidh bhí gleanntáin ghlas’ Ghaoth Dobh…»
atroxaper FROGGS: sorry, but what is 'nr' in release version format? 17:12
japhb atroxaper, number of commits since the tag 17:13
atroxaper, because you're really seeing the result of `git describe` 17:14
I personally use `git describe --always --dirty`, IIRC
(It's wrapped up in something else) 17:15
atroxaper japhb: oh. I see. 17:16
FROGGS atroxaper: so you know you are 12 patches ahead of the 2014.06 release 17:26
atroxaper FROGGS: yes. I see. I updated just now. I my sigsegv has gone :) And now I have a couple of deprecated calls. 17:27
FROGGS atroxaper: yes, that sounds about right :o) 17:28
atroxaper m: my @a = 1,2,3; say @a.WHAT; 17:39
camelia rakudo-moar 87dcd1: OUTPUT«(Array)␤»
atroxaper ok 17:40
hoelzro when does 2014.07 go out, again? the 17th, iirc? 17:49
PerlJam hoelzro: aye. 07-17 17:52
hoelzro I better mosey with the S26 stuff, then =) 17:53
atroxaper I overcame my deprecated calls... But i have a new issue :) gist.github.com/atroxaper/ad0714f011f32fd8cbd9 17:58
emilper perl6: use IO::Socket::INET;
camelia rakudo-moar 87dcd1: OUTPUT«===SORRY!===␤Could not find IO::Socket::INET in any of: /home/p6eval/rakudo-inst-2/languages/perl6/lib, /home/p6eval/rakudo-inst-2/languages/perl6␤»
..rakudo-jvm 87dcd1: OUTPUT«===SORRY!===␤Could not find IO::Socket::INET in any of: /home/p6eval/rakudo-inst-1/languages/perl6/runtime, /home/p6eval/rakudo-inst-1/languages/perl6/lib, /home/p6eval/rakudo-inst-1/languages/nqp/lib, /home/p6eval/rakudo-inst-1/languages/perl6/lib, /home…»
..rakudo-parrot 87dcd1: OUTPUT«===SORRY!===␤Could not find IO::Socket::INET in any of: /home/p6eval/rakudo-inst-2/lib/parrot/6.5.0-devel/languages/perl6/lib, /home/p6eval/rakudo-inst-2/lib/parrot/6.5.0-devel/languages/perl6␤»
..niecza v24-109-g48a8de3: OUTPUT«Unhandled exception: System.Exception: Unable to locate module IO::Socket::INET in /home/p6eval/niecza/lib /home/p6eval/niecza /home/p6eval/.config/niecza␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1502 (die @ 5) ␤ at /home/p6eval/niecza…»
atroxaper if i use just '$anchor."{$name}"() = %args{$name};' than in case $name eq '%storage' then i have a message about deprecated call and should write %storage = %(itemized) 17:59
I mean I am forced to write a type of left argument, but I dont know it. 18:00
atroxaper I can to find out the type of left argument, but I think it is dirty... 18:00
psch hi #perl6 o/ 18:16
github.com/rakudo/rakudo/pull/284 # now rebased and does-what-it-should
atroxaper psch: Hi!
psch if i can't i'd like to defer spectests to someone with a beefier machine; moar alone runs for about 30 minutes or more here with two jobs 18:17
hey atroxaper :) 18:18
psch starts building all backends and spectesting them anyway, because waiting two hours before someone else runs them in 1/5th of the time when it could be finished by then is not really efficient. 18:22
dalek kudo/spesh_decont_contspec: 3c02865 | (Timo Paulssen)++ | src/vm/moar/ops/container.c:
gotta subtract the header's size from the offset
 18:24
rjbs woolfy: When we were in Lyon, I had some of your jenever. Today, I bought a bottle of the kind of gin that is considered a "missing link" between genever and London gin! 18:26
woolfy: en.wikipedia.org/wiki/Old_Tom_Gin
nwc10 mmm, gin
rjbs (I got it after a powerful endorsement by [particle].)
(So this really is on topic for #perl6!) 18:27
TimToady you just have to find some way to tie it to linguistics
rjbs Speaking of drinking and language, today I was trying to find the origin of the meaning "stand you a drink" for stand.
It was verb definition #61 (!) of "stand" in OED, but gave no origination. 18:28
japhb I somehow expected "Speak of drinking and language," to be followed by increasingly unreadable gibberish .... 18:30
PerlJam japhb: maybe that's why rjbs can't find the origin of that phrase. 18:31
atroxaper i've created a bug report rt.perl.org/Ticket/Display.html?id=122274. I still have SIGSEGV 18:34
Mouq_ RE: irclog.perlgeek.de/perl6/2014-07-11#i_9009278 18:37
You need / \N*? (\w+) \n {} /
I believe... 18:38
But the {} is supposed to create $/
atroxaper Mouq_: Hello!
Mouq_: did you see Pod::SAX code?
Mouq_ atroxaper: Hi! :) I did get a chance to look at it a little. I imagine there's gotta be a cleaner syntactic way to write Pod::SAX::Goes::HTML 18:46
japhb Goes? 18:47
Mouq_ (I'd just name it Pod::To::HTML) 18:48
atroxaper Mouq_: Do you imagine some concrete syntactic way? 18:48
Mouq_ m: say (4, 5) ~~ sub ($, $) { True }.signature 18:49
camelia rakudo-moar 87dcd1: OUTPUT«True␤»
Mouq_ m: say (4, 5) ~~ :($, $)
camelia rakudo-moar 87dcd1: OUTPUT«True␤»
Mouq_ m: say (4, 5) ~~ :(4, 5)
camelia rakudo-moar 87dcd1: OUTPUT«True␤»
Mouq_ m: say (4, 5) ~~ :(4, 6)
camelia rakudo-moar 87dcd1: OUTPUT«False␤»
Mouq_ ^^ There's one way to clean it up a little; instead of using sub (...) { True } everywhere, use :(...)
atroxaper Mouq_: Hmm... ok. What else? :) 18:51
Mouq_: Do you like a general idea of start, in, stop callbacks and post calls of anchors? 18:52
Mouq_ Honestly, I really didn't at first, but I can see how it would be quite extensible. 18:55
Mouq_ So yes :) 18:56
Mouq_ If you wanted, you could rewrite all the sub (:@draft, :@meta, etc) { }, as blocks that have @:draft, @:meta, etc, not sure if that would increase or decrease readibility; it's just that the same signatures appear over and over again 18:57
(@:draft is like @^draft, but named instead of positional) 18:58
m: { @:draft.say }( :draft[1,2,3] )
camelia rakudo-moar 87dcd1: OUTPUT«1 2 3␤»
atroxaper Ohh... I thought about repetitive signature... I don't like it too. 18:59
But how can I find out which named arguments uses in each concrete block? 19:01
I use a signature for that in my current realization. or you think that I don't check arguments and put it all?
Mouq_ There also doesn't seem to be a case where you're not just pushing to @draft; maybe make it avaible (since I see where it could be useful to manipulate directly) but maybe just have whatever the block returns be pushed onto @draft?
m: { @:draft.say }( :draft[1,2,3], :passed-anyway ) 19:02
camelia rakudo-moar 87dcd1: OUTPUT«Unexpected named parameter 'passed-anyway' passed␤ in block at /tmp/eYqF7779kW:1␤␤»
Mouq_ aww, what?
m: sub { @:draft.say }( :draft[1,2,3], :passed-anyway )
camelia rakudo-moar 87dcd1: OUTPUT«Unexpected named parameter 'passed-anyway' passed␤ in sub at /tmp/Jznfd2tjsD:1␤ in block at /tmp/Jznfd2tjsD:1␤␤»
Mouq_ That's annoying
m: say { @:draft.say }.signature
camelia rakudo-moar 87dcd1: OUTPUT«:(:draft(@draft)!)␤»
Mouq_ ^^ Looks like the sig is still introspecable, though. I don't think you'd have to change anything in the backend for that 19:03
*introspectable
atroxaper Mouq_: oh... that is great! 19:04
jnthn psch: Looked over your latest PR and I'm happy with it. I guess your spectest run is some way along now...lemme know how it works out :)
atroxaper Mouq_: about @draft. I'm not prefer to push returns value to @draft. I thought, if sub return False, then we call the next sub if it exists. 19:05
psch jnthn: i'm S04 for jvm; parrot and moar come after that... :) 19:06
jnthn k 19:07
I'm here all evening... :)
atroxaper Evening... I have 2am already.
Mouq_ if @returned { push @draft, @returned } else { call-the-next-sub } ? 19:08
You could also return Nil and say if defined @returned, though I'm still just postulating 19:09
It may be better to just push to @draft
atroxaper Mouq_: And then we remove each :@draft in signature, but add "my @returned;" in third part of code :) 19:12
Mouq_: But there will less code in any way. You are right.
Mouq_: Now I'm going to finish with Pod::To::HTML and its tests. And after that refactor the solution. 19:15
Mouq_: it's so sad that i have SIGSEGV on one of my test of Pod::To::HTML on Parrot. 19:16
atroxaper Ok. nights, #perl6 ! 19:20
Mouq_ atroxaper: What about having @*draft be a dynamic variable, and having a sub append (*@args) { push @*drafts, @args } ?
atroxaper: g'night!
atroxaper Mouq_: mmm.... I didn't understand... 19:21
Mmmm.
Mouq_: Good idea!
Mouq_: I would say, Great idea! :)
Mouq_: But... I thinks that I don't understand fully xD 19:23
Mouq_: I need to read about dynamic variable! I'm going to do so tomorrow. Nights! 19:25
Mouq_ atroxaper: Haha, sorry. in !visit or somewhere higher on the call chain, define a my @*draft := self.draft. Dynamically scoped variables are to all the functions that are called in the scope of the variable, so the subs in Pod::SAX::Goes::HTML could just push to @*draft
atroxaper ...
Mouq_ atroxaper: Sorry, I tend to come and go from the computer 19:26
atroxaper Mouq_: yes! Great Idea!
Mouq_ atroxaper: Haha, ok :) 'nighty
atroxaper Mouq_: you too :-)
dalek kudo/spesh_decont_contspec: deef0e3 | (Timo Paulssen)++ | src/vm/moar/ops/container.c:
it helps to have your piece of memory for yourself.
 19:30
dalek osystem: 4a96e8b | (David Warring)++ | META.list:
Renamed a couple of repositories

perl6-CSS-Language => perl6-CSS-Module PDF-Grammar => perl6-CSS-Grammar		 19:41
psch apparently i still had the non-solution in my nqp-j for the missing outer... 19:42
also -Xmx3000m for perl6-eval-server isn't great for 32bit machines :) 19:43
psch restarts spectests
dalek p: efba64d | (Timo Paulssen)++ | tools/build/MOAR_REVISION:
bump moar revision to get bigger ContainerSpec
 19:45
kudo/nom: e9c4921 | (Timo Paulssen)++ | src/vm/moar/ops/container.c:
teach Rakudo_Scalar how to spesh.
 19:46
kudo/nom: 3c02865 | (Timo Paulssen)++ | src/vm/moar/ops/container.c:
gotta subtract the header's size from the offset
kudo/nom: deef0e3 | (Timo Paulssen)++ | src/vm/moar/ops/container.c:
it helps to have your piece of memory for yourself.
kudo/nom: e83c5a3 | (Timo Paulssen)++ | tools/build/NQP_REVISION:
bump NQP to get bigger ContainerSpec
timotimo this ought to make the jit a bunch happier 19:47
fabio__ lhi 19:49
hi
timotimo hello there :)
fabio__ :) 19:50
timotimo that was a short visit :( 19:50
pippo .tell TimToady I hope all goes well with your eyes. Best wishes. 20:07
yoleaux pippo: I'll pass your message to TimToady.
dalek ast: 75b7b15 | coke++ | S02-types/nil.t:
Fix busted test; Should use declared type.

  jnthn++		 20:10
ast: 05d8a73 | coke++ | .gitignore:
To ignore your tmpfiles, use local global gitconf
ast: fcf62f1 | coke++ | / (301 files):
Remove any Pugs fudges

  * or pugs-specific comments
  * one pugs-specific test file
830b276 | coke++ | S12-attributes/delegation.t: remove commented out "cannot work" test
  (marked so by pmichaud++)
masak [Coke]++
jnthn ooh, I go for a walk and lots of commits happen 20:11
I should go for more walks.
masak we all should.
[Coke] dalek gave up: ended with "Thanks for getting us started!"
"so say we all
nwc10 if we all went for walks at the same time, would it still work? 20:12
are there commit faries who only come out when no-one is logging?
masak stranger things have happened 20:14
itz__ does moar work OK with clang? 20:23
vendethiel lots of warning iirc 20:24
[Coke] .tell colomon_ github.com/sorear/niecza/issues/191
yoleaux [Coke]: I'll pass your message to colomon_.
jnthn Yes, not heard any reports of it not working. Just that, while we're warning-free on gcc and msvc, clang likes to warn about everything imaginable. :) 20:25
BenGoldberg That's why it's called "clang"... it's so noisy ;) 20:33
japhb BenGoldberg, "clang" always makes me think of brutish warrior types -- like a name you'd give an ogre. 20:39
BenGoldberg . o O (Clang the conquerer?) 20:52
[Coke] should roast prefer 4-space indents and no tabs? 20:55
(if so, bunch of LH formatting fruit)
timotimo i usually dislike changing only formatting in commits as it makes git blame pretty much useless 21:00
[Coke] you can still follow the trail to find out who the original editor was. 21:01
I dislike it when people commit the wrong formatting!
timotimo fair enough
i suppose it'd take way too long if commits would only fix formatting local to other changes, eh? 21:02
[Coke] but not enough to try to reformat things. (I got to the point where I saw how many tabs were in the source, cried, and then stopped)
masak there should be a formatting-impervious git blame. 21:02
[Coke] I officially have no problem with formatting only commits, as long as the message is explicit.
timotimo japhb: you think it'd be a good idea to add a "rakudo-moar-jit" and "nqp-moar-jit" component to bench that puts --enable-jit into the --moar-options?
[Coke] masak: git is already making stuff up to get from one revision to the next, so that should be doable, yes. :) 21:03
[Coke] (e.g. when you duplicate/edit a file at the same time.) 21:03
Mouq_ ack "\t" | wc -l -> 1908
cognominal japbh: is it dangerous to cling to clang? 21:04
masak [Coke]: right. Git has kind of opened my eyes to how much info can actually be recreated after the fact. it's a counterintuitive but good design.
[Coke] Mouq_: not all of them can be converted to spaces without analysis. 21:05
japhb timotimo, yes, an {nqp,rakudo}-moar-jit seems useful. I guess it depends on how soon JIT merges. Because it might be worth it to have {nqp,rakudo}-moar-nojit. But then you get into -noopt and other combinations as well. 21:09
timotimo yeah :( 21:10
japhb cognominal: If clang is wearing spiked armor or flea-infested skins, then yes. :-)
timotimo maybe we're going to wait with the jit-merge until gsoc has finished? though the official docs also encourage merging in contributions swiftly throughout the GSoC
jnthn I'm aiming we'll include it in 2014.08 21:11
timotimo that sounds good!
japhb I'll call that a birthday present. :-)
[Coke] to about 8 of us. 21:12
timotimo hmpf. i'll have to add --moar-option to the configures of rakudo and nqp it seems
wrong. nqp already has that
jnthn It's not so much about "is GSoC done"; it's more "are the big pieces in place" and "is it doing OK with spectests and ecosystem" 21:15
A merge a bit after 2014.07 gives us time to be happy that's the case. 21:16
Mouq_ Does the jit slow down start-up noticeably? 21:16
Mouq_ std: q \ / \ / \ / q \ / \ / \ 21:20
camelia std 0f2049c: OUTPUT«ok 00:01 123m␤»
jnthn Mouq_: Didn't measure, but probably not a lot, given it only JITs hot things, and producing machine code instead of specialized bytecode is only so much more work... 21:22
Mouq_ jnthn: Ok, cool :)
Mouq_ jnthn: Thank you 21:23
jnthn Getting too quiet here... 21:26
dalek kudo/nom: e0cbe53 | jnthn++ | src/vm/moar/Perl6/Ops.nqp:
Fix a .DEFINITE code-gen bug.

Should not blindly re-use registers that may belong to locals.
jnthn ...so I merged lex2loc 21:27
dalek kudo/nom: 14eedae | jnthn++ | src/vm/moar/ops/container.c:
Fix MSVC build.
 21:28
timotimo jnthn: nice :) 21:28
jnthn Note that it's not that aggressive yet. 21:31
It only does it for parameters and things that we bind to upon initialization.
timotimo oh, i suppose that's all right
jnthn Well, it's a good start :)
timotimo at least the jit can work with lexicals by now, so having lexicals at all isn't going to make it bail any more afaict 21:32
dalek kudo/nom: 1f05e14 | (Timo Paulssen)++ | Configure.pl:
expose --moar-option in Configure.pl
 21:35
timotimo this ought to be enough, right?
jnthn timotimo: Right. 21:38
timotimo it's a bit annoying to massage local commits into perl6-bench :\ 21:46
vendethiel
.oO( I see you take good take of your commits )
 21:49
timotimo %)
vendethiel care*. dammit 21:49
timotimo huh. 21:53
i git clone the moarvm/moarvm.git which has my jit-merged branch in it manually and git checkout jit-merged works
but when i clone it via the configure.pl with --gen-moar=jit-merged it doesn't work
oh! 21:54
japhb: --gen-moar clones a MoarVM into nqp/, but perl6-bench clones moarvm into ./
timotimo oh, suddenly it works much better :) 21:55
japhb That sounds like someone did a copy-pasta from nqp-moar to rakudo-moar in the components build instructions. 21:56
Care to fix, SVP?
timotimo i'm fixing it 21:56
japhb Thanks!
japhb timotimo, What did you mean by "it's a bit annoying to massage local commits into perl6-bench"? 21:58
timotimo that was before i figured out that perl6-bench was hitting up github for each moarvm checkout 21:59
and i got frustrated at trying to get my local "jit-merged" branch in there
dalek rl6-bench: d8d5301 | (Timo Paulssen)++ | components.json:
Configure.pl --gen-moar puts a moarvm under nqp/MoarVM, not moarvm.
 22:01
japhb Ah. Yeah, it's a major design feature that `bench setup` and `bench fetch` should be the only things doing network pulls (except for niecza, which I can't do much about, sadly).
And breaking that was :-(
timotimo++ for noticing and fixing it 22:03
timotimo going to run a benchmark with rakudo-moar-jit now :) 22:04
and a jitless rakudo-moar to compare it against 22:05
this is going to be interesting :)
jnthn timotimo: Given it can't JIT extops yet and just about everything in Rakudo contains one... :)
japhb I bet you'll see a bigger difference with the respective nqp's.
timotimo i suspect it'll be quite visible where OSR and the jit accepts the code and where either bails
jnthn: oh damn! i forgot about that :(
jnthn What japhb said though; at this point the NQPs may come out more itneresting.
timotimo okay, i'll do those instead 22:07
masak 'night, #perl6 22:10
timotimo japhb: got an idea how to make moarvm not need to hit the 'ternets for the submodules it includes? 22:15
japhb timotimo, thinking 22:15
timotimo Configure.pl of moarvm could be taught a "reuse-submodule-repositories" option or something 22:16
japhb timotimo, I think it is possible by cloning into the proper 3rdparty/ subdirs and using `git submodule` carefully. 22:17
With a little experimentation, we ought to be able to come up with the magic sequence of clones and submodule commands so it Just Works. 22:18
japhb Do any of the submodules recurse? 22:20
timotimo right ... yay :|
japhb I can try later, but if you want to try now, I'd do this: 1. Add each submodule as a component. 2. `bench setup` to get the new components. 3. Try changing the rules for moarvm to clone them into the right 3rdparty directories before running configure. 4. Check if this Just Worked. If not, we can try something else. :-) 22:22
japhb timotimo, and you can find the proper submodule clone locations from .gitmodules in the moarvm root. 22:23
timotimo i can try that later, i'll put up the nqp comparison benchmarks first, though 22:24
timotimo t.h8.lv/p6bench/2014-07-12-nqp-jit.html and t.h8.lv/p6bench/2014-07-12-nqp-jit-ignored.html 22:29
it seems like spesh and/or the jit bail on every of those
japhb: i think i'll do the submodule pre-cloning stuff tomorrow rather than today and go get some rest
japhb timotimo, Sleep well! 22:31
timotimo will try :)
woolfy rjbs: thanks for the link and the info. Next time I'm in the UK, I will ask for a Old Tom gin at the bar. Or ask for a small bottle in a liquor shop. 23:24
rjbs: next time we meet, I have to bring some super-old Dutch jenever. I got a bottle of XO, and it almost tastes like old whisky. 23:26
This is of course nothing related to Perl 6. Sorry.
vendethiel is pretty sure beer is related to P6, at that point 23:27
woolfy This is not about beer... :-) 23:28
vendethiel drinks, still 23:30
woolfy Indeed. 23:31
jnthn Yeah, on topic... :P 23:33
jnthn almost froze his beer today while trying to cool it sufficiently...
woolfy TimToady: we will try to tie as many things to linquistics, just to make you happy! Dutch for beer is bier, German word is Bier. French is bière. It is all delicious, or lekker, or lecker, or bon. 23:35
Ow, something Perl 6 related... 23:36
Our French friend elbeho is visiting us in Echt. He is now the 5th official board member of the YAPC Europe Foundation (congratulations are in place!), he is now officially listed at the Chamber of Commerce. 23:37
And of course we have been drinking several types of beer with him... but that's not the point, even though it was very nice.
woolfy I've shown him the book jnthn made about NQP (jnthn++) and the booklet that came with the course in april this year about Perl 6 by Damian Conway. elbeho was very interested and decided he has to look into Perl 6 and try it out. 23:38
Yay.
vendethiel woolfy: mind asking him if he has a timeline for fpw2014 videos :P ?
woolfy (now off to bed)(good night all)
vendethiel: I will do that when we all wake up... (hips) 23:39
vendethiel thankies
psch just realized that (*****) is a valid closure 23:40
std: say (*****)(<1 2 3>, 4) 23:41
camelia std 0f2049c: OUTPUT«ok 00:01 128m␤»
psch note sure if it's ** ** * or ** * **; not even sure it has to start with **
cognominal elbeho++ woolfy++ lizmat++
psch Whatever^n...
jnthn psch: LTM menas it'll be ** if there's a choice 23:42
psch jnthn: right. that includes the second set? i.e. if i wanted to multiple multidimensional Whatevers i'd have to have the parens? (disregarding the obvious cause for confusion without any spaces) 23:43
err, spaces
not parens
jnthn Oh, I forgot about ** 23:44
Hmm :)
m: say (*****)(<1 2 3>, 4)
camelia rakudo-moar 1f05e1: OUTPUT«===SORRY!=== Error while compiling /tmp/qoXdp7mKab␤HyperWhatever (**) not yet implemented. Sorry. ␤at /tmp/qoXdp7mKab:1␤------> say (**⏏***)(<1 2 3>, 4)␤»
psch as an aside, j-spectest is somewhere in t/integration
jnthn Yeah, LTM wants ** as a term first.
vendethiel Hyper* power hyper* ?
psch vendethiel: right, (******) is valid too 23:45
vendethiel
.oO( when I thought I couldn't care less of whatever, the hyper one came to be ... )
psch there's the line noise again... :)
vendethiel you misspelled "perl" :P
jnthn Whatever can be really nice 23:47
vendethiel it indeed can. I don't quite remember what hyper is supposed to do ...
jnthn .map(*.name), .grep(*.type eq 'stout'), etc. 23:47
jnthn No, me either... :) 23:47
psch S02: "[HyperWhatever] is generally understood to be a multidimension form of * when that makes sense." 23:48
i don't know how to figure out the line numbers :)
vendethiel I love the "when that makes sense" part :D.
"put it there, if it makes sense"
flussence
.oO( "**()" means "generate the code I'm thinking of" )
 23:49
psch the examples explain it away as "-> *@x { map { $x { ... } }"
where '...' is not the operator with the same name :)
flussence "..." means "someone else, write the code I'm thinking of" :) 23:50
vendethiel "..." has a pretty convoluted meaning ...
jnthn The spec saying things like, "when it makes sense", is what makes working on a Perl 6 compiler so easy... :P 23:51
vendethiel "so, uh, about the sense you were talking about earlier"... 23:52
vendethiel imagines TimToady with a greyed beard thinking of riddles to add in the spec :P
jnthn "I decided it makes sense on a Tuesday"
psch "don't care anymore--just shoot me"
i have to say though, it seems rather understandable to me
the HyperWhatever as it's spec'd, that is 23:53
of course trying to use it might show more
jnthn We just don't have so much multi-dim stuff around yet...
It's one of the things I'm getting tired of us not having, though... :)
vendethiel and when jnthn gets tired of something ... the ++ fear ! 23:54
jnthn Well, it means it's gradually climbing its way to the top of my "stuff I should work on" list
vendethiel hopes that list is shrinking 23:55
vendethiel doesn't do much to help it shrink, though ... 23:55
psch gist.github.com/peschwa/2b05db181f37ea55d458 # j-spectest results 23:56
it doesn't look good, but i don't know if i can savely blame 32bit and -Xmx1800 23:57
*safely
jnthn psch: I doubt those fails are related to your patch 23:58
psch the random "Missing or wrong version of dependency" really surprised me
jnthn Yes, that is odd
jnthn Though the test probably is shelling out somewhere... 23:59
But still

Powered by Cro and the Raku® Programming Language.

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

Thank you!