Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2011-01-30

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:00 bacek joined #perl6
00:01 Tene sorear: It very much reminds me of conflicts I've seen between dating and married couples, which is subtly hilarious to me.  "We never talk any more.  I don't feel that you understand my needs anymore.  I'm going to start seeing other people."
00:01 Tene :)
00:01 sorear yeah I laughed at the couples therapy line.
00:02 Tene I don't think it's a bad metaphor for trying to address the issues, either.
00:05 bacek left #perl6
00:05 masak left #perl6
00:06 masak joined #perl6
00:10 TiMBuS left #perl6
00:11 TiMBuS joined #perl6
00:11 [Coke] made sense to me.
00:14 patspam joined #perl6
00:17 coldhead it's sad
00:20 pmurias masak: re ++foo --option ++/foo i used to have --option before that, ++foo ++/foo seems to specify what goes to which subpart more specifically
00:20 pmurias masak: as for example different backends can have vastly different options
00:21 masak I'm not sure I see that as a direct answer to my question.
00:21 masak it's more of an elucidation of S19, which I don't think I'm less-than-clear on.
00:25 * shortcircuit sets out to begin translating/rewriting http://rosettacode.org/wiki/RCRPG/Perl to Perl 6.
00:26 pmurias masak: what are the drawbacks of ++foo?
00:26 * shortcircuit wonders what he should use in place of Term::ANSIColor.
00:27 pmurias masak: implementing that isn't very hard
00:27 patspam left #perl6
00:27 pmurias masak: and it seems cleaner
00:27 patspam joined #perl6
00:28 lopaway is now known as lopnor
00:28 pmurias masak: the stuff i wanted to pass to mildew's backends ended up very strange
00:28 colomon shortcircuit: I vote you just implement Tem::ANSIColor first.  ;)
00:29 pmurias like options to the underlying c compiler and stuff
00:29 shortcircuit colomon: I guess I could do that. Though I haven't implemented a P5 module in almost a decade, and have even less of an idea how to do it in P6.
00:29 pmurias masak: i could have --backend-cflags but seems inelegant
00:30 colomon shortcircuit: it's always good to learn.  :)
00:30 masak pmurias: [drawbacks of ++foo] as I see them: specific to Perl 6, no rooting in Unix, introduces mental overhead in the user, introduces model overhead in the implementation.
00:30 pmurias why mental overhead to the user?
00:31 masak I assume openers and closers have to match up, and not mis-nest.
00:31 masak there's mental overhead in adhering to those invariants.
00:31 masak just as it is with writing correct XML or whatever.
00:31 shortcircuit Is there a construct where something like "my %values :="... would make sense. As, "look this up in this hash, if it doesn't exist, run this code to generate it, and cache the result"?
00:31 masak there are not as many ways to go wrong with classical command-line options.
00:32 pmurias how have to parse compound names then
00:32 shortcircuit Basically, I've got a scenario where I precalculate a hash's contents based on a transformation of another hash's contents, and I'd prefer to do that lazily.
00:32 pmurias so like using prefixes instead of namespaces
00:32 pmurias like in C
00:33 pmurias i could live with --backend-* and --parser-* etc.
00:33 sorear shortcircuit: %hash{$key} //= calculate()
00:33 pmurias just when i found S19 i was looking exactly for something like that
00:34 sorear it's called a shortcirtcuit assignment operator :)
00:34 shortcircuit Oh god...
00:34 colomon sorear++
00:34 pmurias masak: ghc uses something like that too
00:35 colomon rakudo: my %h; %h<blue> //= sin(10); %h<blue> //= 1/0; say %h.perl
00:35 p6eval rakudo 549d2a: OUTPUT«{"blue" => -0.54402111088937}␤»
00:35 masak pmurias: tell me more.
00:35 colomon rakudo: my %h; %h<blue> //= sin(10); %h<blue> //= (1..*).elems; say %h.perl
00:35 masak (this sounds familiar for some reason)
00:35 p6eval rakudo 549d2a: OUTPUT«{"blue" => -0.54402111088937}␤»
00:35 colomon rakudo++
00:35 sorear haskell_program +RTS -N2 -RTS  # run program on 2 cores
00:36 sorear actually -RTS can be omitted if it would be at the end.
00:38 colomon rakudo: my %h; %h<blue> //= sin(10); %h<blue> //= (1...*).elems; say %h.perl
00:38 p6eval rakudo 549d2a: OUTPUT«(timeout)»
00:38 colomon rakudo--
00:38 colomon shortcircuit: see above.  I don't think //= shortcircuits properly in Rakudo.
00:38 colomon would love to be wrong about that....
00:41 drbean joined #perl6
00:42 shortcircuit colomon: That looks like a complicated run, which is a little more than what I'm addressing for the moment. (I don't have an infinite series anywhere here...)
00:42 wallberg left #perl6
00:43 sorear perl6: my %h; %h<blue> //= say(1); %h<blue> //= say(2);
00:43 p6eval pugs, niecza v1-203-gbc5d7f7: OUTPUT«1␤»
00:43 p6eval ..rakudo 549d2a: OUTPUT«1␤2␤»
00:43 masak there you go.
00:43 * masak submits rakudobug, Justin Case
00:44 shortcircuit There wouldn't happen to be a fallback or special hash key, for a 'key not found' scenario? That would be ideal for what I'd like to do.
00:46 masak shortcircuit: KeyHash allows you to set a default.
00:46 masak but it's NYI in Rakudo.
00:46 * shortcircuit nods.
00:48 kst left #perl6
00:48 * shortcircuit .oO( One could create a ridiculously simple caching proxy using shortcircuit operators, default key values and a filestore backing to Tie to.)
00:50 masak well volunteered!
00:51 shortcircuit Heh. I'm going to stick with my current P5->P6 translation of a terminal-mode game, for now. :)
00:51 kst joined #perl6
00:57 cafesofie left #perl6
00:59 masak 'night, #perl6
00:59 masak left #perl6
01:03 Trashlord left #perl6
01:03 pmurias left #perl6
01:06 sorear shortcircuit: undefined values are intended to support the "key not found" case
01:07 sorear any time you think you want to distinguish "key not present" from "key is undef", you're probably doing it wrong.
01:10 whiteknight joined #perl6
01:18 sorear hello whiteknight
01:24 whiteknight hello
01:31 shi joined #perl6
01:35 GinoMan left #perl6
01:35 patspam_ joined #perl6
01:38 bemjb left #perl6
01:39 patspam left #perl6
01:39 patspam_ is now known as patspam
01:44 snarkyboojum left #perl6
01:55 aindilis joined #perl6
01:59 VXZ left #perl6
02:13 aesop left #perl6
02:20 aesop joined #perl6
02:46 snarkyboojum joined #perl6
02:48 kst left #perl6
02:49 kst joined #perl6
02:59 noganex joined #perl6
03:02 noganex_ left #perl6
03:10 shi left #perl6
03:24 flussence_ joined #perl6
03:27 flussence left #perl6
03:40 whiteknight left #perl6
03:58 patspam left #perl6
04:01 patspam joined #perl6
04:12 JimmyZ joined #perl6
04:12 patspam left #perl6
04:26 Su-Shee left #perl6
04:28 Su-Shee joined #perl6
04:38 shortcircuit Hm.
04:38 sorear Yes?
04:39 shortcircuit How do I declare constructor arguments for a class?
04:39 shortcircuit (Actually, I'm using class ClassName { ..... }, so I need to see how to declare the implementation, too.)
04:39 sorear Every attribute is a constructor argument
04:39 shortcircuit Ah. Convenient.
04:40 sorear niecza: class Foo { has $.bar }; say Foo.new(bar => 12).bar
04:40 p6eval niecza v1-203-gbc5d7f7: OUTPUT«12␤»
04:40 sorear If you want to accept arguments differently, just override new in your class
04:40 sorear niecza: class Foo { has $.bar; method new($x) { nextwith(self, bar => $x) } }; say Foo.new(12).bar
04:41 p6eval niecza v1-203-gbc5d7f7: OUTPUT«12␤»
04:41 kst left #perl6
04:42 shortcircuit Any advantage to Sweet. Thanks.
04:44 kst joined #perl6
04:44 shortcircuit If I have a case where I could take method($var) {...}, and handle $var with an if/else latter, I could just as easily write it using a set of multis and no ladder. Any particular considerations for favoring one approach over the other?
04:45 shortcircuit (Also, can the overridden new() be a multi instead of a method?)
05:04 sorear multis and methods, per se, cannot substitute for each other
05:04 sorear you can however use a proto, which is just an ordinary method
05:04 sorear any place you can use a method, you can use a proto + a set of methods
05:04 sorear set of multis
05:04 shortcircuit I wound up using a couple batches of multis to defer to from within the method body.
05:04 sorear if you don't have the proto, then things get (not understood by me)
05:05 sorear do you know about given/when?  it's designed to replace if/else ladders
05:05 shortcircuit Not familiar with them.
05:06 sorear rakudo: given 12 { when 10 { say "A" }; when 12 { say "B" }; when 14 { say "C" } }
05:06 p6eval rakudo 549d2a: OUTPUT«B␤»
05:10 shortcircuit This is what I've written so far. http://pastebin.com/BS0SMcHT
05:10 shortcircuit (I haven't tested to see that it runs)
05:10 araujo joined #perl6
05:11 felliott left #perl6
05:12 dalek niecza: ffcfc4a | sorear++ | docs/nam.pod:
05:12 dalek niecza: [nam.pod] Finish documenting low-level format
05:12 dalek niecza: review: https://github.com/sorear/niecza/commit/ffcfc4a79b
05:13 sorear shortcircuit: what you want is ... a hash
05:14 shortcircuit Heh
05:14 sorear shortcircuit: also, I don't recommend assigning (,,,) lists to $variables as this is a rather dark corner of the spec.  [,,,] is much better defined.
05:15 shortcircuit The problem I'm dealing with for the moment is coming up with a nicer way to deal with %directions and %rdirections in http://rosettacode.org/wiki/RCRPG/Perl#Code
05:16 shortcircuit Anyway, I'm overdue for bed. I obviously have some reading to do when I get up. :)
05:17 envi joined #perl6
05:17 sorear I'll try my own hand at that task tonight
05:17 shortcircuit Heh; sweet.
05:19 shortcircuit I haven't written the ANSI module yet, but I figure I'll wind up doing it, just for greater familiarity with consoles. (I'd like to see if I can query termcaps for it; just outputting raw codes assuming a 20-plus-year-old spec is being honored on the client end seems optimistic these days.
05:20 felliott joined #perl6
05:23 cafesofie joined #perl6
05:27 dukeleto left #perl6
05:27 dukeleto joined #perl6
05:39 kaare_ joined #perl6
05:44 chillroses joined #perl6
05:44 chillroses good morning
05:45 sorear shortcircuit: querying termcap is dangerous these days; with a proliferation of slightly different terminals, and ssh, it's increasingly unlikely that the correct terminal description will be available
05:45 sorear hello and welcome chillroses
05:46 chillroses thank you
05:46 chillroses iam a newbee
05:47 chillroses would like to know the differences of 5 and 6
05:47 sorear shortcircuit: I, uh, comaintain a couple of telnet gaming servers and have had no end of trouble with PuTTy in particular
05:47 sorear shortcircuit: PuTTY calls itself "xterm" but defaults to different keypad codes than the ones in xterm-r7's terminfo file
05:48 sorear shortcircuit: postel's prescription is a much safer way to work with terminals, than libtermcap
06:05 satyavvd joined #perl6
06:06 * mberends -> moritz_
06:06 mberends left #perl6
06:15 justatheory left #perl6
06:18 chillroses left #perl6
06:38 lopnor is now known as lopaway
06:39 kst left #perl6
06:44 kst joined #perl6
06:45 kaare_ left #perl6
06:48 satyavvd left #perl6
06:54 felliott left #perl6
07:15 drbean left #perl6
07:16 sorear Who wrote RCRPG/Perl?
07:16 sorear This is ... very unidiomatic code.
07:35 TiMBuS timtaswtdi
07:58 sorear phenny: tell colomon &postfix:<L>
07:58 phenny sorear: I'll pass that on when colomon is around.
08:03 rgrau left #perl6
08:08 Trashlord joined #perl6
08:27 am0c joined #perl6
08:32 Psyche^ joined #perl6
08:33 Patterner left #perl6
08:33 Psyche^ is now known as Patterner
08:38 kst left #perl6
08:41 kst joined #perl6
08:42 sorear RCRPG has found two bugs so far in niecza
08:42 sorear but now I must sleep
08:42 * sorear out
08:48 drbean joined #perl6
08:50 cxreg what happened to rakudo.org?
08:50 cxreg other virtualhosts on that server are responding but that one just dies
08:58 Mowah joined #perl6
09:05 JimmyZ left #perl6
09:45 moritz_ good morning
09:47 masak joined #perl6
09:47 masak hola, zebras.
09:51 _118 joined #perl6
09:51 kjeldahl joined #perl6
09:57 kst left #perl6
09:58 kst joined #perl6
10:01 moritz_ http://www.perl-community.​de/bat/poard/thread/15903 a proposal for a pythonic frontend for Perl 6 (in German)
10:02 kst left #perl6
10:03 kst joined #perl6
10:18 _118 left #perl6
10:18 shi joined #perl6
10:18 _118 joined #perl6
10:20 Su-Shee left #perl6
10:22 Ooblick joined #perl6
10:24 masak I've had a similar idea -- give Perl 6 a mostly Pythonic syntax -- and I'm basically waiting for more advanced parser hooks to try it out.
10:25 masak but... getting rid of the sigils? I don't think that's easily done, and besides I don't think it's a good idea.
10:25 tadzik I don't think the lack of sigils makes stuff more readable
10:25 moritz_ masak: indeed, that's what I argued in my reply
10:26 masak oh, haven't read the reply.
10:26 moritz_ masak: I should blog about the various aspects of sigils, and why removing them isn't easily done
10:26 moritz_ https://github.com/moritz/nqp-cl-parser if you want to look at some funny command line parsing code :-)
10:27 masak moritz_: right, good reply. you brought up more objections than I'd found so far.
10:27 * masak looks expectantly :)
10:28 moritz_ well, I spent a few minutes thinking about it :-)
10:28 masak :)
10:28 masak by the way, http://modules.perl6.org/ has an encoding problem. look at the line with perl6-literate.
10:28 masak I'm pretty sure it's not github that has the encoding problem.
10:28 tadzik masak: btw, there's no META.info for perl6-literate
10:29 tadzik just saying. BTW, >50% of modules use the new infrastructure now
10:29 tadzik now that masak++ has moved his
10:30 masak moritz_: er. variable on https://github.com/moritz/nqp-​cl-parser/blob/master/do.pl#L9 , where is it declared?
10:30 moritz_ you mean $x?
10:30 moritz_ it's a formal parameter
10:31 tadzik I have a plenty of plans to launch some module-related services on feather when my exam session is over
10:31 xinming_ joined #perl6
10:31 masak tadzik: huh? I see https://github.com/masak/perl6-literate/comm​it/736e5e3d422ac6d4627abb23da0c9feb59144386 online, don't you?
10:32 masak moritz_: no, $!stop-after-first-arg
10:32 masak which looks very much like an attr.
10:32 tadzik masak: oh, must've missed it
10:33 masak moritz_: line 9, as I linked to.
10:33 xinming left #perl6
10:33 moritz_ masak: oh, I can't read
10:34 masak moritz_: how can you use an attribute without declaring it?
10:34 moritz_ masak: it's broken, I renamed the attribute to $!b for testing purposes, and accidentally committed
10:34 masak oh :(
10:35 moritz_ masak: fix pushed... i wanted to find out more about the WTF on line 165
10:35 masak lines 16 and 17... sometimes I think about how to write that using real, encapsulated Perl 6 OO.
10:35 masak as they are now, I think they leak read-writeness.
10:37 moritz_ they do, in nqp-rx
10:37 moritz_ they don't in proper Perl 6
10:38 masak are you sure? I think the spec says somewhere that ro-ness only goes down "one level".
10:38 * moritz_ sees only one level
10:38 masak I'm talking about the general case.
10:39 masak a .deep-clone would be needed to do it right in the general case.
10:39 moritz_ well, if you have objects inside the attribute, ... right
10:39 masak objects or no, I wouldn't want to hand out a reference to *anything* that's only meant to be changed from within the object,
10:40 dalek mu: 1288837 | moritz++ | misc/dalek-conf.json:
10:40 dalek mu: [dalek-conf] fix channel for nqp-cl-parser
10:40 dalek mu: review: https://github.com/perl6/mu/commit/1288837d33
10:41 masak Josh Bloch calls it "making defensive copies". it's important both for getters and constructors/setters.
10:41 masak it's basically the OO twist on "user data can't be trusted".
10:44 * moritz_ nods, although he can't connect the name Josh Bloch to anything
10:46 masak moritz_: "Effective Java", "Java Puzzlers", Java Collections, java.math :)
10:47 moritz_ masak: that explains it - I'm not strong in the Java universe
10:48 masak moritz_: [nqp-cl-parser] looks very nice.
10:48 moritz_ masak: thanks
10:58 Ooblick left #perl6
11:00 am0c left #perl6
11:02 flatwhatson_ joined #perl6
11:17 moritz_ phenny: tell felliott that he has commit access to perl6/specs too, and is welcome to commit obvious fixes himself
11:17 phenny moritz_: I'll pass that on when felliott is around.
11:20 dalek specs: a68bbb1 | (Fitz Elliott)++ | S19-commandline.pod:
11:20 dalek specs: [S19] break up unintentional pod formatting code
11:20 dalek specs: review: https://github.com/perl6/specs/commit/a68bbb14c0
11:20 dalek specs: e144eac | (Fitz Elliott)++ | S32-setting-library/Containers.pod:
11:20 dalek specs: [S32:Containers] fix broken link to Type Declarations, now in S29
11:20 dalek specs: review: https://github.com/perl6/specs/commit/e144eacb29
11:20 * moritz_ was just about to commit that too
11:23 rgrau joined #perl6
11:25 _118 left #perl6
11:36 shi left #perl6
11:47 whiteknight joined #perl6
11:57 kst left #perl6
11:58 kst joined #perl6
11:59 dalek nqp-rx/nom: 29337fb | moritz++ | src/metamodel/reprs/P6opaque.c:
11:59 dalek nqp-rx/nom: distinguish "P6opaque attributes" error messages (no additional value for the user)
11:59 dalek nqp-rx/nom: review: https://github.com/perl6/nqp-rx/commit/29337fb49b
12:00 shi joined #perl6
12:14 xinming joined #perl6
12:17 xinming_ left #perl6
12:37 cjk101010 left #perl6
12:43 cjk101010 joined #perl6
12:48 rgrau left #perl6
12:55 kjeldahl left #perl6
13:11 * masak misses hugme
13:19 coldhead left #perl6
13:24 flatwhatson_ left #perl6
13:42 drbean left #perl6
13:51 rgrau joined #perl6
13:52 am0c joined #perl6
13:52 orafu joined #perl6
13:52 MayDaniel joined #perl6
13:52 kst left #perl6
13:53 kst joined #perl6
13:56 whiteknight left #perl6
13:58 felliott joined #perl6
14:05 wamba joined #perl6
14:22 f00li5h is now known as _0000
14:23 _0000 is now known as f00li5h
14:29 colomon sorear++
14:29 phenny colomon: 07:58Z <sorear> tell colomon &postfix:<L>
14:30 colomon testing it now.
14:31 moritz_ masak: it seems that hugme's auto-rejoin doesn't work well on netsplits
14:31 hugme joined #perl6
14:35 colomon pushed.
14:35 felliott Thanks, moritz_!
14:35 phenny felliott: 11:17Z <moritz_> tell felliott that he has commit access to perl6/specs too, and is welcome to commit obvious fixes himself
14:36 MayDaniel left #perl6
14:36 colomon looks like causality is going backwards on #perl6 this morning.  ;)
14:36 felliott I live backwards in time.
14:36 felliott It's a problem.
14:37 xinming left #perl6
14:39 cafesofie left #perl6
14:39 xinming joined #perl6
14:40 shortcircuit sorear: Yeah, I'm familiar with PuTTY troubles. I use PuTTY everwhere I use Windows, and it's royal PITA to get things set up properly.
14:41 shortcircuit sorear: I wrote RCRPG/Perl. I've been hoping someone would clean it up for ages.
14:43 mikehh joined #perl6
14:43 shortcircuit felliott: So it goes.
15:00 mkramer joined #perl6
15:09 xinming_ joined #perl6
15:13 xinming left #perl6
15:14 am0c left #perl6
15:18 mkramer1 joined #perl6
15:19 mkramer left #perl6
15:35 kst left #perl6
15:36 kst joined #perl6
15:39 felliott left #perl6
15:41 jnthn Good afternoon, #perl6
15:41 phenny jnthn: 29 Jan 19:33Z <moritz_> tell jnthn that I get a nasty "P6opaque attributes NYFI" error when I uncomment the hilighted line (165) in https://github.com/moritz/nqp-cl-parser/blob/683​2591a1597ecd34c250257f8ae51fd31a3b9de/do.pl#L165 in what seems like a simple scalar attribute access - any idea what's wrong?
15:42 colomon o\
15:42 jnthn moritz_: I commented on the commit but I think it's just that nqp-rx doesn't parse - in attribute names, and the undeclared attribute check at compile time is NYI.
15:42 jnthn Or maybe something like that
15:42 jnthn Maybe not though
15:42 jnthn o/ colomon
15:42 * jnthn will take a look
15:43 satyavvd joined #perl6
15:44 * jnthn admires the tea selection in his hotel room
15:45 felliott joined #perl6
15:47 cjk101010 left #perl6
15:52 envi left #perl6
15:52 envi joined #perl6
15:54 sorear good * #perl6
15:58 wamba left #perl6
15:58 felliott left #perl6
15:58 moritz_ jnthn: I replied to your comment... change the attribute to $!foo doesn't fix it
15:59 jnthn moritz_: OK, then it's a $weird_bug
15:59 felliott joined #perl6
15:59 moritz_ jnthn: I've added numbers to the error messages to distinguish the cases... it's the first occurence in the file that throws the error
16:00 jnthn Thanks, that's helpful.
16:00 sorear shortcircuit: https://gist.github.com/802968 # My version of RCRPG/Perl 6.  Contains several workarounds for bugs not present in Rakudo
16:01 jnthn moritz_: Just getting an up to date build on the laptop. :)
16:02 jnthn Didn't hack on this machine for a month. :)
16:09 shortcircuit sorear: Wow; there's a lot of syntax in here I'll need to wrap my head around.
16:14 colomon sorear: postfix:<L> was a lovely suggestion, and I have made it so.
16:15 sorear shortcircuit: I will answer any requests for clarification
16:19 [Coke] rakudo: my @a = 1..4; say @a[1..*].perl
16:19 p6eval rakudo 549d2a: OUTPUT«(timeout)»
16:19 moritz_ @a[1..*-1] works
16:19 [Coke] that's new.
16:20 [Coke] ah. not that new. (RT #64566)
16:20 sorear niecza: my %foo; if !%foo<hi>:exists { say "..." } # Is this correct?
16:20 p6eval niecza v1-204-gffcfc4a: OUTPUT«Unhandled exception: Excess arguments to SAFE prefix:<!>, unused named exists␤  at /home/p6eval/niecza/lib/SAFE.setting line 0 (SAFE prefix:<!> @ 0)␤  at /tmp/RTr5rVn9TU line 1 (MAIN mainline @ 2)␤  at /home/p6eval/niecza/lib/SAFE.setting line 1037 (SAFE C432_ANON @
16:20 p6eval ..1)␤  at /home/p…
16:20 * [Coke] wishes that he could use gmail style keyboard shortcuts in RT. :P
16:21 moritz_ sorear: I don't think so
16:21 moritz_ saying ... would be the correct answer
16:23 shortcircuit sorear: sub command() is a very strange beast to me, or, at least, the subsequent lines of command [supported args], (string), number, -> $var { block }
16:24 sorear It's just a function with four arguments
16:25 sorear -> sig { block } is just a lambda
16:29 jnthn moritz_: ping
16:30 moritz_ jnthn: pong
16:31 jnthn moritz_: How do I reproduce the failure?
16:31 jnthn moritz_: If I take https://github.com/moritz/nq​p-cl-parser/raw/master/do.pl and run it, I get all 17 passing.
16:31 [Coke] where would a test for / <[a-z]> /  go? S05-?
16:31 moritz_ jnthn: yep, you need to comment out line 165
16:31 [Coke] metachars?
16:32 jnthn moritz_: d'oh :)
16:33 moritz_ [Coke]: S05-metasyntax/charset.t probably
16:33 [Coke] moritz_: danke.
16:33 moritz_ [Coke]++
16:34 jnthn moritz_: OK, reproduce it.
16:34 jnthn *reproduced
16:35 jnthn moritz_: I'm confused.
16:35 jnthn moritz_: $!stop-after-first-arg is declared as an attribute in CLIParseResult
16:35 jnthn moritz_: But you try to access it in CommandLineParser
16:35 moritz_ jnthn: d'oh
16:36 moritz_ jnthn: when I move it to CommandLineParser, I get "P6opaque attributes NYFI 4"
16:36 moritz_ previously 1
16:36 moritz_ oh wait
16:36 moritz_ I do the initialization in the wrong class too
16:36 jnthn moritz_: You have an accessor for it in CLIParseResult
16:37 moritz_ jnthn: it should all be in the parser class... and works then
16:37 jnthn moritz_: OK
16:37 moritz_ so it seems my problem is just with imperfect error reporting
16:38 jnthn moritz_: Yeah, what should happen is that it tells you at compile time, "hang on...that attr doesn't exist"
16:38 jnthn Will add that check soon. :)
16:39 moritz_ \o/
16:40 jnthn Anyway, glad I've not got a really super-obscure  bug to find. :)
16:40 moritz_ it just needed another pair of eyes :-)
16:40 [Coke] there is no builtin Test sub for checking an exception message, yes?
16:41 [Coke] (so i have to handroll an eval and check the message?)
16:41 moritz_ correct :(
16:42 jnthn No built-in Test sub in...which test module/language/project? :)
16:42 shortcircuit sorear: Any problem with me taking the gist and putting it on RC?
16:43 moritz_ jnthn: Perl 6
16:44 moritz_ jnthn, [Coke] I plan to have a throws_ok which checks the kind of exception that's thrown
16:44 jnthn I thought we had dies_ok?
16:44 [Coke] moritz_: hurm. / <[a-z]> / seems to not be catchable by try {}
16:44 moritz_ [Coke]: it's a compile time bug
16:44 moritz_ and try { } is run time
16:44 moritz_ so you need eval ''
16:45 [Coke] *facepalm*
16:45 [Coke] danke.
16:45 [Coke] jnthn: *dies_ok just checks to see if $! was set.
16:48 [Coke] is there a better way to reset $! than try { 1 } ?
16:50 [Coke] (is lexically scoping it sufficient?
16:51 sorear you need eval_dies_ok
16:52 [Coke] eval_dies_ok doesn't check the value of $! either, methinks.
16:52 sorear shortcircuit: no, although it seems like for consistency it should be ported to Rakudo
16:53 satyavvd left #perl6
16:54 [Coke] is there a doc on writing spec tests? like "here's how to fudge just that one test you rewrote so you can test things reasonably fast." ;)
16:55 shortcircuit sorear: I saw the TODO line at the top, but I'm not familiar enough with Perl6 or rakudo to know the bugs in question (or necessarily where they'd apply in the body of the code.)
16:56 shortcircuit sorear: Which implementation was this tested with?
16:56 envi left #perl6
16:58 kst left #perl6
16:59 sorear niecza
17:00 [Coke] moritz_: I cannot seem to construct something where /<[a-z]>/ is catchable.
17:01 [Coke] rakudo: eval "/<[a-z]>/"
17:01 p6eval rakudo 549d2a:  ( no output )
17:01 [Coke] ... that is not what it does locally. :P
17:01 shortcircuit sorear: Does Niecza have an project page somewhere describing it? (i.e. what it is, how it differs from other P6 implementations, etc)
17:02 [Coke] bah. behaves differently in REPL.
17:02 kst joined #perl6
17:03 * shortcircuit would like to get a P6 Implementation page properly set-up for Niecza.
17:04 shortcircuit Hm. https://github.com/sorear/niecza Looks like a good reference target.
17:07 sorear docs/announce.initial is probably the best executive summary
17:09 moritz_ rakudo: eval '/<[a-z]>/; say "alive, error was $!"
17:09 p6eval rakudo 549d2a: OUTPUT«===SORRY!===␤Confused at line 22, near "eval '/<[a"␤»
17:10 moritz_ rakudo: eval '/<[a-z]>/'; say "alive, error was $!"
17:10 p6eval rakudo 549d2a: OUTPUT«alive, error was Unsupported use of - as character range;in Perl 6 please use .. at line 1, near "z]>/"␤»
17:12 shortcircuit sorear: http://rosettacode.org/wiki/Niecza
17:12 shortcircuit Hack that up however needed.
17:14 sorear I should get an account, huh.
17:14 dalek niecza: b0c5160 | sorear++ | src/ (3 files):
17:14 dalek niecza: Mergeback
17:14 dalek niecza: review: https://github.com/sorear/niecza/commit/b0c5160089
17:14 dalek niecza: ae2f23a | sorear++ | src/STD.pm6:
17:14 dalek niecza: Fix crash in "strange text after block" error
17:14 dalek niecza: review: https://github.com/sorear/niecza/commit/ae2f23affd
17:14 shortcircuit sorear: Nothing stops you from editing anonymously, except a bunch more CAPTCHAs. :)
17:15 shortcircuit (The CAPTCHAs continue while logged in for a while, but die off after an internal metric is met.)
17:15 masak s/metric/threshold/
17:16 dalek nqp-cl-parser: 9bdd79d | moritz++ | do.pl:
17:16 dalek nqp-cl-parser: sometimes I am so stupid it hurts. jnthn++
17:16 dalek nqp-cl-parser: review: https://github.com/moritz/nqp​-cl-parser/commit/9bdd79de7a
17:16 masak moritz_: I know the feeling... :/
17:16 sorear shortcircuit: Where does the documentation still say "Perl 5"?
17:18 sorear ummm... I can't find the login link
17:18 sorear oh found it
17:18 MayDaniel joined #perl6
17:19 [Coke] wow. this is incredibly frustrating. ;)
17:22 [Coke] is there a way to get the harness to spit out "got FOO, expected BAR" ?
17:23 [Coke] ... ah. t/harness --fudge <file> doesn't actually refudge the file.
17:23 [Coke] ... no, it does. I give up. :)
17:24 MayDaniel left #perl6
17:24 kaare_ joined #perl6
17:25 ab5tract joined #perl6
17:28 moritz_ rakudo: use Test; is 'a', 'b';
17:28 p6eval rakudo 549d2a: OUTPUT«not ok 1 - ␤#      got: 'a'␤# expected: 'b'␤»
17:28 sorear shortcircuit: actually I suppose it will still be Perl 5 until tomorrow if you stick to released versions...
17:29 moritz_ [Coke]: that does give the diagnostic message
17:33 sorear shortcircuit: I meant to say bugs *not* in Rakudo
17:33 sorear shortcircuit: for instance
17:34 sorear rakudo: say ['gold' xx 9].perl # This is the best way to write the line
17:34 p6eval rakudo 549d2a: OUTPUT«["gold", "gold", "gold", "gold", "gold", "gold", "gold", "gold", "gold"]␤»
17:34 sorear niecza: say ['gold' xx 9].perl # Oops
17:34 p6eval niecza v1-206-gae2f23a: OUTPUT«[]␤WRONG ExitRunloop TAKEN:␤  at  line 0 (ExitRunloop @ 0)␤Correct:␤  at /home/p6eval/niecza/lib/SAFE.setting line 533 (SAFE List.perl @ 5)␤  at /tmp/ba2ss6hPQs line 1 (MAIN mainline @ 3)␤  at /home/p6eval/niecza/lib/SAFE.setting line 1037 (SAFE C432_ANON @ 1)␤  at
17:34 p6eval ../home/p6eval/nie…
17:34 sorear pugs: say ['gold' xx 9].perl # just curious
17:34 p6eval pugs: OUTPUT«["gold", "gold", "gold", "gold", "gold", "gold", "gold", "gold", "gold"]␤»
17:35 moritz_ where's the bug? in xx or in .perl?
17:36 sji joined #perl6
17:36 sorear moritz_: it appears to be in &map
17:37 shi left #perl6
17:39 ubunturocks joined #perl6
17:40 ubunturocks left #perl6
17:42 tadzik left #perl6
17:43 tadzik joined #perl6
17:43 ubunturocks joined #perl6
17:44 ubunturocks :)
17:44 ubunturocks perl6 gonna be mainstream soon?
17:45 sorear That's up to you.
17:45 ab5tract left #perl6
17:45 masak ubunturocks: it's good you came. we need your help.
17:47 masak we need people to test-drive Perl 6 before it goes mainstream. shake out some bugs, etc.
17:47 masak we have some people doing this, but it'd be great with one more.
17:47 masak ubunturocks: do you have what it takes? :)
17:47 ubunturocks left #perl6
17:48 cafesofie joined #perl6
17:48 masak are you willing and able?
17:48 sorear guess not.
17:48 jnthn ...he may, but his connection doesn't.
17:48 sorear oh.
17:48 * jnthn is hungry
17:48 jnthn bbiab
17:49 ponbiki left #perl6
17:50 ponbiki joined #perl6
17:54 justatheory joined #perl6
17:57 ab5tract joined #perl6
18:00 ab5tract left #perl6
18:15 kst left #perl6
18:17 kst joined #perl6
18:35 dalek niecza: 786a885 | sorear++ | lib/Builtins.cs:
18:35 dalek niecza: Fix lazy cases of map and grep
18:35 dalek niecza: review: https://github.com/sorear/niecza/commit/786a885f0c
18:40 cafesofie left #perl6
18:50 thundergnat joined #perl6
18:50 thundergnat Hello #perl6
18:51 thundergnat masak: BTW,  RT #83002 is a dup of RT #77420
18:52 thundergnat rakudo: my @g; @g[^2]».[^6] »=» 0; @g.perl.say
18:52 p6eval rakudo 549d2a: OUTPUT«[[0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0]]␤»
18:52 thundergnat rakudo: my @g; @g[^2]>>.[^6] >>+=>> 0; @g.perl.say;
18:52 p6eval rakudo 549d2a: OUTPUT«[[0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0]]␤»
18:52 thundergnat rakudo: my @g; @g[^2]>>.[^6] >>=>> 0; @g.perl.say;
18:52 p6eval rakudo 549d2a: OUTPUT«===SORRY!===␤Missing << or >> at line 22, near "> 0; @g.pe"␤»
18:53 thundergnat No it isn't :-/
18:53 thundergnat std: my @g; @g[^2]>>.[^6] >>=>> 0; @g.perl.say;
18:53 p6eval std 625303c: OUTPUT«[31m===[0mSORRY![31m===[0m␤Missing << or >> at /tmp/Bc93YUWpWg line 1:␤------> [32mmy @g; @g[^2]>>.[^6] >>=>[33m⏏[31m> 0; @g.perl.say;[0m␤    expecting infix_circumfix_meta_operator␤Parse failed␤FAILED 00:01 121m␤»
18:54 thundergnat Phooey.
18:55 thundergnat Seems kind of crappy that »=» is ok but >>=>> isn't.
18:59 thundergnat left #perl6
19:01 Su-Shee joined #perl6
19:02 dukeleto left #perl6
19:02 dukeleto joined #perl6
19:12 dalek nqp-cl-parser: 9c18ca4 | moritz++ | do.pl:
19:12 dalek nqp-cl-parser: test for stop-after-first-arg
19:12 dalek nqp-cl-parser: review: https://github.com/moritz/nqp​-cl-parser/commit/9c18ca4719
19:13 cafesofie joined #perl6
19:21 dual left #perl6
19:36 molaf joined #perl6
19:37 dukeleto left #perl6
19:38 dukeleto joined #perl6
19:39 orafu left #perl6
19:39 awwaiid left #perl6
19:40 awwaiid joined #perl6
19:41 kst left #perl6
19:41 orafu joined #perl6
19:41 kst joined #perl6
19:45 cafesofie left #perl6
19:55 envi joined #perl6
20:05 sorear moritz_, et al: What kind of attribution should I have when I steal functions from Rakudo's setting one at a time?
20:07 masak Rakudo's LICENSE: "You may Distribute verbatim copies of the Source form of the Standard Version of this Package in any medium without restriction [...] provided that you duplicate all of the original copyright notices and associated disclaimers."
20:09 masak "You are permitted to link Modified and Standard Versions with other works, to embed the Package in a larger work of your own [...] provided the result does not expose a direct interface to the Package."
20:11 masak sorear: since Niecza is also licensed with Artistic 2.0, I see very little problem.
20:13 masak of course, a little attribution is a kind gesture, if nothing else.
20:18 Mowah left #perl6
20:22 masak left #perl6
20:24 jnthn sorear: Seems like since you're under the same license, then the associated disclaimers are covered; that just leaves the copyright notices.
20:24 Mowah joined #perl6
20:24 zby_home left #perl6
20:25 jnthn sorear: I very much doubt there'd be any expectation to flag up with routines have which copyright, though.
20:25 jnthn sorear: Just some blanket "Portions copyright ..." or so would (IMHO, IANAL) do.
20:26 moritz_ "Part of this file is Copyright (C) 2010, 2011 by The Perl Foundation"
20:26 jnthn *nod*
20:26 jnthn Something like that.
20:28 masak joined #perl6
20:36 coldhead joined #perl6
20:48 sorear ok.
20:49 zby_home_ joined #perl6
20:51 tadzik left #perl6
20:52 tadzik joined #perl6
20:52 tadzik screen--
20:52 zby_home_ left #perl6
20:52 envi left #perl6
20:54 kjeldahl joined #perl6
21:23 sorear tadzik: try tmux
21:23 tadzik sorear: is it on feather?
21:24 sorear if it's not, I'll install it
21:24 cognominal left #perl6
21:24 tadzik it is, I'll check it out in some spare time
21:25 sorear I switched a few months ago (doy++)... it's much better at KISS
21:26 wooden left #perl6
21:29 cognominal joined #perl6
21:31 wooden joined #perl6
21:34 maja_ joined #perl6
21:40 Limbic_Region joined #perl6
21:40 maja_ left #perl6
21:43 kaare_ left #perl6
21:48 pmichaud jnthn: ping
21:50 pmichaud jnthn, moritz, others:   draft nqp plan at http://pmichaud.com/sandbox/nqp.txt
21:50 pmichaud comments and revisions welcomed
21:51 Limbic_Region left #perl6
21:52 jnthn pmichaud: Great! Will read now. :)
21:52 pmichaud I'll mention on #parrot shortly -- when we're happy with the text I'll post to parrot-dev and perl6-compiler
21:52 pmichaud and maybe also to my blog
21:53 jnthn Your blog lives now? :)
21:53 pmichaud no, but it's a reasonable first-post :)
21:53 pmichaud and I can make it live pretty much whenever I want
21:53 pmichaud (everything is in place, I'm just needing a good post)
21:53 pmichaud I expect I'll be tweaking the blog a fair bit over time
21:54 jnthn pmichaud: "multisub dispatch" in first para - just "multiple dispatch"
21:54 jnthn pmichaud: It handles multi-method dispatch too.
21:55 pmichaud right
21:55 pmichaud fixed
21:55 pmichaud (reload to see changes if/when you want)
22:03 jnthn The new nqp also contains a replacement implementation of multiple
22:03 jnthn dispatch -- it no longer makes use of the Parrot MultiSub PMC type.
22:04 jnthn That's true, but may be worth mentioning that it's to get consistent semantics with Perl 6 so we're a truer subset.
22:04 kst left #perl6
22:06 kst joined #perl6
22:09 Mowah left #perl6
22:10 PacoLinux left #perl6
22:15 jedai_ joined #perl6
22:19 jedai left #perl6
22:19 aindilis left #perl6
22:20 aindilis joined #perl6
22:25 felliott left #perl6
22:27 * allbery_b observes, apropos of several discussions elseirc, that while people grumble about p6's incompatibilities with p5, they're ignoring the fact that python and ruby are already causing havoc in that way
22:27 pmichaud yes, python has been incompatible with perl 5 for quite a while.  :-P :-P
22:27 colomon Yes, both python and ruby are quite in.... what pmichaud said.  :)
22:28 pmichaud I think we hope to be able to fix that in Perl 6 :-P
22:29 jnthn They're just DSLs in Perl 6, right? ;_)
22:29 colomon pmichaud: just built a new layout for my guy.  Two different elevated sections, the turntable, and Brendham Docks.  ;)
22:29 perigrin pmichaud: you have plans in perl6 to make python more compatible with perl5?
22:29 perigrin *that* I have to see.
22:29 colomon perigrin: we're quite ambitious.  ;)
22:34 sji left #perl6
22:37 tadzik hey, that was supposed to be Parrot, right? ;)
22:37 tadzik . o O ( dollar_underscore )
22:39 pmurias_ joined #perl6
22:40 pmurias joined #perl6
22:42 pmurias_ left #perl6
22:45 allbery_b heh
22:45 allbery_b no, ruby isn't compatible with itself between releases, and the problems with pythion 2.x vs python 3 are approaching the status of legend
22:45 allbery_b ...yet it's perl 6 that everyone seems to whine about
22:46 whiteknight joined #perl6
22:47 pmichaud perl 6 is often held to a different standard.  That's okay, we hold it to different standards also.  :-)
22:47 tadzik python has problems between 2.6 and 2.7 even. 3 was _supposed_ to be incompatibile
22:48 perigrin perl has (some) problems between 5.10 and 5.12 and 5.13
22:48 perigrin 6 is spposed to be incompatible :)
22:49 perigrin There are moments where I have a horrible suspicion that if someone were to release Perl6 as a brand new langauge without a name that matched qr/^Perl/;
22:49 perigrin people would love it
22:49 allbery_b I hear that from a lot of people, actually
22:50 allbery_b it isn't perl 6 that annoys them per se, they're annoyed that it's still called perl
22:50 perigrin I'm thinking less about the Perl community
22:50 allbery_b which is pretty much what I was ranting about there :)
22:50 perigrin and more about the people who hate on anything that matches qr/^Perl/
22:51 allbery_b I'm not talking about the perl community either, I'm talking random programmers, sysadmins, etc. who are annoyed that their little pigeonholes are being violated (possibly in several senses, in their opinions)
22:51 * perigrin nods
22:52 perigrin it's water under the bridge now
22:52 perigrin we have the brandings we have
22:53 perigrin also I don't have a single perl6 talk submitted to YAPC::NA yet
22:53 perigrin is everybody gonna wait till the deadline?
22:53 perigrin *sniff*
22:53 colomon when's the deadline?
22:53 colomon I've been pondering going.
22:53 perigrin the deadline is in March
22:53 colomon actually, when's the conference, for that matter?  :)
22:53 perigrin but we've been doing early acceptances
22:54 perigrin the conference is June 27-29th
22:54 pmichaud I've been considering skipping conferences this year
22:54 colomon the thing is, there's not a lot of point for me to talk if people like pmichaud are around.
22:54 perigrin I would consider that except I'm the freeking organizer
22:54 perigrin they expect me to attend.
22:54 perigrin colomon: um
22:54 shi joined #perl6
22:54 perigrin Perl6 has the same problems Perl5 has except larger
22:55 perigrin not enough people talking about the cool stuff that their language allows them to do
22:55 pmichaud colomon: it's *always* worth hearing from people other than the "principals"
22:55 perigrin exactly
22:55 pmichaud colomon: I think a talk describing what you've done with the various math components of Rakudo would be.... very well received
22:55 pmichaud and much more impressive than a talk from me on a topic like that
22:55 perigrin people *expect* pmichaud and jnthn and TimToady to be excited about Perl6 ... they tune it out after a while
22:56 perigrin having new people show new excitment is important
22:56 pmichaud I know that for the places I've been, it's the talks from people other than the principals that get the most buzz
22:57 colomon I've also never given any sort of talk like this before.  :)  Though I am tentatively scheduled to give a talk at Penguicon in late April.
22:57 perigrin colomon: honestly YAPC or a Perl workshop is the easiest place to give your first talk
22:57 pmichaud it's definitely worth making this a first talk then.  yapc audiences tend to be very friendly
22:58 perigrin we have a *far* lower bar than OSCON or some such :)
22:58 pmichaud exactly as perigrin++ says
22:58 perigrin I gave my first talk at a YAPC
22:58 perigrin and I sucked horribly
22:58 pmichaud imo, the friendly nature of yapc audiences makes for more interesting and useful talks
22:58 shortcircuit sorear: Sorry about that; was afk for a few hours.
22:58 perigrin we have a really good crop of talks submitted so far
22:59 perigrin but I've been pushing non-technical talks in the perl5 world
23:00 perigrin colomon: also if you submit *now*
23:00 perigrin you get a free slurpee
23:00 perigrin cause I promised free slurpees to the next 5 submissions
23:00 perigrin and I've gotten 3 so far
23:02 colomon My wife thinks I should give one too.  (Conversation here now.)  Though it may be she just wants to go to North Carolina.  :)
23:02 shortcircuit sorear: I don't recall where I found that Niecza was written in Perl 5, but I found it in what seemed a cannon source. I'll see if I can find it again; it sounds like that'll need to be corrected.
23:02 perigrin colomon: http://www.yapc2011.us/yn2011 # the picture doesn't lie :)
23:03 colomon "Oh, that's pretty!"
23:03 perigrin that's about 1hr north of Ashville looking back towards it
23:03 perigrin but it's representative of the mountains
23:04 colomon I don't suppose you can guarantee John Doyle will be in town?  ;)
23:04 perigrin uh ... I'd have to know who he is first I suspect
23:04 colomon http://www.johndoylemusic.com/
23:06 perigrin no sadly I can't.
23:06 colomon Didn't figure you could.  I'm pretty sure he lives in Ashville these days, when he's not out touring.
23:06 shortcircuit sorear: Ah. I think it was this bit: http://www.nntp.perl.org/group/perl​.perl6.announce/2010/11/msg641.html (followed by your noting https://github.com/sorear/niecza/​blob/master/docs/announce.initial )
23:07 pmurias shortcircuit: niecza wa rewritten in Perl 6
23:07 shortcircuit pmurias: So I gather.
23:09 perigrin colomon: I wouldn't be surprised. The area is *packed* with traditional and folk music.
23:09 colomon It's getting harder and harder to say no.... :)
23:09 perigrin which considering the scotch / irish settled the area isn't surprising
23:15 colomon Okay, I'm tentatively planning on going and giving a talk, but I owe Ambercon a slew of game submissions first, so I probably will have have to forego the slurpie.
23:15 hudnix left #perl6
23:15 shi left #perl6
23:17 hudnix joined #perl6
23:17 kst left #perl6
23:22 kst joined #perl6
23:30 hudnix left #perl6
23:32 MayDaniel joined #perl6
23:33 sorear pmurias: do you want to write anything about -Bclisp for the v2 mailing?
23:35 shortcircuit pmurias, sorear: Would either of you like to go in, edit and improve http://rosettacode.org/wiki/Niecza ?
23:35 shortcircuit Rosetta Code isn't Wikipedia; I _like_ having people involved who know what they're talking about.
23:36 sorear shortcircuit: you misread a line of mine.  I said the code has workarounds for bugs NOT in Rakudo
23:37 sorear i.e. the code will be simpler if Rakudo is used
23:37 shortcircuit sorear: Yeah, I saw your correction, and corrected the relevant page on RC.
23:39 sorear shortcircuit: I'm not really skilled at this "writing for an unknown audience" thing
23:39 icwiener joined #perl6
23:41 shortcircuit sorear: Understood. RC has a built-in audience of programmers, so a perspective as assumed for your original Niecza announcement is reasonable.
23:43 shortcircuit Hm. If you have any concerns about the content of the page, and aren't comfortable making the edit,  mention it here, in #rosettacode, or on the page's talk page. Something will likely come of it.
23:44 * shortcircuit needs to give thought to the perspective of facing the 'unknown audience'; that might be a major existing non-technical barrier to contributing to the wiki in general.
23:45 pmurias sorear: it's not really at the stage when it's ready to be annonuced
23:49 dju joined #perl6
23:57 pmurias left #perl6
23:58 * diakopter seems to arrive at #perl6 often slightly after pmurias' departure
23:59 dorlamm joined #perl6
23:59 masak lol; blogged. http://strangelyconsistent.​org/blog/the-thing-with-nim
23:59 MayDaniel left #perl6

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

Perl 6 | Reference Documentation | Rakudo