Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2014-11-19

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:10 Augustus how do you add the current directory to the include path?
00:11 Augustus nm it's the usual -I sry
00:17 Augustus is there an easy way to have your pm file export all of its symbols when you use it?
00:18 BenGoldberg joined #perl6
00:29 Augustus is there a reason why I have to put "is export" on my grammar but not on my actions class?
00:38 Exodist joined #perl6
00:45 Mouq joined #perl6
00:52 peterete1 I can't imagine anyone was following along at home, but my supervisor has talked me out of my CPAN6 proposal, so I won't be working on that this year
00:52 Mouq Augustus: Depending on what you want, something like `sub EXPORT (|) { OUR:: }` may be what you want?
00:54 Mouq peterete1: Aww. why?
00:55 peterete1 I think he could see my heart wasn't in it, and it felt like a moving target considering the active and not-so-active work on it
00:57 Mouq peteretep: Yeah, I could see that as being the case. Do you have any idea what you're planning on instead?
00:58 peteretep I'll be writing a specialized document similarity engine and DB for job descriptions
01:00 Mouq Nice!
01:02 peteretep Just trying to pick between Go and Rust
01:03 bjz perturbation: What is your project?
01:04 bjz woops - peteretep
01:04 Mouq I really like what Rust aspires to, though the strictness was tough for a newb (me) to write in
01:05 peteretep bjz: The specialized document similarity engine I was talking about
01:05 peteretep Mouq: I'm a big fan of HAskell, so I find it quite exciting
01:05 * bjz apologises for butting in - he has a notification on
01:06 peteretep I am also on Team Mozilla, rather than Team Google, philosophically, which makes Rust more exciting
01:06 bjz peteretep: you would probably get more out of Rust then - the type system is much more similar to Haskell
01:06 peteretep bjz: That's the conclusion I am coming to
01:07 peteretep The "terrible syntax" looks no worse than Erlang or Perl 5
01:07 bjz heh
01:07 peteretep let me take that back. Erlang is HORRIBLE, but you do get over it
01:07 adu joined #perl6
01:07 peteretep You become mindful of it - you note the syntax is still horrible whenever yo uuse it, but you cease to get upset about it
01:08 bjz yeah I don't like the syntax either - it has got a great deal better in the last year, but it could be worse. it tends to recede into the background if the semantics are good enough
01:08 bjz like Erlang
01:09 perturbation after mumps, everything looks clean... guess what 'f p=2,3:2 s q=1 x "f f=3:2 q:f*f>p!'q  s q=p#f" w:q p,?$x\8+1*8' does?
01:09 KCL_ joined #perl6
01:10 perturbation one-liner for primes, formatted for 80-col terminal... nifty, but kinda terrible too..
01:10 bjz Mouq: you have to treat it differently than a dynamic language where the interpeter is like the clay you mould. In a statically the compiler is a far more active participant in your work, and you are having a conversation with it to work towards figuring out what you actually meant to say.
01:10 peteretep perturbation: Nothing beats APL
01:11 bjz Mouq: And the rust compiler has much more to say about memory management than almost any other language out there, so it can be a little off-putting at first.
01:12 Augustus If I have a QAST object at runtime, is there a simple way to eval it?
01:14 bjz Mouq: either way isn't better or worse, but for stuff where you are working intimately with memory, you want a friend (ie. rustc) to be watching your back. :) anyway,
01:15 bjz perturbation: your project sounds cool. we are very friendly over at irc.mozilla.org #rust - I don't want to blabber with off topic stuff here.
01:16 perturbation bjz: I think you mean peteretep again :)
01:16 bjz arrrrgh
01:16 peteretep :-)
01:16 bjz sorry
01:16 peteretep I think so too
01:16 * peteretep needs to part some channels, and will join
01:17 Mouq bjz: :) I wasn't complaining, per se -- I simply never got into it. I think it's a super cool project, and I plan to use it at some point :)
01:20 bjz Mouq: no worries! as I say, we are trying hard to be a friendly community, so if you have trouble we would love to help.
01:27 Augustus what is the perl6 equivalent of eval?
01:28 BenGoldberg EVAL
01:29 Augustus ah ok all caps, thanks. Anybody know where I can find the EVAL implementation?
01:30 ugexe https://github.com/rakudo/rakudo/bl​ob/b6b1b7cb2277a5801abebb94e2240f99​ecaed8f2/src/core/control.pm#L204
01:33 Augustus awesome thx
01:37 colomon FROGGS is probably in bed, but the smoke run came out much better second time around
01:38 dayangkun joined #perl6
01:41 [Coke] Did we lose dalek again?
01:59 xinming_ joined #perl6
02:06 BenGoldberg .seen dalex
02:06 yoleaux I haven't seen dalex around.
02:06 BenGoldberg .seen dalek
02:06 yoleaux I saw dalek 18 Nov 2014 06:54Z in #perl6: <dalek> rakudo/nom: review: https://github.com/rakudo/rakudo/commit/f2ffb9a384
02:11 jimmy_ joined #perl6
02:12 [Coke] .time
02:27 adu joined #perl6
02:28 adu joined #perl6
02:29 Augustus currently, does nqp work as a slang? Can you mix and match nqp and Perl6 in the same file, or do you have to put all your nqp in a separate file?
02:30 Augustus also, if I just put NQP libs directory on the path, I can just use nqp libs directly right?
02:34 ugexe m: nqp::say('hello'); say 'world';
02:34 camelia rakudo-moar 93c2b4: OUTPUT«world␤hello␤»
02:34 ugexe nqp: nqp::say('hello'); say 'world';
02:34 camelia nqp-moarvm: OUTPUT«Confused at line 2, near "say 'world"␤   at gen/moar/stage2/NQPHLL.nqp:485  (/home/camelia/rakudo-inst-2/languag​es/nqp/lib/NQPHLL.moarvm:panic:105)␤ from gen/moar/stage2/NQP.nqp:913  (/home/camelia/rakudo-inst-2/language​s/nqp/lib/nqp.moarvm:comp_unit:872)␤ f…»
02:34 camelia ..nqp-jvm: OUTPUT«(signal ABRT)»
02:34 camelia ..nqp-parrot: OUTPUT«Confused at line 2, near "say 'world"␤current instr.: 'panic' pc 15673 (gen/parrot/stage2/NQPHLL.pir:5731) (gen/parrot/stage2/NQPHLL.nqp:425)␤»
02:37 ugexe m: nqp::say('hello'); say 'world'; say 'X';
02:37 camelia rakudo-moar 93c2b4: OUTPUT«world␤X␤hello␤»
02:38 ugexe i thought i was answering your question but now im asking one myself
02:40 Augustus let me be more specific. I'm trying to create ASTs. Do I have to do that from pure nqp? I can put the QAST module in my path, but when I load it from perl6 it seems the symbols aren't exported.
02:41 Augustus When I load it from nqp, it works fine.
02:41 Augustus can I pass a parameter to the use statement to tell it to import all the symbols?
02:45 JimmyZ m: use QAST:from<nqp>;
02:45 camelia rakudo-moar 93c2b4: OUTPUT«===SORRY!===␤Do not know how to load code from nqp␤»
02:45 JimmyZ m: use QAST:from<NQP>;
02:45 camelia rakudo-moar 93c2b4: ( no output )
02:45 KCL joined #perl6
02:46 Augustus JimmyZ: Thanks, that seems to work, at least no immediate errors pop.
02:50 Mouq Augustus: There's no need to put them in-path unless you wrote them
02:52 Augustus Mouq: I see now, I wasn't aware of the :from attribute to pull from the other language
02:56 telex joined #perl6
03:06 chenryn joined #perl6
03:12 kjs_ joined #perl6
03:17 zakharyas joined #perl6
03:30 Augustus when I pass a QAST object from perl6 code back into NQP I am getting "This type does not support positional operations" errors. Do I have to do something special for NQP objects to pass through my p6 code unscathed?
03:32 Augustus also, if I use := in my p6 code rather than plain =, then when I pass the object back to NQP it doesn't recognize the object's class. Can anyone explain?
03:36 Mso150 joined #perl6
03:39 JimmyZ NQP doesn't support  =
03:40 JimmyZ and NQP knowns nothing about Perl 6
03:40 JimmyZ You can't pass p6 object to nqp, methinks
03:41 Augustus i'm not trying to pass a p6 object to nqp, i'm tring to pass fetch an nqp object in p6 and pass it back to nqp
03:41 JimmyZ I don't know why you want to use nqp in perl6 thought
03:42 JimmyZ *though
03:42 Augustus i'm just trying to generate QAST and execute it without having to learn NQP
03:43 Augustus in general, we just be able to create grammars and actions in perl6 not just nqp
03:43 Augustus just ==> should
03:44 ugexe https://github.com/tony-o/perl6-slan​g-sql/blob/master/lib/Slang/SQL.pm6
03:44 ugexe its not *all* nqp
03:45 JimmyZ Augustus: Do you mean something like this? https://github.com/moritz/js​on/tree/master/lib/JSON/Tiny
03:45 JimmyZ Slang/SQL.pm6 is a different example
03:46 Augustus well right now for testing I'm trying to just manually create QAST and execute it
03:46 Augustus I'm doing my $comp_unit = QAST::CompUnit.new(...); etc
03:47 Augustus and then getting the backend and trying to get it to compile and run that
03:47 tinyblak joined #perl6
03:48 Augustus while EVAL will happily take source code and run it, it seems oddly difficult to EVAL starting from QAST
03:49 JimmyZ I would suggest use NQP, since NQP is Not Quite Perl 6 ;)
03:50 Augustus I understand that may get me over my current hump but it obviously not correct
03:51 Augustus it is possible to pass objects back and forth with Java or C, so it must be possible to round trip objects with NQP
03:51 JimmyZ AFAIK, there is some reason why rakudo is not written in rakudo
03:53 ugexe have you looked at qast.t?
03:53 Augustus i don't know what a .t file even is, so no
03:53 ugexe https://github.com/perl6/nqp/blob/00b94e415741​a014fb507b7d4f033d383901fca8/t/qast/01-qast.t
03:55 Augustus ah .t is a test file and it appears that the nqp compiler has a :from('ast') option
03:55 Augustus that may solve my problem without having to roundtrip an object
03:57 Mouq Augustus: Are you trying to create a Slang or a new language?
03:58 Mouq i.e, are you just trying to a "simple" augementation of the Perl 6 grammar, or are you doing something else entirely?
03:59 Augustus right now I'm just testing QAST, what I do with later is secondary
03:59 Augustus So I'm trying to manually build up a QAST structure and then execute it
04:00 Augustus the test case ugexe showed me might get me where I need to go except now nqp::getcomp('nqp') is returning null
04:00 Mouq Augustus: Ah, yeah, I imagine that would be a pain in the butt
04:00 Mouq do nqp::getcomp('perl6') or whatever it is..
04:01 ugexe m: say nqp::getcomp('nqp');
04:01 camelia rakudo-moar 93c2b4: OUTPUT«(Mu)␤»
04:01 Mouq m: say nqp::getcomp('perl6');
04:01 camelia rakudo-moar 93c2b4: OUTPUT«No such method 'gist' for invocant of type 'Perl6::Compiler'␤  in sub say at src/gen/m-CORE.setting:16539␤  in block <unit> at /tmp/jaNNTba0XW:1␤␤»
04:01 Mouq ^^
04:02 ugexe nqp: say(nqp::getcomp('nqp'));
04:02 camelia nqp-parrot: OUTPUT«NQP::Compiler<-1070294649291487057>␤»
04:02 camelia ..nqp-moarvm: OUTPUT«cannot stringify this␤   at gen/moar/stage2/NQPCORE.setting:682  (/home/camelia/rakudo-inst-2/languages/n​qp/lib/NQPCORE.setting.moarvm:print:13)␤ from gen/moar/stage2/NQPCORE.setting:688  (/home/camelia/rakudo-inst-2/language​s/nqp/lib/NQPCORE.setting.moarvm:sa…»
04:02 camelia ..nqp-jvm: OUTPUT«(signal ABRT)»
04:02 Mouq m: say nqp::getcomp('perl6').^name;
04:02 camelia rakudo-moar 93c2b4: OUTPUT«Perl6::Compiler␤»
04:02 Mouq nqp: my $c := nqp::getcomp('perl6'); say($c.HOW.name($c));
04:02 camelia nqp-parrot: OUTPUT«Can only use get_how on a SixModelObject␤current instr.: '<mainline>' pc 46 ((file unknown):41) (/tmp/tmpfile:1)␤»
04:02 camelia ..nqp-moarvm: OUTPUT«VMNull␤»
04:02 camelia ..nqp-jvm: OUTPUT«(signal ABRT)»
04:02 Mouq oop
04:03 Mouq nqp: my $c := nqp::getcomp('nqp'); say($c.HOW.name($c));
04:03 camelia nqp-jvm: OUTPUT«(signal ABRT)»
04:03 camelia ..nqp-{moarvm,parrot}: OUTPUT«NQP::Compiler␤»
04:04 rmgk_ joined #perl6
04:04 Augustus it looks like you can't get the nqp compiler from inside perl6 code
04:04 ugexe nqp: my $c := nqp::getcomp('QAST'); say($c.HOW.name($c));
04:04 camelia nqp-jvm: OUTPUT«(signal ABRT)»
04:04 camelia ..nqp-moarvm: OUTPUT«QAST::MASTCompiler␤»
04:04 camelia ..nqp-parrot: OUTPUT«QAST::Compiler␤»
04:04 Augustus i wonder if the perl6 compiler supports the :from('ast') command
04:06 kurahaupo joined #perl6
04:06 Mouq Augustus: I'm fairly certain it does. That's implemented in HLL::Compiler, which both NQP::Compiler and Perl6::Compiler inherit from
04:06 Augustus hmm, still getting the "type does not support positional operations" error
04:07 Mouq Can you gist/nopaste what you're doing?
04:08 Augustus sure. also I just changed the = back to := and now I get "Optimizer could not find UNIT"
04:10 Augustus Here's a gist: https://gist.github.com/ano​nymous/82bed5ca4d38288dbd1b
04:10 Augustus it's pretty simple
04:12 Augustus it does seem like when I switch everything to nqp it stops complaining
04:13 Mouq You're breaking assumptions the optimizer has about about Perl 6... getcomp('anything') probably isn't the way to go if you're just evaling random QAST::Ops, since they're agnostic to language...
04:13 Augustus i don't really want to
04:14 Augustus but when I tried going to the backend to execute directly it was giving me the other error from earlier
04:14 Augustus it won't recognize the [] notation
04:14 Augustus for accessing QAST child nodes
04:16 Augustus also there doesn't seem to be a way to get executable code from AST in a backend agnostic way
04:17 Mouq m: use QAST:from<NQP>; say nqp::atpos(QAST::Op.new( :op<call>, :name<say>, QAST::IVal.new(42) ), 0).^name
04:17 camelia rakudo-moar 93c2b4: OUTPUT«Too many positionals passed; expected 1 argument but got 2␤  in any new at gen/moar/stage2/QASTNode.nqp:275␤  in block <unit> at /tmp/ukBqIutY5r:1␤␤»
04:17 Mouq m: use QAST:from<NQP>; say nqp::atpos(QAST::Op.new( :op<call>, :name<say>, QAST::IVal.new(:value(42)) ), 0).^name
04:17 camelia rakudo-moar 93c2b4: OUTPUT«QAST::IVal␤»
04:17 Mouq That's the best we have at the moment
04:18 Augustus I'm not trying to access the children, the compiler is
04:18 Augustus and the compiler is written in NQP trying to access children of an NQP object
04:19 Augustus but somehow by passing it through a p6 variable in between, this breaks
04:19 Augustus how do you tell p6 to treat an object as a foreign, opaque object and not mess with it?
04:20 Augustus i'm guessing it's perl6-ifying the object when I assign to a variable in p6 code
04:20 Augustus which breaks it when I pass it back to nqp code
04:25 ugexe maybe you mean unboxing?
04:25 ugexe https://github.com/perl6/nqp/blo​b/master/docs/ops.markdown#unbox
04:27 kurahaupo joined #perl6
04:27 Augustus hmm that only applies to integers, floats and strings
04:27 Augustus i don't know what kind of a thing an NQP object is
04:30 Mouq Augustus: Can I see what you're talking about?
04:30 * Mouq is very confused
04:31 kaare_ joined #perl6
04:34 Mouq Well, maybe some other time :P
04:38 Augustus Sorry I missed you Mouq. If anybody else is willing to take a look, I have made another gist. It runs fine if you use the nqp runtime but errors out if you use perl6
04:38 Augustus https://gist.github.com/ano​nymous/5c9be05d7e2ac123ccb0
04:39 Augustus in order to run from nqp you of course have to remove the :from<NQP> on the use QAST statement
04:39 Augustus I think this should illustrate the problem in a very simple way
04:40 kurahaupo joined #perl6
04:44 ugexe nqp: my $x := CompUnit.new; $x[0]; $x[0][0]
04:44 camelia nqp-jvm: OUTPUT«(signal ABRT)»
04:44 camelia ..nqp-{moarvm,parrot}: ( no output )
04:45 ugexe m: my $x := CompUnit.new; $x[0]; $x[0][0]
04:45 camelia rakudo-moar 93c2b4: OUTPUT«Too few positionals passed; expected 2 arguments but got 1␤  in method new at src/gen/m-CORE.setting:23663␤  in block <unit> at /tmp/jI0slltWDA:1␤␤»
04:46 Augustus maybe try with a different node than comp unit which requires children
04:46 Augustus or actually pass the required params
04:46 Augustus but obviously with no children [0] will probably fail, but perhaps a different error
04:48 ugexe m: my $x = 1; say $x[0];
04:49 camelia rakudo-moar 93c2b4: OUTPUT«1␤»
04:50 Augustus m: use QAST:from<NQP>; my $x := QAST.Block.new(QAST::IVal.new( :value(42) )); say $x[0];
04:50 camelia rakudo-moar 93c2b4: OUTPUT«===SORRY!===␤Object of type QAST in QAST::WVal, but not in SC␤»
04:51 Mouq joined #perl6
04:51 Mouq o/, sorry, but I had to run off real quick
04:52 Mouq Augustus: like I said, you have to use nqp::atpos
04:52 Mouq It's not a matter of the object being p6-ified as it is that [] translates into something different in NQP and Perl 6
04:53 Augustus well there must be a way to force nqp code to use nqp semantics
04:53 Mouq However, if you want to get around that, (and I don't guarantee that this won't break things), you can try defining `multi postcircumfix:<[ ]> (Mu \k, $v) { nqp::atpos(k, $v) }`
04:54 Augustus it's obviously a problem if nqp code breaks when called from perl6 code
04:56 Mouq NQP wasn't designed to be used in Perl 6 code :P
04:57 ugexe you can use nqp::atpos in perl6
04:58 Augustus ugexe: the objective is for nqp code to preserve nqp semantics regardless of whether it is invoked by a perl6 runtime or an nqp runtime
04:58 Augustus in other words we want semantics baked in at compile time
04:59 Augustus having the runtime choose between atpos and operator<[]> means that code compiled with nqp is still breaking when invoked by the perl6 runtime
05:00 dj_goku joined #perl6
05:01 Augustus either nqp compiled code should work when called from perl6 100% or it should be prohibited
05:02 gtodd is there an  Inline::NQP ?  :-)
05:03 Mouq In Perl 6, calling {foo bar}[baz] is always going to call &postcircumfix:<[ ]>. We don't really have the semantics in place for dealing with NQP objects because NQP is a compiler-detail -- it's my understanding that NQP is allowed for those who wish to muck about with internals within Perl 6, but it is never the ideal in Perl 6 code itself
05:04 tinyblak_ joined #perl6
05:06 Mouq In this case, "slangs," and ast-building still need a massive overhaul that's expected to have something to do with the macro one. You're right, a user shouldn't have to do contortions to create an AST or Slang, but there isn't really a high-level way around it implemented right now, unfortunately
05:06 Augustus again, I just want to reiterate that we are talking about the perl6 runtime messing up precompiled nqp code compiled by the nqp compiler
05:06 Mouq Sorry, I'm a very rant-y mood
05:07 Augustus no please, rant away, I just want to make sure you understand the issue at hand
05:08 Augustus everybody keeps thinking I want to mess with NQP internals from p6
05:08 Augustus I do not
05:09 Augustus the point here is that the compiler should dictate semantics that are stable regardless of runtime
05:09 Augustus just like it doesn't matter if you run on MOAR or JVM
05:09 Augustus once you compile your NQP code it should have stable semantics
05:09 Mouq It does
05:09 Augustus that's the point it doesn't
05:09 Mouq The object itself doesn't change
05:09 Mouq Well
05:09 Mouq okay
05:09 Augustus invoking nqp code from perl6 is breaking
05:10 Augustus because nqp is not getting the atpos behavior
05:10 Mouq The syntax changes
05:10 Mouq Because you're writing in a different language
05:10 Augustus PRECOMPILED CODE!!!!
05:10 Mouq so?
05:11 Augustus once I compile C code its semantics don't change if I invoke it from C++, java, python, or anything
05:11 Mouq The code from NQP isn't changing.
05:11 Augustus exactly
05:12 Augustus but the perl6 runtime is messing up the nqp runtime
05:12 ugexe are you saying nqp::atpos does not work as expected in p6 runtime? or something else specifically
05:13 Augustus well nqp's operator<[]> does not work from the p6 runtime
05:13 Augustus the nqp code is not calling atpos directly
05:13 Mouq How? What I get is that it's messing it up because you expect [] to automatically mean nqp::atpos even though it means &postcircumfix<[ ]>. Just like `:=` in NQP and Perl 6, they share the same symbols but mean different things
05:14 Augustus but from what you guys are saying that's going on behind the scenes
05:16 Mouq Currently, we don't have a candidate for &postcircumfix:<[ ]> defined for NQP objects, however I think it would be very convenient if we did. `multi postcircumfix:<[ ]> (Mu \k, $v) { nqp::atpos(k, $v) }` fixes that, although it might break other things so I'd be weary using it.
05:16 Augustus let me be more specific. I'm not sure how you guys link to source code but take a look at nqp/gen/moar/stage2/QAST.nqp line 4433
05:17 Augustus if +@($cu) != 1 || !nqp::istype($cu[0], QAST::Block) {
05:17 Augustus where $cu is a QAST::CompUnit
05:17 Augustus the $cu[0] fails when you are using the p6 runtime
05:18 tinyblak joined #perl6
05:18 TimToady yes, because .[] is a different operator in a different language
05:18 TimToady it's not a method
05:18 Augustus yes so what I'm saying is the p6 runtime shouldn't stomp all over the nqp runtime
05:19 TimToady it's not the runtime, it's the language definition, so it's lexically determined, not determined by runtime dispatch
05:19 Augustus it is the runtime
05:19 Augustus it works fine when using the nqp runtime
05:19 TimToady it works fine when using the nqp *languge*
05:20 TimToady *uage
05:20 Augustus no i mean when I use perl6-m it breaks
05:20 Augustus and when I use nqp-m it works
05:20 TimToady because now you're using a different language
05:20 TimToady which dispatches to a different place
05:20 Mouq If you took an object from Perl 6 into C and tried to do foo[123] on it, it would fail
05:20 TimToady same deal
05:20 Augustus right so you're saying that the p6 runtime is stomping on the nqp runtime
05:21 TimToady no
05:21 TimToady that's like saying my German is stomping on my French
05:21 Augustus Mouq: but if you took a C pointer into perl and passed it back to to C code it would not break
05:21 Augustus because perl doesn't stomp on the C runtime
05:21 Augustus of course in C there's no dynamic dispatch but you understand
05:21 Mouq Augustus: And if you take a NQP object and pass it back to NQP code it does not break
05:22 Augustus indeed it does
05:22 TimToady but you aren't passing $cu back to nqp
05:22 Augustus that is the whole point of this conversation
05:22 TimToady so the argument is oot
05:22 TimToady *moot
05:22 Augustus no, I am writing my $x = QAST.CompUnit.new(...); then passing $x into precompiled code written in NQP
05:22 TimToady you're passing it to the P6 .[] operator and expecting it to make sense
05:23 Augustus not unless the nqp compiler called the p6 .[]
05:23 Augustus the code was compiled by the nqp compiler
05:23 TimToady nqp knows nothing about p6
05:23 Augustus it is an nqp file after all
05:24 Augustus I think you guys are having a very difficult time understanding what's going on
05:24 Mouq I really don't
05:24 Augustus there is an nqp file
05:24 Augustus nqp/gen/moar/stage2/QAST.nqp
05:24 Augustus it has been compiled by the nqp compiler
05:24 Augustus it now exists as precompiled bytecode
05:25 Augustus hmmm wait a second
05:25 Augustus when I use a module it's loading the precompiled bytecode right?
05:25 Augustus the problem may be the use ...:from<NQP>
05:26 chenryn joined #perl6
05:26 Augustus when you use ...:from<NQP> is it recompiling it with p6 semantics?
05:26 Augustus i am taking it as a given that using a module is using the precompiled version just like in Java
05:27 Mouq Regardless, that module is parsed and executed correctly. The error doesn't come from the module.
05:27 Augustus well that is where the error occurs
05:27 TimToady how are you creating the $cu object on the p6 end?
05:27 Augustus there's only 1 way
05:27 Augustus QAST::CompUnit.new
05:27 TimToady how are you adding bits to it?
05:28 Augustus Just passing stuff to new
05:28 Mouq (from earlier: https://gist.github.com/ano​nymous/5c9be05d7e2ac123ccb0 )
05:29 Mouq Oi, bad link. https://gist.github.com/ano​nymous/5c9be05d7e2ac123ccb0
05:29 Augustus well and you need the earlier one
05:29 Augustus in that example I am indeed using p6 .[]
05:29 Augustus i was trying to simplify for you
05:30 Augustus https://gist.github.com/ano​nymous/82bed5ca4d38288dbd1b
05:31 Mouq That one fails for a completely different reason though, I though. That one fails because you're constructing an AST tree the Perl 6 optimizer doesn't see as valid and saying it comes from Perl 6...
05:31 Augustus hmm and that's still not a good example because it runs into the optimizer error, which is a different problem
05:31 Mouq *I thought
05:31 Augustus yes
05:33 TimToady going the other way, it would be really easy to call P6's comma operator by accident, which is going to make a P6 list or parcel rather than an nqp::list, which would prevent subscripting in nqp
05:34 Augustus Ok, one more time, this demonstrates the original error I had
05:34 Augustus https://gist.github.com/ano​nymous/8f471e531cb4b5f4d05a
05:35 Augustus so you're suggesting that passing parameters from p6 to nqp is tricky in some way?
05:36 Augustus this could make sense, if nqp can't unpack the parameters then everything will break
05:36 TimToady it's also possible you're getting your nqp objects wrapped in containers by using = rather than :=
05:36 TimToady nqp won't know what to do with scalar containers
05:37 TimToady (P6 automatically derefs them for you most of the time)
05:37 Augustus well if I use := instead (I tried this) then NQP doesn't recognize the object as being a CompUnit at all
05:37 * Mouq would like to note he's been frustrated and stressed recently, and it likely a tad more argumentative than he should be
05:37 Augustus :)
05:37 Augustus or well I guess :(
05:37 TimToady NO YOU're NOT!!!
05:37 TimToady :)
05:37 Augustus what is the sympathy emoticon lol
05:38 Augustus when I use := I get "Top-level MAST node must be a CompUnit"
05:38 TimToady hugme: hug Mouq and Augustus
05:38 * hugme hugs Mouq and
05:38 TimToady hugme: hug Mouq and Augustus
05:38 * hugme hugs Mouq and Augustus
05:38 TimToady there we go
05:38 Augustus nice
05:38 Augustus hugme: Mouq
05:38 Augustus hmm did I do it wrong?
05:38 TimToady this sentence no verb
05:39 Augustus hugme: hug Mouq
05:39 * hugme hugs Mouq
05:39 Augustus there we go lol
05:40 TimToady you have to use an unbreakable space to get more than one word hugged though
05:40 Augustus lol
05:41 TimToady well, perhaps $comp_unit needs to be = but $qast needs to be raw
05:41 Augustus hmm I'm guessing then I'm mistaken thinking that the = sign worked better
05:41 TimToady or something like that
05:41 Augustus i like your idea of messing up the parameter passing
05:41 Mouq :) thanks. It seems pretty strange that you'd be using .mast: on a QAST...
05:41 TimToady nqp and qast are just very, very picky, without being terribly informative
05:42 Augustus .mast converts qast to mast
05:42 Augustus i would like a backend agnostic way of doing that of course
05:42 TimToady they were originally just a tool for the original implementors to get rakudo implemented in, not something that has been productized at all
05:42 Augustus but from the compiler stages theres no real way to tell where to jump in
05:42 Augustus and I think people get that, I'm happy to dig through source :)
05:43 TimToady that's just how nqp is currently, somewhere on the other side of the bleeding edge
05:43 TimToady it's just a matter of how anemic you're willing to end up :)
05:43 Augustus for right now I'm fine hard coding for MOAR because that's a side issue
05:44 Augustus if i can just pass ASTs in and successfully compile them I'll be happy
05:44 TimToady eventually we'd like to have a better higher level AST interface, but it's just not there yet
05:45 TimToady well, it's possible, but it's just a complete pain in the butt to get right currently
05:45 Augustus hmm, there's probably a much fafster way to get to the bottom of this, but I've never used the p6 debugger
05:45 Augustus it should be possible to just trace in to the nqp code
05:45 Augustus and see whether the parameters got passed correcty right?
05:46 kaare_ joined #perl6
05:46 * TimToady doesn't know how well the debugger does with nqp, having seen only p6-level demos
05:46 kurahaupo joined #perl6
05:46 TimToady but both languages are based on 6model, which helps
05:47 Augustus cuz otherwise I'd have to modify and recompile nqp which seems a dubious undertaking
05:47 TimToady what you don't have in nqp is a lot of introspective methods directly accessible
05:48 TimToady so most of the reflection must be done through 6model
05:48 Augustus is there anything I can do to be extra, extra careful that my parameters are being passed in the appropriate nqp format?
05:48 Augustus is there any way to eliminate that as a possible source of the problem?
05:49 Augustus like create an nqp parcel by hand or something
05:49 * TimToady is not an expert on those bits of it
05:50 Augustus hmm d'oh I should just create a simple package that takes an AST
05:50 Augustus i don't need to modify the compiler just make my own package
05:50 Augustus what's the command line switch to output a compiled module?
05:51 kjs_ joined #perl6
05:51 TimToady I dunno, my current make is spitting out too much debug info, and the command scrolled away...
05:52 TimToady but however the standard make does it
05:52 TimToady for like Test.pm or so
05:53 kurahaupo joined #perl6
05:55 Mouq Augustus: The problem, after you either bind to $comp_unit or use \comp_unit, is that $mast needs to be either bound to or be \mast
05:58 Augustus maybe i didn't understand correctly, i tried changing my $mast = $backend.mast($comp_unit); to my $mast := $backend.mast($comp_unit); and I also tried replacing $mast with \mast
05:59 Augustus neither made a difference
05:59 Augustus oh but I need to do the same with comp_unit you're suggesting
06:01 Augustus hmm now I'm just getting "Cannot invoke comp unit object" when I try and run the compiled code
06:02 Mouq Because it's not a code object, so you can't just () it
06:02 smls joined #perl6
06:03 Augustus the problem apparently that the compiling function is just passing back the comp unit rather than the compiled code
06:03 Augustus I was basing this off of how EVAL works
06:04 Augustus maybe I do need to invoke the moar stage even though it appeared to be a non-op
06:05 Augustus no difference
06:06 Mouq Oh, that's ... curious
06:07 Mouq Try $backend.compunit_mainline($compiled)()
06:07 Mouq That's the missing peice
06:07 Mouq piece?
06:08 kjs_ joined #perl6
06:09 smls good morning perl 6
06:09 Augustus well no errors!
06:09 Augustus now i just have to change the ast to actually print hello world or whatever
06:09 Augustus to see that it actually works
06:11 smls I know I can convert a decimal number to base 16 using «.base(16)» or «.fmt("%02x")», but how can I do the reverse?
06:11 smls ...i.e. get the decimal representation *from* a base-16 number that is stored as a string
06:11 Mouq :16<foo>
06:13 smls thanks!
06:14 moritz \o
06:14 moritz or :16($variable)
06:16 smls m: <00 aa ff>.map({:16($_)}).say
06:16 camelia rakudo-moar 93c2b4: OUTPUT«0 170 255␤»
06:17 smls ^^ the bracket density is getting a little high there for my taste
06:18 moritz m: say :16(*)
06:18 camelia rakudo-moar 93c2b4: OUTPUT«Cannot call 'UNBASE'; none of these signatures match:␤:(Int:D $base, Cool:D $num)␤:(Int:D $base, Str:D $str)␤  in block <unit> at /tmp/uGSpL2tjpR:1␤␤»
06:18 smls can the :16() construct be hypere'd like a method call to avoid the map?
06:18 anaeem1 joined #perl6
06:18 moritz smls: I don't think so
06:18 moritz smls: but we could think about giving it list input in the first place
06:19 moritz m: say :16( <00 aa ff> )
06:19 camelia rakudo-moar 93c2b4: OUTPUT«You have confused the number 00 aa ff with the textual representation "00 aa ff";␤if you wanted to render the number in the given base, use $number.base($radix)␤  in block <unit> at /tmp/w80JIAa6hT:1␤␤»
06:21 Augustus ok, finally, it does appear to be working! So it was an issue with = vs :=
06:22 Augustus You have to be meticulous to use := everywhere. I guess earlier I did something else wrong and convinced myself to go with =
06:24 JimmyZ moritz: http://irclog.perlgeek.de/​perl6/2014-11-18#i_9680184 # in case you missed it
06:25 moritz JimmyZ: I didn't miss it, thanks
06:26 moritz JimmyZ: I'm just not thrilled to learn two technologies :-)
06:26 chenryn joined #perl6
06:26 JimmyZ :)
06:28 Augustus ok new question. I can :op('say') hello world, but it breaks if I try :op('call'), :name('&say') it breaks
06:28 smls joined #perl6
06:28 JimmyZ I can help build kvm virtual machine, but the network between host and guest sometimes misses after the host reboot
06:29 JimmyZ and needs a shell or something to re-create the network
06:29 moritz Augustus: you have to do a lookup
06:29 moritz Augustus: so :op<call>, QAST::Var.new(:name<&say>)
06:30 moritz possibly with a scope too, let me look that up
06:30 moritz yes, the QAST::Var needs a :scope<lexical> too
06:31 Augustus ok let me try that
06:34 Augustus hmm that's still not working, can you see what I'm doing wrong here:
06:34 Augustus my $qast := QAST::Block.new(QAST::Op.new( :op<call>, QAST::Var.new(:name<&say>, :scope<lexical>), QAST::SVal.new(:value('Hello World'))));
06:35 moritz Augustus: what error do you get?
06:35 Augustus it accepts that but when I run "cannot invoke null object"
06:35 Augustus so the QAST::Block constructs ok but fails during compilation to byte code I guess
06:36 Augustus with no indication of what failed
06:36 anaeem1 joined #perl6
06:36 moritz presumably the QAST::Var returned a null
06:37 moritz Augustus: I haven't backlogged yet; where is your code? is it somewhere in rakudo?
06:38 Augustus ok here's one that essentially works
06:38 Augustus https://gist.github.com/ano​nymous/c77f22162dda98efc968
06:39 Augustus the previous stuff all had other problems
06:42 ugexe nqp: my $qast := QAST::Block.new(QAST::Op.new( :op<call>,:name<&say>,QAST::SVal.new(:value('Hello World'))));
06:42 camelia nqp-{moarvm,parrot}: ( no output )
06:42 camelia ..nqp-jvm: OUTPUT«(signal ABRT)»
06:42 Augustus oh and if you switch the op to say and get rid of name, it does in fact print out Hello World
06:43 Augustus so the other parts do seem to work
06:44 moritz Augustus: what I don't see in this code is anything that connects it the setting, in which &say is defined
06:44 moritz or the current scope, for that matter
06:44 moritz which would explain why the QAST::Var lookup fails
06:45 moritz whereas ops are available everywhere
06:47 Augustus well say is a global function to it is essentially available everywhere, right?
06:48 Augustus or do I have to import it
06:48 Augustus because you're right I haven't set up any kind of scope
06:49 Augustus and obviously upon further reflection it would make no sense
06:49 Augustus since other languages don't have a global say function
06:50 Augustus and since I'm not really looking to try and import a bunch of p6 stuff (trying to keep it simple)
06:50 Augustus I should stick to built in ops and functions I actually write
06:51 Augustus a better test then would be for me to make my own say function that invokes the say op
06:51 Augustus and see if I can call that
06:52 Augustus Ok thanks for all your help everyone, it's late here so I'm gonna call it a night
06:52 Augustus you've all been super helpful! :)
06:52 TimToady 'night
06:53 moritz Augustus: built-ins are not global in Perl 6
06:53 moritz good night :-)
06:54 smls S19:159 say that «-i» (for one-liners that want to do in-place file editing) is still under "review"
06:54 synopsebot Link: http://perlcabal.org/syn/S19.html#line_159
06:54 smls What's athe argument against it?
06:54 smls I would find it useful... :)
06:56 TimToady maybe the kind of review it needs is someone trying to implement it and seeing what happens :)
06:57 smls Hm. Would that be junior job material?
06:59 smls On a similar note, while trying to further golf down some of the one-liners at https://github.com/sillymoose/Perl6-One-Liners,
06:59 smls it struck me that it would be cool to have a command-line switch that flattens the result of the «-e» expression and prints out the elements one per line
07:00 smls (or however the shell in question usually represents lists; i.e. possibly a native list in Powershell)
07:01 TimToady so a .say for ... basically
07:01 smls i.e. basically wrap the expression in «print "$_\n" for do { ... }»
07:01 smls yeah
07:02 smls maybe «-l» since the P5 meaning of the switch has not been carried over
07:02 smls perl6 -le '2, 4, ... 100'   # print even numbers up to 100
07:03 smls short and sweet :)
07:04 JimmyZ m: say (2, 6, 7 ... 100)
07:04 camelia rakudo-moar 93c2b4: OUTPUT«===SORRY!===␤Cannot assign to a readonly variable or a value␤»
07:06 smls because working with 1-element-per-line lists is such a common thing in the shell, that the «.say for ( ... )» or «.Str.say for ( ... )» feels like boilerplate
07:07 kaleem joined #perl6
07:10 smls JimmyZ: Hm, locally I get the more sensible error message "Unable to deduce arithmetic or geometric sequence from 2,6,7"
07:12 JimmyZ locally I get the same error message '
07:13 JimmyZ Cannot assign to a readonly variable or a value'
07:16 TimToady so do I
07:19 TimToady looks like it's failing to produce the error message somehjow
07:19 TimToady m: say (1,2,42 ... )
07:19 camelia rakudo-moar 93c2b4: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/g948ClqJR8â�¤Bogus statementâ�¤at /tmp/g948ClqJR8:1â�¤------> [32msay (1,2,42 ... [33mâ��[31m)[0mâ�¤    expecting any of:â�¤        postfixâ�¤        infix stopperâ�¤        infix or meta-infix…»
07:19 TimToady m: say (1,2,42 ... *)
07:19 camelia rakudo-moar 93c2b4: OUTPUT«===SORRY!===␤Cannot assign to a readonly variable or a value␤»
07:20 TimToady m: say (1,2,4 ... *)
07:20 camelia rakudo-moar 93c2b4: OUTPUT«1 2 4 8 16 32 64 128 256 512 1024 2048 4096 8192 16384 32768 65536 131072 262144 524288 1048576 2097152 4194304 8388608 16777216 33554432 67108864 134217728 268435456 536870912 1073741824 2147483648 4294967296 8589934592 17179869184 34359738368 68719476736…»
07:21 kurahaupo joined #perl6
07:21 anaeem___ joined #perl6
07:21 TimToady m: say (1,2,3 ... *)
07:21 camelia rakudo-moar 93c2b4: OUTPUT«1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 8…»
07:22 TimToady m: sink (1,2,3 ... *)
07:22 camelia rakudo-moar 93c2b4: ( no output )
07:23 TimToady m: sink (1,2,33 ... *)
07:23 camelia rakudo-moar 93c2b4: ( no output )
07:23 TimToady m: eager (1,2,33 ... *)
07:23 camelia rakudo-moar 93c2b4: ( no output )
07:24 JimmyZ joined #perl6
07:30 dj_goku joined #perl6
07:30 dj_goku joined #perl6
07:32 smls TimToady: Another idea: Change «-p» to print the result of the last statement rather than $_, but print the original line when that result is Nil
07:32 jluis joined #perl6
07:32 smls It might be a better fit for the more 'functional' way of doing things in P6
07:33 smls things like «perl6 -pe 's:g/foo/far' file.txt» would still work the same
07:34 smls but we'd no longer have to go through contortions to to mutate $_ in cases like «perl6 -pe '$_ = .uc if /foo/' example.txt»
07:36 espadrine_ joined #perl6
07:50 FROGGS joined #perl6
07:53 moritz .=uc if /foo/ doesn't look too bad to me
07:53 moritz m: $_ = 'ab'; .=uc; .say
07:53 camelia rakudo-moar 93c2b4: OUTPUT«AB␤»
07:56 Isp-sec joined #perl6
08:01 smls moritz: Well it may only be one character longer, but takes me two or three times as long to write due to having to stop and think where the dot goes
08:02 smls it adds an extra level of complexity
08:03 smls the current behavior made sense in P5 where things like s/// could only be used for mutating and not functionally
08:03 smls but in P6, functional expressions are not only possible for pretty much everything, but also feel more natural and elegant
08:03 smls s/feel/often feel/
08:03 moritz smls: that's a good point
08:04 moritz smls: but I wonder if  return value OR $_  might be a bit too magical
08:13 smls I think it would look quite nice in practice
08:13 smls «cat input.txt | perl6 -pe '.uc'»
08:13 smls «cat input.txt | perl6 -pe '.uc if /foo/'»
08:15 moritz smls: how would one filter out lines with -pe then?
08:15 smls 'next'
08:15 moritz fairy nuff
08:15 smls or use -ne with manual say
08:16 FROGGS next does already work, right?
08:17 smls FROGGS: yes
08:18 moritz $ echo -e "a\nb" |perl6-m -pe 'next if /a/'
08:18 moritz b
08:18 FROGGS proper compiler design++
08:18 anaeem1 joined #perl6
08:19 darutoko joined #perl6
08:19 anaeem1 joined #perl6
08:23 Mso150 joined #perl6
08:28 brrt joined #perl6
08:28 brrt \o
08:28 brrt i wanted to point out that my static perl6 thingy is slowly moving ahead
08:29 moritz \o/
08:29 virtualsue joined #perl6
08:31 smls static perl6 thingy?
08:31 brrt :-) i can use it to statically embed perl6.moarvm yet
08:31 brrt yes
08:31 brrt basically, you know how perl6 (on moar) is just a moarvm binary coupled with a runner script?
08:31 brrt that runner script is annoying
08:32 brrt and frankly, having perl6 depend on consistent structure of 3 library directories isn't ideal, either
08:32 anaeem___ joined #perl6
08:33 smls so you're putting it all in a single binary?
08:33 brrt trying to
08:33 brrt i'm not sure if i'm correct in doing so
08:34 brrt whether it makes sense.. i'm doubting that now
08:35 brrt especially as libmoar.so is dynamically linked.. hmm
08:35 pdcawley joined #perl6
08:35 JimmyZ you can build static moar.exe though
08:36 brrt probably, but more importantly i'd need a static libmoar :-)
08:37 JimmyZ how does parrot do that?
08:37 FROGGS pbc2exe? I guess this will still link against the libparrot.so
08:38 FROGGS ldd /home/froggs/dev/nqp/install/bin/perl6-p
08:38 FROGGS linux-vdso.so.1 =>  (0x00007fff65ffe000)
08:38 FROGGS libparrot.so.6.9.0 => /home/froggs/dev/nqp/insta​ll/lib/libparrot.so.6.9.0 (0x00007f2643b90000)
08:38 FROGGS [...]
08:39 brrt making a static libmoar should be only a bit of work... but still
08:40 brrt and i'd need to know which .moarvm files are actually needed by perl6
08:40 bjz_ joined #perl6
08:40 * JimmyZ wonders why we can't steal the idea from parrot :)
08:40 moritz ALL OF THEM
08:41 moritz JimmyZ: we can, but parrot doesn't generate static, fully-contained executables (afaict)
08:41 JimmyZ oh, that's right
08:41 brrt so, cause ambition
08:41 moritz JimmyZ: they still load the ops DLL/.so
08:42 * brrt would argue that since at build time we still have the .o from them, we can do that ahead-of-time too
08:49 grondilu joined #perl6
08:49 FROGGS m: say EXPORT::<DEFAULT>.WHO
08:49 camelia rakudo-moar 93c2b4: OUTPUT«␤»
08:52 molaf_ joined #perl6
08:53 rindolf joined #perl6
08:57 FROGGS m: sub a is export { "a" }; say &EXPORT::ALL::a ~~ Callable
08:57 camelia rakudo-moar 93c2b4: OUTPUT«True␤»
08:58 FROGGS m: sub a is export { "a" }; say &EXPORT::ALL::
08:58 camelia rakudo-moar 93c2b4: OUTPUT«(Any)␤»
08:58 FROGGS m: sub a is export { "a" }; say EXPORT::ALL::
08:58 camelia rakudo-moar 93c2b4: OUTPUT«===SORRY!===␤Object of type ALL in QAST::WVal, but not in SC␤»
09:08 pdcawley joined #perl6
09:14 abraxxa joined #perl6
09:15 timotimo i seem to be having a lot of trouble backlogging ... the backlogs keep piling up and i can't get terribly much read
09:18 ptc_p6 joined #perl6
09:20 smls maybe perl6/rakudo as a project has grown beyond the size where IRC can efficiently be used as the main communication channel
09:20 smls in fact I was surprised that it does; most other mid-sized open-source projects tend to use mailing lists
09:24 ptc_p6 joined #perl6
09:26 lizmat Public Service Announcement: tomorrow I will be doing the 2014.11 release of Rakudo
09:27 virtualsue \o/
09:28 rurban joined #perl6
09:28 lizmat I will be naming the release "Helsinki" in honour of the Nordic Perl Workshop (1st in Helsinki ever, afaik)
09:28 JimmyZ What about the release of MoarVM :)
09:28 lizmat well, that would be up to jnthn or FROGGS or timotimo perhaps?
09:28 zakharyas joined #perl6
09:28 lizmat today I will be de-commuting most of the day
09:29 osfameron lizmat: are you both in Helsinki?
09:30 lizmat no, in Copenhagen at the moment
09:30 * osfameron really wanted to go to NPW this year :-(  (will probably make 2015 Oslo)
09:30 brrt \o/
09:30 osfameron ah, still nice
09:30 brrt npw = nordic perl workshop? hadn't really imagined that
09:30 brrt oh hopefully we'll have a static perl6 by christmas then
09:31 brrt for the 2014.12 release :-)
09:31 lizmat osfameron: yes, we went to the Tivoli Gardens yesterday evening
09:31 lizmat brrt: that would be nice!
09:31 lizmat brrt: yes, Nordic Perl workshop
09:31 dakkar joined #perl6
09:31 woolfy brrt: it's not confusing at all: NPW = Nordic Perl Workshop, NLPW = Netherlands Perl Workshop
09:32 brrt and LPW?
09:32 osfameron London
09:32 woolfy London
09:32 woolfy :-)
09:32 brrt ah
09:32 * osfameron hugs woolfy
09:32 brrt logic
09:32 tadzik logic perl workshop
09:32 * woolfy hugs osfameron
09:32 * osfameron considers Nordic London Perl Workshop
09:32 * brrt is hugely excited for perl6-in-2015 :-D
09:32 lizmat DPW = Dutch Perl Workshop, or Deutscher Perl Workshop ?
09:32 lizmat brrt is not alone  :-)
09:32 brrt i think it should be deutscher
09:33 woolfy PPW = Pittsburgh Perl Workshop, PLPW = Polish Perl Workshop, PTPW = Portuguese Perl Workshop
09:33 brrt although that isn't actually consistent with the 'workshop' part
09:33 woolfy GPW = German Perl Workshop
09:33 kurahaupo joined #perl6
09:33 woolfy We've got several t-shirts with enormous characters GPW for the workshop!
09:34 woolfy osfameron should organise a Nordic London Perl Workshop and not call it NLPW!
09:34 brrt that settles it
09:36 osfameron the Scandiwegian Perl Workshop = SPW
09:36 FROGGS lizmat: I can only prepare the release, but jnthn has to upload it
09:36 osfameron me and ilmari can meet and talk in Finnish (I'd have to give a very short talk)
09:36 lizmat I guess jnthn will be done $working for the day in a few hours...
09:37 lizmat I guess we only picked up 1 important word in Finnish: Alco
09:37 lizmat (or was it Alko?)
09:37 osfameron Alko is the beer shop
09:38 lizmat alko, yes, the finnish version of the swedish systembolaget
09:38 lizmat osfameron: more like wine / liquor / whisky, at least the Alko we've been to  :-)
09:39 osfameron yes, true, "alcoholic beverages" by metonymy :D
09:40 sqirrel_ joined #perl6
09:40 lizmat brrt: last year, some visitors thought the NLPW was the Neuro-Linguistic Programming Workshop
09:41 FROGGS and then they came and wiped your branes?
09:41 lizmat no, the *expected* to have their branes wiped  :-)
09:41 lizmat *they
09:41 brrt seriously.. wow
09:41 virtualsue they hadn't actually signed up, had they?
09:42 lizmat no, they were accidental passers by at the University of Humanistics, where the NLPW was held
09:42 virtualsue that's ok then
09:43 lizmat :-)
09:45 zakharyas joined #perl6
09:47 woolfy left #perl6
09:48 lizmat decommute&
09:55 JimmyZ joined #perl6
09:56 KCL_ joined #perl6
09:57 pecastro joined #perl6
09:58 espadrine_ joined #perl6
10:14 fhelmberger joined #perl6
10:27 pdcawley joined #perl6
10:29 bjz joined #perl6
10:31 masak evening, #perl6
10:31 yoleaux 17 Nov 2014 14:16Z <Ven> masak: typo on your last blog post "Macro paramters/"
10:33 masak vendethiel: fixed, thanks.
10:35 masak 3 replacements s/\.slurp/\.slurp-rest/ in my blogging software.
10:35 masak lizmat++ # most probably
10:41 lucas___ joined #perl6
10:43 lucas___ I'm sorry to be picky about it, but last specs commit accidentally removes some line breaks from S17.
10:47 virtualsue joined #perl6
10:50 masak I can revert it, but that doesn't fix the fact that blame won't give much info on that file next time.
10:50 kaleem joined #perl6
10:53 Ugator joined #perl6
10:57 kaleem joined #perl6
11:02 eLtJgTM joined #perl6
11:05 regreg joined #perl6
11:08 rindolf joined #perl6
11:15 eLtJgTM joined #perl6
11:19 eLtJgTM left #perl6
11:22 jluis joined #perl6
11:29 sqirrel_ joined #perl6
11:32 prevost joined #perl6
11:44 vendethiel joined #perl6
11:44 telex joined #perl6
11:45 chenryn joined #perl6
11:50 bjz joined #perl6
11:55 Woodi joined #perl6
12:23 darutoko joined #perl6
13:08 PZt joined #perl6
13:12 zakharyas joined #perl6
13:16 guru joined #perl6
13:21 firefish5000 joined #perl6
13:28 daxim http://planetsix.perlfoundation.org/    # daed?
13:29 masak seems so. I just get domain spam here.
13:33 vendethiel joined #perl6
13:36 moritz daxim: http://pl6anet.org/ has the current feed
13:37 daxim cool.  who's the hostmaster or webmaster for perlfoundation.org? needs a redirect
13:38 nwc10 daxim: strictly, needs its redirect updating
13:39 chenryn joined #perl6
13:39 nwc10 http://www.perlfoundation.org/contact_us says Casey West - webmaster(at)perlfoundation.org
13:40 nwc10 but given that the CNAME for planetsix.perlfoundation.org points to develooper.org
13:40 nwc10 er
13:40 nwc10 develooper.com
13:40 nwc10 I'd suggest this contact:
13:40 nwc10 Admin Email: ask@develooper.com
13:40 nwc10 ie ask Ask
13:43 moritz but the destination admin can't do much about the wrong redirect, can he?
13:44 nwc10 I am inferring that the redirect is in the varnish config on a machine that Ask adminsters
14:00 * masak .oO( Perl 6 for Morpheusphytes )
14:00 arnsholt Any 6model experts around? Method lookups can be customised in a HOW via find_method, but is there something similar to customise attribute handling?
14:01 arnsholt (Because everything in Python is an attribute. Methods are just magical subs created on attribute lookup in a class object)
14:02 jnthn Attribute access is all about representation.
14:02 yoleaux 18 Nov 2014 23:36Z <bronco_creek> jnthn: When I try to use panda to install JSON::Path with my 2014/9 rakudo* on moarvm it fails a test.
14:02 jnthn For a Python object, isnt it really just a hash?
14:02 jnthn *isn't
14:02 arnsholt Yeah, basically
14:02 arnsholt But there's a bit of magic on top
14:03 jfredett joined #perl6
14:03 jnthn So probably you want the meta-object to use the representation that provides a hash-based attribute storage
14:03 arnsholt Yeah, I'm starting to realize that now
14:03 jnthn You'll need to do the clever stuff in high level code I guess...
14:04 jnthn But inlining and spesh could make it not too costly.
14:04 arnsholt Attribute lookups need to invoke attribute-getting code in the meta-object that does the correct stuff behind the scenes
14:04 jnthn Attribute access doesn't go via the meta-object in stuff we have so far...
14:05 arnsholt Yeah, I'll have to generate that code whenever I see an attribute access
14:05 jnthn Tjough there's no reason you can't compile them thta way.
14:05 jnthn Well, it'd be nice if there was a fast path/slow path thing though
14:06 arnsholt The thing is that when an attribute doesn't exist in an instance, you have to search the class object for an attribute of that same name
14:07 arnsholt If the class object (or one of its parents) has that member return that. And if it's a callable, you return (more or less) sub(*args) { $meth($invocant, |@args) }
14:07 jnthn I'd probably code-gen a simple getattr
14:08 arnsholt Probably wise, yeah
14:08 jnthn Followed by a if not null goto fo ; if not invokable goto foo ; call slow path
14:08 jnthn *foo
14:08 jnthn And then spesh can probably too those checks for guards
14:08 arnsholt Troo, troo
14:08 jnthn And so eliminate the branches.
14:08 masak arnsholt: methods in Python are descriptors.
14:09 masak arnsholt: which is how you bind an unbound method, for example. you call __get__ on it.
14:10 arnsholt Yeah, there's that stuff too
14:10 * masak .oO( yer a wizard, @arry[2] )
14:11 arnsholt But the stuff above is more or less verbatim from the spec
14:15 masak arnsholt: https://gist.github.com/masak/864795f3af67359d9a9f
14:15 arnsholt Oh, blerg. There's an implementation of those by default of course
14:16 arnsholt The description of descriptors in the spec starts like this: "The following methods only apply when an instance of the class containing the method [...]" =)
14:17 masak lizmat++ # http://www.liz.nl/Perl6ForNeophytesNPW2014.pdf
14:18 masak arnsholt: you're building an object system, and complaining about circularity? :P
14:18 arnsholt I know, I know!
14:18 chenryn joined #perl6
14:18 arnsholt Objects are hard =)
14:19 jnthn .oO( Circularity in object systems? I saw it all before... )
14:22 onebitboy joined #perl6
14:23 masak arnsholt: jnthn and I had a nice discussion of 2-categories over dinner. (still quite jetlagged.) :)
14:23 jluis joined #perl6
14:23 masak in a 2-category, even the arrows have arrows!
14:24 arnsholt Gah! =D
14:25 sunnavy joined #perl6
14:25 rurban joined #perl6
14:25 arnsholt masak: Oh, it's actually not a problem
14:26 arnsholt It's the unbound method `C.foo` that has a __get__, not C
14:26 masak yep.
14:27 masak the bound methods don't. I don't think there's a way to rebind a method.
14:27 arnsholt Yeah, you can read the invocant member (name is __self__ IIRC), but it's RO
14:27 arnsholt Anyways, I'm gonna ignore descriptors for the time being =)
14:29 masak probably wise.
14:29 masak arnsholt++ # tinkering with snake
14:35 arnsholt It's fun, so far
14:35 virtualsue joined #perl6
14:35 arnsholt Making me want to put stuff into the dormant NQP book too
14:36 masak \o/
14:38 ghostlines joined #perl6
14:40 arnsholt Also, having a GitHub project is a good idea
14:40 arnsholt All of the notes about how to implement stuff can be stuffed into issues for later reference
14:43 KCL joined #perl6
14:45 cognominal joined #perl6
14:57 * [Coke] is very confused at what augustus is trying to do.
14:59 spider-mario joined #perl6
14:59 BinGOs found an empire.
14:59 masak name a month after himself.
15:00 jnthn ceaser the day
15:02 masak or the month.
15:04 kaare_ joined #perl6
15:07 rindolf joined #perl6
15:10 * [Coke] sees the discussion has been over for a while, nevermind he said anything.
15:11 masak [Coke]: I'll say. Augustus died 2000 years ago. :P
15:11 raiph joined #perl6
15:13 timotimo he died for our sins, didn't he?
15:17 kaleem joined #perl6
15:18 FROGGS joined #perl6
15:19 raydiak \o
15:20 FROGGS o\
15:20 adu joined #perl6
15:21 pecastro joined #perl6
15:30 guru joined #perl6
15:32 chenryn joined #perl6
15:34 [Coke] man, I wish I timed the daily runs. :P
15:44 FROGGS jnthn: you know about release day tomorrow?
15:44 firefish5000 joined #perl6
15:45 jnthn FROGGS: Uh...not really :/
15:46 jnthn $dayjob is nomming all the time :(
15:46 jnthn Until the weekend or so, then it may get a bit better.
15:46 FROGGS jnthn: can you (re)upload a moarvm tarball tomorrow?
15:46 FROGGS I'd prepare All The Things like last time
15:47 jnthn FROGGS: That would be awesome
15:47 FROGGS cool
15:47 jnthn FROGGS: I can, though please note I'm currently 7 timezones ahead of you.
15:47 FROGGS ahead...
15:47 FROGGS uhh
15:47 jnthn (So now it's about to be midnight.)
15:47 FROGGS ahh, alright
15:48 jnthn I could also swing by on Friday morning if that's too rushed, before I head out for teaching.
15:48 jnthn Which'd be about midnight your time
15:48 jnthn Anyways, please /msg me the url
15:48 FROGGS then it might make sense that I prepare it tonight (within seven hours), then you've time tomorrow while the shines on your place of the earth :o)
15:48 FROGGS the sun*
15:48 FROGGS k
15:49 jnthn The sun shines, but the smog layer was making a good effort of hiding the fact today...
15:49 FROGGS hehe
15:49 FROGGS (Berlin suburbs)++ I'd say
15:49 PerlJam jnthn: would it be possible to share the love with FROGGS (or someone else) so that they can do the upload?
15:49 chenryn joined #perl6
15:51 jnthn That's the right way to fix it medium term, but it'll take me more effort to set up than to do a wget tomorrow... But yeah, will look into it once I've time again. :)
15:51 jnthn Sleep time now...'night
15:51 colomon o/
15:51 FROGGS gnight jnthn
15:53 brrt lizmat++ for the presentation. stuff i didn't know :-)
15:59 timotimo m: say 10 + 3i % 2 + 2i
15:59 camelia rakudo-moar 93c2b4: OUTPUT«Cannot call 'Real'; none of these signatures match:␤:(Mu:U \v: *%_)␤  in sub infix:<%> at src/gen/m-CORE.setting:4588␤  in sub infix:<%> at src/gen/m-CORE.setting:4588␤  in block <unit> at /tmp/98S_RX1orG:1␤␤»
15:59 timotimo is it a dumb idea to want to have a component-wise % operation on complex numbers?
16:02 treehug88 joined #perl6
16:03 brrt modulus for complex numbers?
16:03 Util timotimo: Do you have a use case (instead of just an example)? Does this come up in math that you have seen, or is it a clever idea of your own?
16:03 brrt i'm not sure that makes sense
16:04 timotimo Util: i have not seen it in math. i want to use it to warp the player's position (represented as a Complex) into a rectangular game field
16:04 brrt well... that's wrong
16:05 brrt :-)
16:05 brrt unless you want a players (0,1) position multiplied to become (1,0)
16:05 brrt eh (-1,0)
16:05 brrt represent it as something else instead :-)
16:06 Util timotimo: OK, then I would implement it as a <%> operator on (Complex,Complex), with the operator defined in your program or library, not in the language itself.
16:07 Util You are semi-abusing Complex, to exploit the fact that it has 2 (paired) component Reals, so just extend the operator on Complex to extend this use.
16:08 Util IMHO, you would want to keep the core language Complex pure to the behaviour that Math people would expect.
16:08 Util Otherwise, if you dif add it to the core language, I would make it a method instead of an op, like Slurp is added to Quaternions,
16:09 * timotimo has implemented Snake
16:10 Util because lots of people who use Quaternions are exploiting their 3-D rotations for Game programming, and they all need Slerp, even though Slerp is not actually a part of Quaternion nature.
16:10 timotimo fair enough :)
16:11 timotimo 69.5% time spent in sleep ... well ... %)
16:12 FROGGS O.o
16:12 * FROGGS .oO( students.... )
16:12 arnsholt I was under the impression that 3D rotation is basically the only thing people use quaternions for
16:12 arnsholt Apparently it's a bit marginal in mathematics =)
16:23 nanis joined #perl6
16:23 Util arnsholt: Not nearly as marginal as Octonions and Sedenions :^)
16:24 Util The pure research (I think) is in the *full* Cayley–Dickson algebras, not restricted to 4|8|16 elements.
16:24 Util Dr. John Bales (who is in my LUG), and has written papers on Quaternions.
16:24 Util I think that the C-D calculator he wrote goes up to 256 elements. http://jwbales.us/
16:24 Util Disclosure: /me is just an "armchair mathematician", but is responsible for https://github.com/Util/Perl6-Math-Quaternion/ and http://rosettacode.org/wiki/Quaternion_type#Perl_6 .
16:25 masak heathens! quaternions are not "used" -- they are appreciated for their intrinsic beauty (and lack of commutativity).
16:25 timotimo :)
16:25 timotimo the GC pauses in my snake implementation take at least 9ms each time
16:26 regreg joined #perl6
16:27 Util :)
16:28 arnsholt timotimo: For a second I thought you were talking about Snake my Python implementation, rather than Snake the game
16:28 arnsholt =D
16:28 timotimo hah. making it force_gc after doing every single frame gives me better worst-case frame-rates, but the time spent in gc goes up to 60% from 13%
16:29 arnsholt Yeah, better interactivity through many smaller GCs
16:29 timotimo actually, i wasn't measuring the GC time for the frame counter
16:30 sqirrel_ joined #perl6
16:30 gtodd am I dreaming or was there a page somewhere that had per release set of benchmarks with pretty graphs and such  ?
16:31 timotimo http://t.h8.lv/p6bench/
16:31 gtodd :-)
16:32 timotimo if you want to build something way cool that generates very helpful overview pages for bunches of outputs ... that'd be greatly appreciated :)
16:34 gtodd you mean jnthn is not working on that in secret ? :-D
16:36 timotimo perl6-bench isn't jnthn's area of expertise
16:37 bjz joined #perl6
16:46 gtodd timotimo: well if the pages are autogenerated (I assume they are .... with App::Dapper or something like it) and the data automatically captured and jsonified onto the page for jqplot then that would exhaust my areas of knowledge
16:48 timotimo i autogenerated the index.html by piping the output of ls into a .html file and running a vi macro over it ...
16:49 timotimo i don't perl5. ever.
16:51 TimToady m: (1,2,3)[*-5]
16:51 camelia rakudo-moar 93c2b4: OUTPUT«Unhandled exception: Effective index [-2] is out of range for a Parcel of 3 elements (*1)␤   at <unknown>:1  (/home/camelia/rakudo-inst-2/languages/perl6/r​untime/CORE.setting.moarvm:throw:4294967295)␤ from src/gen/m-CORE.setting:13812  (/home/camelia/r…»
16:52 TimToady that's still kinda LTA
16:53 kst` joined #perl6
16:54 flussence .oO( I get worried every time I see a new Parrot release... they're going to run out of bird names eventually. )
16:55 TimToady m: (1,2,3)[-> $i { $i - 42 }]
16:55 camelia rakudo-moar 93c2b4: OUTPUT«Unhandled exception: Effective index [-39] is out of range for a Parcel of 3 elements (*-36)␤   at <unknown>:1  (/home/camelia/rakudo-inst-2/languages/perl6/r​untime/CORE.setting.moarvm:throw:4294967295)␤ from src/gen/m-CORE.setting:13812  (/home/cameli…»
16:56 TimToady m: (1,2,3)[-> $i { $i / 2 - 20 }]
16:56 camelia rakudo-moar 93c2b4: OUTPUT«Unhandled exception: Effective index [-18.5] is out of range for a Parcel of 3 elements (*-15.5)␤   at <unknown>:1  (/home/camelia/rakudo-inst-2/languages/perl6/r​untime/CORE.setting.moarvm:throw:4294967295)␤ from src/gen/m-CORE.setting:13812  (/home/ca…»
16:57 rurban joined #perl6
16:59 gtodd flussence: work to preserve wildlife and biodiversity
16:59 TimToady m: (1,2,3)[0-1]
16:59 camelia rakudo-moar 93c2b4: OUTPUT«Unhandled exception: Unsupported use of [-1] subscript to access from end of Parcel; in Perl 6 please use [*-1]␤   at <unknown>:1  (/home/camelia/rakudo-inst-2/languages/perl6/r​untime/CORE.setting.moarvm:throw:4294967295)␤ from src/gen/m-CORE.setting:1…»
17:00 TimToady m: my $x = -1; say (1,2,3)[$x]
17:00 camelia rakudo-moar 93c2b4: OUTPUT«Unsupported use of [-1] subscript to access from end of Parcel; in Perl 6 please use [*-1]␤  in method gist at src/gen/m-CORE.setting:13804␤  in sub say at src/gen/m-CORE.setting:16535␤  in block <unit> at /tmp/lptxgA2NtO:1␤␤»
17:00 TimToady m: my $x = -3; say (1,2,3)[$x]
17:00 camelia rakudo-moar 93c2b4: OUTPUT«Unsupported use of [-3] subscript to access from end of Parcel; in Perl 6 please use [*-3]␤  in method gist at src/gen/m-CORE.setting:13804␤  in sub say at src/gen/m-CORE.setting:16535␤  in block <unit> at /tmp/msNYi9U0xN:1␤␤»
17:00 TimToady that's also LTA
17:01 TimToady it should not be assuming that relative addressing was intended
17:02 timotimo one week left to get GLR up in time for thanksgiving ...
17:02 kjs_ joined #perl6
17:06 timotimo 25% of run time of my snake game is GC (because i force GC with a probability of 20% or something each frame) and 25% of the execution time itself is spent inside NativeCall's postcircumfix:<( )>
17:07 TimToady obviously you need a 4 core machine or so :)
17:13 gtodd timotimo: there's a thing in north american consumerism ... was mostly US but now big in Canada too called "Black Friday"
17:13 Ulti what's the most efficient way to match and replace a fixed substring in Rakudo on moarvm atm? I'm doing $str ~~ s:g/this/that/
17:14 timotimo gtodd: is that the thing where sometimes people die because they get trampled?
17:16 gtodd timotimo: :-|  argh ... yeah ... and it is destroying Thanksgiving traditions !
17:17 gtodd so I think GLR should aim for post Thanksgiving but it would be nice if some elements of GLR were ready for Perl6 Advent :-)
17:17 timotimo oh no, the sanctity of thanksgiving!
17:17 gtodd heheh
17:17 guru joined #perl6
17:18 gtodd m:   ( super => "perl5" , fantastic => "perl6" ).hash.reverse.say
17:18 camelia rakudo-moar 93c2b4: OUTPUT«"fantastic" => "perl6" "super" => "perl5"␤»
17:18 gtodd ( super => "perl5" , fantastic => "perl6" ).hash.invert.say
17:18 gtodd m: ( super => "perl5" , fantastic => "perl6" ).hash.invert.say
17:18 camelia rakudo-moar 93c2b4: OUTPUT«"perl5" => "super" "perl6" => "fantastic"␤»
17:19 cognominal joined #perl6
17:19 gtodd «Useless use of reverse on hash ... to invert a hash; in Perl 6 please use invert␤
17:22 [Sno] joined #perl6
17:23 kaleem joined #perl6
17:33 ugexe perl6: nqp::say("hi"); say("second"); say("third");
17:33 camelia rakudo-moar 93c2b4: OUTPUT«second␤third␤hi␤»
17:33 camelia ..rakudo-parrot 93c2b4: OUTPUT«hi␤second␤third␤»
17:33 timotimo https://gist.github.com/timo/326205e041d3c184a5c7
17:33 timotimo try and comment plox :)
17:34 gtodd timotimo: cool :)  .... are you using waterfall or whirlpool approach to developing Snake ? :-)
17:36 ugexe any idea what that output order is wrong? from the command line the output is correct
17:37 ugexe (for moar)
17:40 TimToady nqp does better buffering?
17:41 gtodd better than moar but worse than parrot or ... ?
17:41 TimToady or worse?
17:42 anaeem1_ joined #perl6
17:43 TimToady anyway, it would seem that nqp is doing block buffering on moarvm, while p6 is doing one of: no buffering, line buffering, or autoflushing
17:43 TimToady so the nqp output waits until the final flush
17:44 ugexe i see. and that somehow is affected by the facts its an eval server?
17:44 TimToady could be
17:44 TimToady line buffering is usually not turned on unless the output is a terminal
17:45 ugexe ah
17:45 ugexe also, re -1 index: my @a; @a[@a.end]; It does the proper thing, but seems wrong?
17:45 TimToady so my guess is nqp/moar is just defaulting to what stdio supplies while p6 is autoflushing
17:46 TimToady the current error messages are making unwarranted assumptions about how the index was calculated
17:47 TimToady m: my @a; say @a[@a.end]
17:47 camelia rakudo-moar 93c2b4: OUTPUT«Unsupported use of [-1] subscript to access from end of Array; in Perl 6 please use [*-1]␤  in method gist at src/gen/m-CORE.setting:13804␤  in sub say at src/gen/m-CORE.setting:16535␤  in block <unit> at /tmp/8TmC83f_h9:1␤␤»
17:48 TimToady it should say something more like: Negative calculated subscript (-1) is not allowed because Array is always 0-based
17:48 TimToady it should not guess that you said *-1
17:49 TimToady it should not assume that you wrote [-1]
17:49 TimToady it should not guess that you meant *-1, I should say
17:51 TimToady and we *still* don't detect a -1 subscript at compile time, which is the correct time to tell them about *-1
17:51 FROGGS_ joined #perl6
17:51 vendethiel masak: (.tell :p)
17:52 TimToady m: sub foo { @_[-1] }
17:52 camelia rakudo-moar 93c2b4: ( no output )
17:52 TimToady std: sub foo { @_[-1] }
17:52 TimToady oops
17:52 TimToady I successfully created a snap directory yesterday, but that's not enoughj
17:52 ugexe i tried hack it to detect at compile time but i think i just wasted my time randomly changing crap with gimmie and reify lol
17:52 FROGGS_ TimToady: in the repo or on that machine?
17:53 TimToady on that machine
17:53 camelia joined #perl6
17:53 TimToady when it rebrews it goes away again
17:53 FROGGS_ so it should live in the repo too?
17:53 TimToady so something needs fixing in the build process
17:54 TimToady build.pl std calls make snap, which should create it, but it doesn't, so the make is failing somehow, probably a missing dep
17:54 camelia std : OUTPUT«Can't chdir to '/home/camelia/std/snap': No such file or directory at lib/EvalbotExecuter.pm line 166.␤    EvalbotExecuter::_auto_execute(HASH(0x1c009c8), "sub foo { \@_[-1] }", GLOB(0xada3d68), "/tmp/wfuVBD9hmT", "std") called at lib/EvalbotExecuter.pm line 114␤    Evalb…»
17:55 TimToady there's a build.pl std entry in crontab, but the output goes to /dev/null
17:57 TimToady can't do anything much about it currently; trying to look through migraine auras makes it really hard to read
17:57 TimToady fortunately my migraines tend not to be very painful
17:58 ptc_p6 joined #perl6
18:03 moritz "Can't locate YAML/XS.pm in @INC"
18:04 * moritz installs it
18:05 TimToady lesse, next will be Moose, or Try::Tiny
18:06 TimToady seems like there was one other I often ran into
18:07 * moritz is now installing Moose
18:07 moritz wasn't there a time when Moose had only a handful of non-core deps?
18:08 moritz Text::Glob, Number::Compare, Test::CleanNamespaces
18:08 TimToady wasn't there a time when Perl 6 was much more like Perl 5?
18:08 moritz gotcha :-)
18:10 TimToady are you doing this on nine's machine?
18:11 TimToady or changing the brewery?
18:12 kjs_ joined #perl6
18:15 moritz TimToady: on nine's machine
18:15 moritz TimToady: and logs are in log/std.log
18:16 moritz File::ShareDir missing...
18:16 TimToady that was the other one :)
18:17 moritz now it takes much longer
18:17 moritz that's a good sign
18:17 * moritz hopes TimToady++'s migraine retreats quickly
18:18 kurahaupo joined #perl6
18:19 TimToady the auras are now far enough away from my macula that I can read again :)
18:19 timotimo gtodd: i didn't follow a particular development strategy ... which i suppose is similar to what whirlpool means?
18:20 TimToady I still get the "mean" feeling you get with migraines, but I probably burned out the pain nerves with shingles long ago, so that part isn't a problem
18:21 moritz std: 42
18:21 camelia std 76ccee1: OUTPUT«ok 00:00 135m␤»
18:21 TimToady \o/
18:21 moritz that looks fairly recent.
18:21 timotimo wow, std can parse integer literals now?
18:21 TimToady std: sub foo { @_[-1] }
18:21 camelia std 76ccee1: OUTPUT«[31m===[0mSORRY![31m===[0m�Unsupported use of [-1] subscript to access from end of array; in Perl 6 please use [*-1] at /tmp/IixH2BAjik line 1:�------> [32msub foo { @_[-1][33m�[31m }[0m�Parse failed�FAILED 00:00 140m�»
18:21 timotimo we're really getting somewhere here!
18:21 TimToady std: sub foo { @_[-2] }
18:21 camelia std 76ccee1: OUTPUT«[31m===[0mSORRY![31m===[0m�Unsupported use of [-2] subscript to access from end of array; in Perl 6 please use [*-2] at /tmp/NueXthJ1SE line 1:�------> [32msub foo { @_[-2][33m�[31m }[0m�Parse failed�FAILED 00:00 140m�»
18:22 TimToady std: sub foo { @_[0;-2] }
18:22 camelia std 76ccee1: OUTPUT«ok 00:00 140m␤»
18:22 TimToady didn't think so
18:22 telex joined #perl6
18:26 gtodd timotimo: sorry OT :-) I just read a book I think it was called "Implementation by Example" ... which was sort of a downer ... it made me think of "whirlpool"
18:26 gtodd (which I think TimToady gave a name to)
18:34 timotimo it was a downer in what sense? you were disappointed with its quality/helpfulness?
18:37 ennio joined #perl6
18:40 [Sno] joined #perl6
18:40 Mso150 joined #perl6
18:41 timotimo i have lots of ideas for more features in that game, fwiw
18:41 PerlJam in retrospect, "whirlpool" is kind of an unfortunate name as they go round and round without ever making progress in a particular direction
18:41 timotimo mostly graphical
18:44 * arnsholt can haz objects o/
18:44 arnsholt Objects without attributes or methods, but still =D
18:45 FROGGS_ I... object
18:48 TimToady PerlJam: no, eventually the whirlpool sucks you down the middle
18:55 ennio left #perl6
18:58 PerlJam TimToady: I dunno.  Most whirlpools that I've run across don't have much of a downward pull; they just lazily go round and round
18:59 Sqirrel joined #perl6
19:02 PerlJam wikipedia also seems to make a distinction:  "Vortex is the proper term for any whirlpool that has a downdraft."  from http://en.wikipedia.org/wiki/Whirlpool
19:03 TimToady obviously a tornado is not a vortex then :P
19:03 TimToady and Captain Nemo was scared of nothin'
19:04 PerlJam also, from that page: "In popular imagination, but only rarely in reality, whirlpools can have the dangerous effect of destroying boats."   ;)
19:04 TimToady so a process that may or may not converge is certainly well described by "whirlpool" :)
19:06 TimToady and "eddy" generally describes the non-convergent whirlpool, so we don't need to force "whirlpool" to mean that
19:07 * TimToady hates it when so-called experts try to make general terms mean specific things, except when the expert is him, of course...
19:07 regreg joined #perl6
19:08 PerlJam Isn't that what language design is all about?  ;-)
19:09 ennnio joined #perl6
19:11 PerlJam ennnio: I must be projecting my inner state out into the world a little bit as I read your nick as "ennuio" (that is ennui-o)
19:17 guru joined #perl6
19:21 kurahaupo1 joined #perl6
19:23 ennnio PerlJam, may be I'll change my nick to improve your inner state?
19:24 raiph lucas++, masak++ thanks for notes about my S 17 commit; lesson learned
19:24 raiph sorry it broke stuff and revert won't fix it /o\
19:26 virtualsue joined #perl6
19:27 Mso150 joined #perl6
19:27 sjn_ joined #perl6
19:27 dagurval joined #perl6
19:31 KCL_ joined #perl6
19:35 ghostlines joined #perl6
19:37 rurban joined #perl6
19:41 grondilu left #perl6
19:41 adu joined #perl6
19:43 Ven joined #perl6
19:44 Ven dammit, I always type "perl6weekly.wordpress.com" :(
19:56 travis-ci joined #perl6
19:56 travis-ci NQP build failed. Tobias Leich 'bump parrot and moar revision to current release'
19:56 travis-ci http://travis-ci.org/perl6/nqp/builds/41520148 https://github.com/perl6/nqp/com​pare/54ee66700d5b...d0d274e13454
19:56 travis-ci left #perl6
19:58 travis-ci joined #perl6
19:58 travis-ci Rakudo build failed. Tobias Leich 'bump nqp/moar/parrot revision'
19:58 travis-ci http://travis-ci.org/rakudo/rakudo/builds/41520304 https://github.com/rakudo/rakudo/c​ompare/93c2b4c4e16d...031e8af114e2
19:58 travis-ci left #perl6
20:12 [Coke] parrot the only clean rakudo implementation. Release is RSN.
20:14 [Coke] jvm failing 30 tests; moar-jit 103, moar-nojit 21.
20:14 FROGGS_ that much? unbelievable
20:15 FROGGS_ I've got a clean spectest run on moar+jit
20:15 [Coke] every day. haven't had a clean moar run in weeks.
20:15 [Coke] you're not on OS X, then.
20:15 [Coke] I am regularly getting segfaults and memory corruption errors.
20:16 [Coke] none of the jit, no-jit failures coincide, btw.
20:18 bartolin [Coke]: after your last fudge for parrot I had clean runs for parrot, moar and moar-jit. jvm failed 23 tests (all data from last night on FreeBSD)
20:18 [Coke] I assume the jvm failures are non-controversial: https://github.com/coke/perl6-roast-data/bl​ob/master/log/rakudo.jvm_summary.out#L1712
20:19 [Coke] the roll failures may be pointing at bad tests, or bad rand stuff on jvm. I unno, those flap a bit.
20:19 [Coke] bartolin++
20:19 bartolin I have not seen any flappers there for a while. also I had less failures: https://github.com/usev6/perl6-roast-data/b​lob/master/log/rakudo.jvm_summary.out#L1710
20:19 travis-ci joined #perl6
20:19 travis-ci NQP build passed. Tobias Leich 'adopt to new MAST::Ops/MAST::Nodes location'
20:19 travis-ci http://travis-ci.org/perl6/nqp/builds/41522573 https://github.com/perl6/nqp/com​pare/d0d274e13454...86f18410156b
20:19 travis-ci left #perl6
20:19 bartolin [Coke]++ ;-)
20:20 [Coke] so, might be the rolls
20:20 bartolin yes, it seems so
20:23 [Coke] if you dig through the java history on my site, you'll those come and go, I think (even before the switch to OS X)
20:23 FROGGS_ valgrind does not report anything for the failing moar tests :/
20:24 [Coke] bartolin: if you're using my stuff, have you re-synced lately? some changes to at least get the builds going in parallel when possible.
20:24 [Coke] FROGGS_: there's a moarvm ticket I opened for one of the tests (now passing, of course) that might have some useful detail.
20:25 [Coke] https://github.com/MoarVM/MoarVM/issues/154
20:26 [Coke] looks like moar-nojit has 3 memory errors which i assume are related to #154, and one segfault.
20:27 [Coke] looks like jit is all segfaults.
20:27 bartolin [Coke]: no, I'm haven't pulled in your changes. if I find some time, I'll do that, thanks! (since the spectests run over night, I don't care too much if it takes a bit longer)
20:30 adu joined #perl6
20:33 bartolin btw, I started a new job this month, so atm I've only occasionally some free time for Perl 6
20:34 bartolin (in case someone wondered what happened after my burst of activity in october ;-)
20:35 [Coke] I have an old job, and I have similar issues. :)
20:35 travis-ci joined #perl6
20:35 travis-ci Rakudo build passed. Tobias Leich 'pull in a build fix for nqp'
20:35 travis-ci http://travis-ci.org/rakudo/rakudo/builds/41523730 https://github.com/rakudo/rakudo/c​ompare/031e8af114e2...2a2dd43e8eef
20:35 travis-ci left #perl6
20:35 * PerlJam too
20:36 felher r: my (\x) = 3; say x;
20:36 camelia rakudo-{parrot,moar} 031e8a: OUTPUT«(Any)␤»
20:36 FROGGS_ bartolin: I still see that you are quite active :o)
20:36 felher Shoudn't that say "3" or am I mistaken?
20:36 FROGGS_ [Coke]: either I don't understand how unions work or Apple doesnt...
20:37 FROGGS_ r: my (\x = 3); say x;
20:37 camelia rakudo-{parrot,moar} 031e8a: OUTPUT«(Any)␤»
20:37 FROGGS_ r: my \x = 3; say x;
20:37 camelia rakudo-{parrot,moar} 031e8a: OUTPUT«3␤»
20:37 FROGGS_ I guess it does not see the initializer with parens
20:39 TimToady it's not an initializer; it's a parameter with a default
20:39 TimToady m: my (\x = 3) ::= (); say x;
20:39 camelia rakudo-moar 031e8a: OUTPUT«(Any)␤»
20:39 TimToady that one should be 3
20:40 felher But "my (\x, \y) = 1, 2" should work too, right?  Or is that only supposed to work with normal variables?
20:42 ghostlines joined #perl6
20:42 felher r: my (\a, \b) = 1,2; my ($x, $y) = 3, 4; say a, b, $x, $y;
20:42 camelia rakudo-{parrot,moar} 031e8a: OUTPUT«(Any)(Any)34␤»
20:46 felher I did expect to get 1234 :)
20:47 bartolin FROGGS_: did you intend to close RT #118501 or was that by mistake?
20:47 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=118501
20:47 FROGGS_ ups
20:48 travis-ci joined #perl6
20:48 travis-ci Rakudo build passed. TimToady 'unify the Negative exceptions
20:48 travis-ci http://travis-ci.org/rakudo/rakudo/builds/41523915 https://github.com/rakudo/rakudo/c​ompare/2a2dd43e8eef...10aa8bf41c81
20:48 travis-ci left #perl6
20:48 FROGGS_ bartolin: I reopened it, thanks
20:49 bartolin FROGGS_++
20:50 FROGGS_ ohh, then let me close+reopen another bunch :P
20:50 bartolin *g*
20:50 FROGGS_ such karma, much happy /o/
20:54 bartolin felher: afaik there are some issues with parsing things like 'my ( ... ) = $foo', because ( ... ) is parsed as a signature. in an old ticket I found an short explanation from jnthn++: https://rt.perl.org/Ticket/Dis​play.html?id=74664#txn-686876
20:55 felher bartolin: thanks :)
20:55 bartolin m: my Int (Str $x); say $x.WHAT # RT #73102
20:55 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=73102
20:55 camelia rakudo-moar 031e8a: OUTPUT«(Int)␤»
20:56 FROGGS_ [Coke]: this might fix it: https://github.com/MoarVM/MoarVM/commit/​33970c6dfdddb5f1d3d667bf1e294ee8f09185d0
20:56 bartolin m: my (Str $a) = 3; say $a # RT #115916
20:56 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=115916
20:56 camelia rakudo-moar 031e8a: OUTPUT«3␤»
20:56 FROGGS_ [Coke]: is there a chance to test that commit? it is >2014.11, and therefore I cannot bump the revision in nqp
20:58 eclipse joined #perl6
21:01 FROGGS_ m: say join '', ('' xx 64)
21:01 camelia rakudo-moar 031e8a: OUTPUT«␤»
21:01 TimToady .oO(Much Ado About Nothing)
21:01 FROGGS_ [Coke]: I dunno if something like the above might blow up on OSX
21:02 FROGGS_ and there are certainly many kinds of nothing
21:05 Mso150 joined #perl6
21:06 gtodd1 joined #perl6
21:07 ghostlines joined #perl6
21:08 ghostlines joined #perl6
21:09 netstar Is there an IO::Socket implemented?
21:09 FROGGS_ m: say IO::Socket
21:09 camelia rakudo-moar 031e8a: OUTPUT«IO::Socket is disallowed in restricted setting␤  in sub restricted at src/RESTRICTED.setting:1␤  in method gist at src/RESTRICTED.setting:22␤  in sub say at src/gen/m-CORE.setting:16539␤  in block <unit> at /tmp/AE1vl051QD:1␤␤»
21:09 FROGGS_ netstar: yes, but not here in IRC as you can see :o)
21:10 netstar lol
21:10 FROGGS_ also, and IO::Socket::INET
21:10 netstar where can i find her??
21:10 FROGGS_ her?
21:10 netstar IO::Socket* modules
21:11 tadzik they're not modules
21:11 tadzik except IO::Socket::SSL
21:11 FROGGS_ here is the source if you are interested: https://github.com/rakudo/r​akudo/tree/nom/src/core/IO
21:11 netstar thanks
21:13 netstar Hmm I have Debian's rakudo installed
21:14 netstar use IO::Socket::INET ... cant be found
21:15 [Coke] FROGGS_: doing a fresh build with master...
21:16 moritz netstar: you don't need to 'use' it; it's already there
21:17 netstar ahar!
21:17 netstar :)
21:20 FROGGS_ [Coke]: nice, please ping me when you've got results :o)
21:23 [Coke] . tell lizmat - if we're doing a 6.0 release in 2015, we might want to pull the trigger on any pending deprecations before then. (right now some of them are scheduled in Q3 of 2015)
21:23 [Coke] .tell lizmat - if we're doing a 6.0 release in 2015, we might want to pull the trigger on any pending deprecations before then. (right now some of them are scheduled in Q3 of 2015)
21:23 yoleaux [Coke]: I'll pass your message to lizmat.
21:24 [Coke] .tell lizmat - this might mean updating the info in the release before we cut the release this week. (Not sure if any of these dates have gone out in a release yet.)
21:24 yoleaux [Coke]: I'll pass your message to lizmat.
21:28 moritz t/spec/S02-types/array.rakudo.moar fails three tests for me, and t/spec/S02-types/lists.rakudo.moar fails 5 tests
21:28 moritz both repeatable
21:29 moritz and TimToady++ touch both today
21:29 netstar okay sorry for the newbieness. Why is my perl6 program sticking on old behaviours whereas the source code is modified
21:30 TimToady moritz: testing with an old rakudo?
21:30 ugexe netstar: youll have to be more specific. is it a module from a script you are running?
21:31 TimToady or happened to hit the gap between updating rakudo vs roast?
21:31 firefish5000 joined #perl6
21:31 Guest39179 joined #perl6
21:32 TimToady moritz: fresh download here is All Tests Successful, at least under mora
21:32 TimToady *moar
21:37 colomon joined #perl6
21:38 TimToady m: sub foo { @_[-1] }
21:38 camelia rakudo-moar 10aa8b: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/0APWFlNlUY�Unsupported use of a negative -1 subscript to index from the end; in Perl 6 please use a function such as *-1�at /tmp/0APWFlNlUY:1�------> [32msub foo { @_[-1][33m�[31m }[0m�»
21:38 [Coke] netstar: perhaps gist the source?
21:38 TimToady m: say (1,2,3)[-5]
21:38 camelia rakudo-moar 10aa8b: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/iHQlCnqvir�Unsupported use of a negative -5 subscript to index from the end; in Perl 6 please use a function such as *-5�at /tmp/iHQlCnqvir:1�------> [32msay (1,2,3)[-5][33m�[31m<EOL>[0m�»
21:39 TimToady m: say (1,2,3)[*-5]
21:39 camelia rakudo-moar 10aa8b: OUTPUT«Calculated index (-2) is negative, but Parcel allows only 0-based indexing␤  in method gist at src/gen/m-CORE.setting:13776␤  in sub say at src/gen/m-CORE.setting:16507␤  in block <unit> at /tmp/AcpKhHxCuF:1␤␤»
21:39 raiph joined #perl6
21:40 TimToady m: my @a; say @a[@a.end]
21:40 camelia rakudo-moar 10aa8b: OUTPUT«Calculated index (-1) is negative, but Array allows only 0-based indexing␤  in method gist at src/gen/m-CORE.setting:13776␤  in sub say at src/gen/m-CORE.setting:16507␤  in block <unit> at /tmp/CsI5Ydt6jo:1␤␤»
21:41 TimToady m: my @a; say @a[ 10 .. -20 ]
21:41 camelia rakudo-moar 10aa8b: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/eTTf1nkK5C�Unsupported use of a negative -20 subscript to index from the end; in Perl 6 please use a function such as *-20�at /tmp/eTTf1nkK5C:1�------> [32mmy @a; say @a[ 10 .. -20 ][33m�…»
21:43 ghostlines joined #perl6
21:44 lizmat joined #perl6
21:45 [Coke] "function"?
21:45 FROGGS_ m: say 42[*-2] # :P
21:45 camelia rakudo-moar 10aa8b: OUTPUT«Calculated index (-1) is negative, but Int allows only 0-based indexing␤  in method gist at src/gen/m-CORE.setting:13776␤  in sub say at src/gen/m-CORE.setting:16507␤  in block <unit> at /tmp/5d1DfgMniC:1␤␤»
21:45 FROGGS_ m: say 42[*-1]
21:45 camelia rakudo-moar 10aa8b: OUTPUT«42␤»
21:46 bjz_ joined #perl6
21:47 TimToady m: say (*-2).arity
21:47 camelia rakudo-moar 10aa8b: OUTPUT«WhateverCode.new()␤»
21:47 TimToady heh
21:48 TimToady m: say .arity given *-2
21:48 camelia rakudo-moar 10aa8b: OUTPUT«1␤»
21:48 TimToady sure looks like a function to me
21:48 gtodd joined #perl6
21:49 colomon m: say (*-2).signature.arity
21:49 camelia rakudo-moar 10aa8b: OUTPUT«WhateverCode.new()␤»
21:49 colomon m: (*-2).signature.arity.say
21:49 camelia rakudo-moar 10aa8b: ( no output )
21:49 TimToady methods are subject to autopriming
21:49 * colomon has no idea what that means
21:50 TimToady *.foo doesn't call foo on the *
21:51 TimToady m: say (*.abs.floor.sqrt)(-123)
21:51 camelia rakudo-moar 10aa8b: OUTPUT«11.0905365064094␤»
21:51 colomon ah!
21:51 FROGGS_ it is like a placeholder parameter
21:51 TimToady m: say (my $ = *-2).arity
21:51 camelia rakudo-moar 10aa8b: OUTPUT«1␤»
21:51 woolfy joined #perl6
21:51 FROGGS_ m: say (* * *)(21, 2)
21:51 camelia rakudo-moar 10aa8b: OUTPUT«42␤»
21:52 FROGGS_ :D
21:52 TimToady basically * autocurries only with prefixes/infixes/postfixes, but method calls are considered postfixes
21:53 ghostlines joined #perl6
21:56 [Coke] Yes, I know that whatevercode is a function, but it doesn't seem like the most helpful word there for a n00b.
21:57 * FROGGS_ .oO( "Nah, Don't Do *That*!" )
21:58 TimToady it tells them precisely what to say, and teaches them it's really a function.  what's not t olike
21:59 TimToady it was 'closure', but I decided that was worse, especially since this one isn't closing over anything
21:59 [Coke] FROGGS_: aborted even more tests. sorry.
22:00 [Coke] TimToady: ok.
22:05 FROGGS_ ewww
22:08 [Coke] (did a default build with jit)
22:11 martinjungblut joined #perl6
22:20 bloonix joined #perl6
22:22 rurban joined #perl6
22:42 colomon joined #perl6
22:43 bronco_creek joined #perl6
22:44 bronco_creek I have a beginner-ish syntax question.
22:44 regreg joined #perl6
22:47 bronco_creek In a program $obj.WHAT.say says "(list)" and $obj.Int.say says "4". But   for $obj -> $item { $item.say }  says all four items concatenated into one string.  Not what I expected.
22:47 timotimo it's because it is in a $ container
22:47 timotimo that forces it into item context
22:47 timotimo and thus the for loop will only iterate once and have the whole $obj in the single iteration
22:48 timotimo you can write @$obj or $obj.list to get the individual items
22:48 * timotimo goes to bed
22:48 timotimo oh, i should mention:
22:48 cognominal joined #perl6
22:49 timotimo the good thing about this is that you can always tell that "for $obj" will only give a single iteration, whatever the value of $obj is
22:49 timotimo whereas "for @something" will give as many iterations as are in @something
22:49 timotimo o/
22:51 netstar what is the preferred method of writing to a socket?
22:52 netstar I come from a C background
22:53 lizmat computer, messages
22:53 yoleaux 21:23Z <[Coke]> lizmat: - if we're doing a 6.0 release in 2015, we might want to pull the trigger on any pending deprecations before then. (right now some of them are scheduled in Q3 of 2015)
22:53 yoleaux 21:24Z <[Coke]> lizmat: - this might mean updating the info in the release before we cut the release this week. (Not sure if any of these dates have gone out in a release yet.)
22:54 lizmat [Coke]: last months release had the first dates in them
22:54 lizmat but I will change the wording to something like "or when Perl 6.0.0 comes out"
22:55 lizmat [Coke]: I also see unclean spectests for weeks now (on OS X)
22:55 lizmat there is no pattern in the failing test-files that I could determine  :-(
22:55 bronco_creek timotimo: Thank you.
22:55 lizmat when run by themselves, they're 99.5% of the time ok  :-(
22:56 Timbus netstar, IO::Socket i would imagine
23:00 Timbus it needs work =/
23:02 Timbus personally, with async implemented as well as it is, I'd go so far as to recommend making it the only (core) way to work with sockets
23:05 kurahaupo joined #perl6
23:06 bronco_creek Perl6 just told me to "Please contact the author to have these calls to deprecated code adapted..."  I wonder how.
23:10 lizmat bronco_creek: could you tell me the exact message ?
23:13 bronco_creek lizmat: %ANON_VAR__2 = itemized hash called at: src/gen/m-CORE.setting, line 1634 Please use %ANON_VAR__2 = %(itemizedd hash) instead.
23:17 lizmat hmmm... I guess I need more info  :-(
23:19 netstar this is fun
23:21 bronco_creek I guess there is a way from me to cut and past the code, but I don't know how.
23:22 bronco_creek Bumbling n00bs are at least good at breaking stuff...
23:22 raydiak bronco_creek: if you're using the windows command line, I believe you right-click and select "Mark" to highlight, then 'Copy' in the same menu to copy, or something along those lines
23:23 raydiak might have to right-click in the titlebar, i dunno, been a while since I used windows more than required
23:23 bronco_creek Yah, but won't that creat havoc if I just paste it into my irc client?
23:23 raydiak if you highlight a big block, yes, but at least even if you only highlight one line at a time, you don't have to manually re-type
23:24 raydiak for big blocks sometimes I paste the whole block into notepad to clean it up and re-copy
23:24 bronco_creek Rough outline:
23:27 bronco_creek use v6; use JSON::Path; my $data = {...largish JSON data structure omitted...} my $path = JSON::Path.new('$.store.*'); my $obj = $path.values($data); for %$obj.keys -> $key { $key.say }
23:30 bronco_creek I guess it is late for EU folks.  I'll need to start playing with Perl6 in my morning.
23:32 lizmat bronco_creek: yeah, sorry, was distracted
23:33 lizmat and about to go to bed...
23:33 lizmat sleep&
23:34 bronco_creek In case someone backlogs, here is the whole compiler message:
23:34 bronco_creek Saw 1 call to deprecated code during execution. ========================================​========================================  %ANON_VAR__2 = itemized hash called at:   src/gen/m-CORE.setting, line 1643 Please use %ANON_VAR__2 = %(itemized hash) instead. ----------------------------------------​----------------------------------------  Please contact the author to have these calls to deprecated code adapted, so that this mess
23:37 ugexe i looked at that the other day, i think its from my &condition = EVAL '-> $_ { my $/; ' ~ ~$<code> ~ ' }';
23:38 ugexe you can do --notests if you dont plan on turning off safe mode (which i dont think is documented)
23:38 ugexe it passes all tests except for 2 safemode tests
23:40 bronco_creek In installed the module with --notests after it failed one test on my initial attempt.
23:43 ugexe ah i think i see the problem
23:44 ugexe its in the test so it should still function fine
23:49 bronco_creek Yes, it ran.  I would not have bothered the channel with this except that the comiler messages asked for follow-up.
23:50 ugexe i sent a PR to the author bronco_creek thanks for letting us know
23:51 bronco_creek ugexe: Thank you.
23:51 jack_rabbit joined #perl6
23:53 Alina-malina joined #perl6
23:53 BenGoldberg joined #perl6
23:55 netstar okay sorry to keep interrupting. I'm trying to use a scalar as a key for a hash, it seems to escape the $ sign each time
23:56 telex joined #perl6

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo