»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
00:00 dayangkun left 00:04 dayangkun joined 00:08 wamba left 00:13 wk__ joined 00:15 fgomez left 00:16 wk_ left
raiph irclog.perlgeek.de/out.pl?channel=p...;summary=1 #perl6 curated backlog for 2012-09-15 00:35
irclog.perlgeek.de/out.pl?channel=p...;summary=1 #perl6 curated backlog for 2012-09-16
Use the existing navigation links above, and the "summary 00:36
The above links are what you get if you click to get to the corresponding day, then click the "summary" link that appears next to Camelia (or wraps if your screen is too narrow) 00:39
00:41 fgomez joined 00:43 Chillance left
raiph phenny, tell moritz gist.github.com/3732447#gistcomment-574072 00:45
phenny raiph: I'll pass that on when moritz is around.
00:49 Circlepuller joined 00:52 Circlepuller__ left 01:03 anuby joined 01:05 dayangkun left 01:07 Circlepuller_ joined 01:10 Circlepuller left 01:11 hypolin joined 01:18 jaldhar left 01:19 fgomez left 01:25 cognominal left 01:28 thou left 01:31 cognominal joined 01:39 whiteknight left, cognominal left 01:40 cognominal joined 01:41 fgomez joined 01:46 jeffreykegler joined 01:50 FROGGS__ joined 01:53 flightrecorder left 01:54 flightrecorder joined 01:55 FROGGS_ left 01:58 dayangkun joined 02:04 jaldhar joined 02:05 orafu left, orafu joined 02:09 yarp joined, _jaldhar joined 02:10 jaldhar left 02:15 jeffreykegler left 02:18 _jaldhar left 02:19 _jaldhar joined 02:20 gootle joined 02:22 thou joined 02:32 pmurias left 02:33 thou left 02:36 thou joined
yarp r: @*INC.say; 02:39
p6eval rakudo 5aa57b: OUTPUT«/home/p6eval/.perl6/lib /home/p6eval/nom-inst/lib/parrot/4.7.0-devel/languages/perl6/lib␤»
03:38 ponpon left 03:44 fgomez left 03:53 fgomez joined
skids r: (0x1800000000000000 +& (+^0x1000000000000000)).fmt("%x").say 04:01
p6eval rakudo 5aa57b: OUTPUT«0␤»
skids hmmmm. that gives me f800000000000000 locally (star08). I wonder what's hosed. 04:02
TimToady r: (0x1800000000000000 - (+^0x1000000000000000)).fmt("%x").say
p6eval rakudo 5aa57b: OUTPUT«1␤»
sorear skids: the perl6 executable on p6eval is 32-bit 04:10
skids: I think rakudo uses machine integers for bitops
which is kind of odd because both feather1 and feather3 are x86_64 04:11
but we have:
sorear@feather3:~$ ls ~p6eval/nom-inst/bin/perl6 -lh
-rwxr-xr-x 1 p6eval p6eval 57K Sep 17 06:08 /home/p6eval/nom-inst/bin/perl6
sorear@feather3:~$ file ~p6eval/nom-inst/bin/perl6
/home/p6eval/nom-inst/bin/perl6: ELF 32-bit LSB executable, Intel 80386, version
TimToady Perl 6 shuld just do bitops in arbitrary precision by pretending it's infinite precision 2's complement 04:14
a negative number just has an infinitude of 1's instead of 0's 04:15
skids That does work to some extent. 04:16
I think there's some problems when things are near the 64-bit border and representations might be promoted/demoted. 04:17
TimToady of course, if you know you're gonna store the result in a native location, you can optimize
we want fast crypto too :) 04:18
skids r: (0x1800000000000000 +& (0xffffffffffffffff +^ 0x1000000000000000)).fmt("%x").say # workaround, works locally here.
p6eval rakudo 5aa57b: OUTPUT«0␤»
skids hrm.
r: (0x1800000000000000 +& (0xfffffffffffffffff +^ 0x1000000000000000)).fmt("%x").say 04:26
p6eval rakudo 5aa57b: OUTPUT«0␤»
04:34 fgomez left 04:42 fgomez joined 04:45 satyavvd joined 04:47 raiph_ joined 04:48 raiph_ left
moritz good morning 04:48
phenny moritz: 16 Sep 17:09Z <raiph> tell moritz gist.github.com/3732447
moritz: 16 Sep 18:20Z <not_gerd> tell moritz there appears to be some confusion on left<->right in the section on item assignment of doc.perl6.org/language/operators
moritz: 16 Sep 21:16Z <raiph> tell moritz What about logmarks? Click time field toggles a logmark. Logmark'd lines are highlighted. Does everything we need for phase 1. Very simple. What's not to love?
moritz: 00:45Z <raiph> tell moritz gist.github.com/3732447#gistcomment-574072
04:49 _jaldhar left 04:50 _jaldhar joined 04:54 pjcj joined
raiph blogs.perl.org/users/perl_6_reports...tware.html 05:06
moritz what's not to love? that clicking on the time already has another function that I'd like to preserve 05:07
05:07 wk__ left 05:09 adu joined
sorear moritz: could have a separate mode for raiph, no? 05:23
raiph moritz: ok.
sorear: that's basically what's there now. works great. 05:24
(but I don't let "works great" for certain use cases get in the way of "works better" for more use cases) 05:25
moritz: forgetting clicking on time, what do you think of the basic approach? click a line, it's logmarked. that's it. 05:26
moritz raiph: it'd like to hear more feedback on the summary mode from others first 05:27
raiph: if it's something that most people don't bother with, I don't want to integrate it that tightly with the rest of the page
if it's something most users pick up, I'm all for tighter integration 05:28
raiph moritz: "more feedback". makes sense. 05:31
05:38 yarp left
raiph moritz: While I hugely want this new summary mode to be successful, I don't want it to rely on me. 05:42
moritz: I'm already concerned because no one but me has done the summarizing in the day since you enabled the feature.
moritz: While that may of course change, it's not an auspicious start. 05:43
moritz: It doesn't take long to do a whole day, maybe an hour or two, but it's all relative.
I was really hoping for an approach that would lend itself to communal summarizing. 05:45
We shall of course see. I'm doing my best to encourage folk to help out. For example my latest blog: 05:47
blogs.perl.org/users/perl_6_reports...tware.html 05:48
Here's hoping.
*communal and continuous (up-to-the-minute) summarizing 05:51
05:57 skids left 05:59 fgomez left 06:00 fgomez joined 06:04 FROGGS__ left 06:05 adu left, flightrecorder left 06:09 tokuhiro_ joined 06:25 fgomez left 06:26 mdupont left 06:35 wtw joined 06:37 kaleem joined 06:38 raiph left 06:40 fgomez joined 06:42 mdupont joined 06:46 Cooqle joined 06:47 Cooqle left 06:49 fhelmberger joined 06:51 kaleem left 06:56 thou left, tokuhiro_ left, tokuhiro_ joined 07:01 tokuhiro_ left 07:04 thou joined 07:13 brrt joined 07:15 kaleem joined, flightrecorder joined 07:16 FROGGS joined
FROGGS good morning 07:21
sorear o/ 07:25
07:27 marmay joined 07:33 thou left 07:34 wk_ joined 07:45 kresike joined
kresike good morning all you happy perl6 people 07:46
sorear o/
mathw Good morning 07:49
anuby 0/ 08:00
brrt \o kresike 08:05
08:20 Psyche^ joined 08:23 Patterner left, Psyche^ is now known as Patterner
dalek rlito: 408fa18 | (Flavio S. Glock)++ | Makefile:
Perlito5 - add "make minify"
 08:24
08:24 dakkar joined 08:36 tokuhiro_ joined 08:44 yarp joined
yarp nr: my Num @number=<3 4 5 6>; @number[4] = [+] @number; @number.say; 08:44
p6eval niecza v21-11-g09989b4: OUTPUT«3 4 5 6 18␤»
..rakudo 5aa57b: OUTPUT«Type check failed in assignment to '$v'; expected 'Num' but got 'Int'␤ in block at /tmp/ByQwh5XleO:1␤␤»
08:50 Chillance joined, Chillance left 08:51 Chillance joined 08:52 Chillance left 08:53 Chillance joined, Chillance left, return joined
return r: '1a2b3c4d5'.subst(/(\d+)/,{$0~$0},:g) 08:54
p6eval rakudo 5aa57b: OUTPUT«use of uninitialized value of type Any in string context in block at /tmp/SpZ3bZi89C:1␤␤use of uninitialized value of type Any in string context in block at /tmp/SpZ3bZi89C:1␤␤use of uninitialized value of type Any in string context in block at /tmp/SpZ3bZi89…
return r: '1a2b3c4d5'.subst(/(\d+)/,{$1~$1},:g) 08:55
p6eval rakudo 5aa57b: OUTPUT«use of uninitialized value of type Any in string context in block at /tmp/C9vHJNhAY9:1␤␤use of uninitialized value of type Any in string context in block at /tmp/C9vHJNhAY9:1␤␤use of uninitialized value of type Any in string context in block at /tmp/C9vHJNhAY…
return n: '1a2b3c4d5'.subst(/(\d+)/,{$0~$0},:g) 08:56
p6eval niecza v21-11-g09989b4: ( no output )
return hello \o/ perl6, why the code wrong? 08:57
08:57 kaleem left 08:59 cognominal left
return n: say '1a2b3c4d5'.subst(/(\d+)/,{$0~$0},:g) 09:00
p6eval niecza v21-11-g09989b4: OUTPUT«11a22b33c44d55␤»
return |o/ n++ 09:01
jnthn morning o/
return morning jnth \o\ 09:02
n: say '1a2b3c4d5'.subst(/(\d+)/,{sprintf("%02d",$0)},:g)
p6eval niecza v21-11-g09989b4: OUTPUT«01a02b03c04d05␤»
return r: say '1a2b3c4d5'.subst(/(\d+)/,{sprintf("%02d",$0)},:g) 09:03
p6eval rakudo 5aa57b: OUTPUT«No such method 'Int' for invocant of type 'Any'␤ in any at src/gen/BOOTSTRAP.pm:91␤ in sub sprintf at src/gen/CORE.setting:2131␤ in block at /tmp/SyyL1YY6LX:1␤␤»
jnthn r: say '1a2b3c4d5'.subst(/(\d+)/,-> $/ {sprintf("%02d",$0)},:g) 09:04
p6eval rakudo 5aa57b: OUTPUT«01a02b03c04d05␤»
return oh r++ 09:05
FROGGS hi jnthn 09:08
return jnthn: thanks. 09:09
09:09 mucker joined 09:10 return left 09:22 kaleem joined 09:30 mucker left 09:41 Chillance joined 09:42 Chillance left 09:53 tokuhiro_ left 09:54 tokuhiro_ joined 09:59 tokuhiro_ left 10:04 dorlamm joined 10:05 cognominal joined, hypolin left 10:06 cognominal left 10:09 gootle left 10:11 gootle joined 10:16 anuby left 10:20 wamba joined 10:32 dayangkun left 10:45 cognominal joined 10:51 cognominal left 10:52 daxim joined 10:53 dayangkun joined 10:55 Targen left 10:56 tokuhiro_ joined 11:00 tokuhiro_ left, tokuhiro_ joined 11:01 cognominal joined, tokuhiro_ left 11:02 tokuhir__ joined 11:05 dorlamm left
jnthn std: -foo => 'bar' 11:08
11:08 Su-Shee left
p6eval std b842bb3: OUTPUT«ok 00:00 41m␤» 11:08
jnthn std: +foo => 'bar' 11:09
p6eval std b842bb3: OUTPUT«ok 00:00 41m␤»
11:10 Su-Shee joined, gootle left
jnthn In glitchmr.github.com/perl-6-changes-2012W37.html it mentions that -foo => 'bar' pair form is now supported, but the commit was to STD_P5, not STD: github.com/perl6/std/commit/942bb1...41e9956a74 11:11
Otherwise, nice summary :)
11:13 am0c joined 11:14 Chillance joined, Chillance left 11:19 am0c left 11:24 pmurias joined 11:27 _jaldhar left 11:31 am0c joined 11:32 Chillance joined, Chillance left 11:34 Chillance joined, Chillance left 11:35 Chillance joined, Chillance left
pmurias jnthn: does nqp -e 'use nqp' work for your? 11:37
jnthn pmurias: No; I fixed the compreg approach though 11:39
(The one you were originally trying)
pmurias thanks, worked 11:40
sirrobert what mechanism is there for class-level variables? 11:51
jnthn sirrobert: You, er, declare a variable inside of the class. :-)
hoelzro sirrobert: I think you can just do this:
sirrobert via closure?
hoelzro class MyClass { my $foo; } 11:52
jnthn Right. :)
sirrobert then manually do an accessor to get it from outside?
jnthn my $.foo; # generates an accessor for you
Su-Shee that's what $.foo is for..
sirrobert ohhhhhhhhhhhhhhhhhhhhh
not *has* $.foo
*my* $.foo
jnthn Right, "has" means "one per instance" 11:53
sirrobert yeah
ok, that's great
trying it out now =)
FROGGS ohh, good to know
11:53 dayangkun left
sirrobert that's great. 11:53
I notice that "my $.bar" doesn't need "is rw" the same way "has $.bar" does. 11:54
Su-Shee I thought rw is default anyways?
sirrobert not for "my $.foo" (at least not in my version of rakudo =)
moritz Su-Shee: no, default is ro 11:55
jnthn defualt is ro. Hm. :)
I think that wants to be the case for the "my" one too but...we don't really have variable traits yet. Unless moritz++ added 'em...
11:56 cognominal left
Su-Shee ah well, I rarely use instance/class vars anyways.... :) d 11:56
moritz jnthn: no, I haven't 11:57
jnthn oh, you did constants...
moritz the problem is really that traits work at compile time, but at compile time we only see the entry in the proto lexpad
so there's not much we can do in terms of traits that affect the value of a variable
jnthn Well, traits are more about the container I geuss 11:58
*guess
11:58 am0c left 11:59 JimmyZ joined, satyavvd left 12:03 crab2313 joined 12:04 cognominal joined 12:12 cognominal left 12:19 cognominal joined 12:22 mucker joined 12:26 cognominal left 12:27 atrodo joined
[Coke] sees "since the day you enabled the feature" in backlog, and wonders what feature, and if you want it to be more noticed, mentioning someplace other than backscroll is a good idea. 12:28
esp. since feather was rebooted in the past week. 12:29
12:34 GlitchMr joined 12:36 cognominal joined 12:38 mucker left
FROGGS is there a way to open pipes? 12:38
[Coke] jnthn: no worries on the autoviv thing for now; I already worked around that in partcl-nqp there weren't many cases of it.
FROGGS my $fh = open( "| cc -xc -", :w ) doesnt work for me :/
moritz FROGGS: I think there's an unspecced adverb to open() that does that
[Coke] jnthn, pmichaud: latest stick on partcl-nqp update is issue #53 on nqp, but also: converting PAST to QAST is not straightforward. 12:39
12:39 adu joined
moritz FROGGS: search for pipe in src/core/IO.pm in rakudo 12:39
FROGGS thanks
jnthn [Coke]: What are you finding non-straightforward, ooc? 12:42
[Coke] jnthn: many of the :adverbs are missing.
lack of docs on both versions makes it hard to say "oh, this is now spelled that way"
jnthn Lack of docs is a problem, yes.
[Coke] Doesn't help that pmichaud setup most of the original PAST, so I had it handed to me the first time. 12:43
jnthn Ah
[Coke] jnthn: for example, here's one that i just ripped out but probably shouldn't have:
FROGGS moritz: this is working: my $fh = open( "cc -xc -", :w, :p )
[Coke] QAST::Var.new( :name<lexpad>, :scope<register>, :isdecl ... 12:44
jnthn :scope<local> :decl<var>
er, with the comma
Problem is that PAST treated parameter as a scope. 12:45
Which meant you couldn't have parameters that went into registers.
You were forced to do lexicals
Which ain't always desirable.
[Coke] jnthn: thank you. 12:46
what about :viviself(...) ?
12:46 wamba left
jnthn That one really did get killed. It also did a couple of distinct things. If you were using it on an isdecl, just do a bind to the thing you're declaring. If you really did want to do some kind of "if it's null" logic per lookup (discouraged unless you must do it), see VarWithFallback 12:48
[Coke] how to do the bind? :bind<thing> ? 12:49
12:49 yarp_ joined
jnthn QAST::Op.new( :op('bind'), ... ) 12:49
First arg is the QAST::Var, second is the thing to bind
[Coke] Thanks. 12:50
what about :hll ? 12:51
12:51 yarp left, yarp_ is now known as yarp
jnthn There's now a QAST::CompUnit that wraps the entire thing to be compiled, and the :hll goes on that. 12:51
QAST::CompUnit.new( :hll('tcl'), $top-level-block ) 12:52
12:57 wamba joined, Chillance joined, Chillance left
Woodi raiph: just learned today about summarizer and reports and like both :) 12:58
but me thinks joins can be filtered out in summary view or maybe displayed in smallest rows ? 13:00
13:01 mucker joined
Woodi and maybe "navigation links" can use bigger font ? 13:01
13:01 PacoAir joined
moritz why would the summary contain joins? 13:09
just don't add them to the summary :-)
13:18 crab2313 left
sirrobert r: class A { my $.foo; has $.foo is rw; }; A.foo = "bar"; my $a = A.new; $a.foo = "baz"; say $a.foo; say A.foo; 13:25
p6eval rakudo 5aa57b: OUTPUT«baz␤baz␤»
sirrobert Is there a way to have a class version separate from an instance version ?
13:26 PerlJam joined
jnthn By calling them different things? :) 13:27
sirrobert I'm hoping for something like setting something at the class level (at runtime) as a default for instances. In particular, db config stuff 13:28
jnthn That's fine, just have a my $.foo-default; has $.foo = $foo-default; 13:29
or something
sirrobert fair enough
moritz (the accessors are just regular methods, which is why you can't have accessors for two attributes of the same name) 13:31
13:31 orafu left
sirrobert ok, that makes sense =) 13:31
13:34 tokuhir__ left, tokuhiro_ joined 13:38 yarp_ joined 13:39 tokuhiro_ left 13:40 yarp left, yarp_ is now known as yarp 13:42 sisar joined 13:50 skids joined 13:56 stopbit joined
pmurias jnthn: what does --vmlibs=perl6_group,perl6_ops do? 13:57
jnthn pmurias: Ensures that .loadlib directives for the Perl 6 PMCs and dynops are emitted. 13:59
pmurias does it stop the nqp via compreg from working?
13:59 flightrecorder left
jnthn Don't see why it'd do that; they feel quite orthogonal. 14:00
pmurias jnthn: I get pastie.org/4738542
Makefile: paste.ubuntu.com/1211092/ 14:01
jnthn bah, I get an Application error when looking at that paste...
14:01 FROGGS left
pmurias jnthn: which one? 14:01
14:02 flightrecorder joined
jnthn the first one 14:03
yarp r: my Num @number = <1 2 3 4>; @number[4] = [+] @number; 14:05
p6eval rakudo 5aa57b: OUTPUT«Type check failed in assignment to '$v'; expected 'Num' but got 'Int'␤ in block at /tmp/uqd_5X2XmR:1␤␤»
yarp nr: my Num @number = <1 2 3 4>; @number[4] = [+] @number;
p6eval rakudo 5aa57b: OUTPUT«Type check failed in assignment to '$v'; expected 'Num' but got 'Int'␤ in block at /tmp/00B337Ld2O:1␤␤»
..niecza v21-11-g09989b4: ( no output )
14:05 kaare__ joined
yarp nr: my Num @number = <1 2 3 4>; @number[4] = [+] @number; @number.say; 14:06
p6eval niecza v21-11-g09989b4: OUTPUT«1 2 3 4 10␤»
..rakudo 5aa57b: OUTPUT«Type check failed in assignment to '$v'; expected 'Num' but got 'Int'␤ in block at /tmp/Vr6pUabaJ7:1␤␤»
14:07 benabik left
cognominal jnthn, would be possible to derive a rakudo grammar from HLL::Grammar ? Currently, it is only possible from nqp 14:07
jnthn No. 14:08
cognominal one cannot use at once nqp and rakudo pbc?
jnthn Well, maybe you can load it, but then the bigger problem is that HLL::Gramamr ultimately derives from NQPCursor and thus NQPMu, and so you'd get NQP match objects, not Perl 6 ones. 14:09
cognominal ha, thx
jnthn We could either: 14:10
1) factor this with roles instead
2) make a Perl 6 module with some of the HLL stuff in and put it on modules.perl6.or
*org
I suggest doing it as roles though 14:11
A QuoteParsing role, a ExpressionParsing one, etc. Could be quite nice.
pmichaud++ may have other suggestions
moritz btw the name conflict between methods in Any and grammar rules really bothers me 14:12
r: grammar A { token TOP { <tree> }; token tree { a } }; say A.parse('a') 14:13
p6eval rakudo 5aa57b: OUTPUT«Cannot call 'tree'; none of these signatures match:␤:(Any:U : Mu *%_)␤:(Any:D : Mu *%_)␤:(Any:D : &c, Mu *%_)␤:(Any:D : Cool $count, Mu *%_)␤␤ in method tree at src/gen/CORE.setting:1224␤ in any !reduce at src/stage2/QRegex.nqp:582␤ in any !cursor_pass at src/st…
14:13 FROGGS joined
moritz and I wonder why. Isn't A much more specific than Any? 14:13
r: grammar A { token TOP { <tree> }; token tree { a } }; say A.^mro
p6eval rakudo 5aa57b: OUTPUT«A() Grammar() Cursor() Any() Mu()␤»
jnthn moritz: The problem is that we ended up with Any in $*ACTIONS 14:17
cognominal++ already pointed out this issue once before
So it calls Any.tree as an action method 14:18
14:19 tomaw_ joined
moritz I wonder if we could set $*ACTIONS to a singleton that either doesn't have any methods, or on which we don't call methods 14:19
pmurias jnthn: error re paste: pastie.org/4738666
moritz hm, or maybe a catch-all method that doesn't do any harm 14:20
doesn't sound very convincing to me though :(
14:22 kaleem left
jnthn pmurias: That appears when you add the --vmlibs line? 14:23
14:24 tomaw_ is now known as tomaw, doy joined
jnthn Can't reproduce that here 14:24
nqp -e "pir::compreg__Ps('nqp').compile('say(42)')()" 14:25
42
nqp --vmlibs=perl6_group -e "pir::compreg__Ps('nqp').compile('say(42)')()"
42
pmurias jnthn: I got confused 14:26
moritz I could imagine that another process might be having trouble to compreg('nqp') without loading those libs first, no?
pmurias yes, that is likely what's happening
jnthn moritz: But if this *is* NQP... :) 14:27
moritz but then I don't think that NQP doesn't emit the necessary compreg__SP call to allow that in first place
jnthn moritz: but yeah, there's an issue here in that compreg is so global it oversteps the global boundary...
moritz: It does :)
moritz it does?
where? 14:28
jnthn $nqpcomp.language('nqp');
.language calls compreg
moritz but that only happens if you call that method, no? 14:29
I mean, if I load the compiled NQP .pir file, it's not called, is it?
masak oh hai, #perl6
moritz so I can use compreg from within an nqp process
14:30 Celelibi left
moritz but I can't use it load the compiler from another process that loads NQP.pbc 14:30
oh hai masak
FROGGS hi masak
14:31 benabik joined 14:32 Celelibi joined
masak been teaching git today :) 14:33
jnthn moritz: ah, I see what you mean. Yeah, that's an issue 14:34
14:34 spider-mario joined
sirrobert r: class A { has $.foo; submethod BUILD { say "FOO: $!foo"; } }; say A.new(foo => 'bar'); 14:37
p6eval rakudo 5aa57b: OUTPUT«use of uninitialized value of type Any in string context in submethod BUILD at /tmp/IdJh1JxoqR:1␤␤FOO: ␤A.new(foo => Any)␤»
tadzik ohhai
sirrobert some day I'll grok build... =) How can I access $!foo with value "bar" in this build?
Su-Shee masak: I wish git would make its interface as nice and clean as darcs' made theirs.
jnthn sirrobert: BUILD runs *before* any defaults are applied
sirrobert: You need to take :$foo as an argument in the BUILD 14:38
sirrobert ahhhh ok, thanks
I did something like that but wrong syntax
jnthn ah :)
masak Su-Shee: git's interface is its weak spot, yes.
sirrobert is there anything that runs after defaults are applied? 14:39
moritz I don't think so 14:40
sirrobert ok, thanks
moritz note that if you write
submethod BUILD(:$!foo) { }
then it will fill in the attribute at the time that BUILD is called
(ie before the body of BUILD is run)
see also: doc.perl6.org/language/objects#Obje...nstruction 14:41
14:43 cognominal left
[Coke] jnthn: thanks for the pointers. I will fixup all that after $DAYJOB and see what's left. 14:44
jnthn [Coke]: you're welcome :)
14:48 cognominal joined
kresike bye folks 14:49
14:49 kresike left 14:50 kst left, yarp left 14:51 yarp joined 14:59 thou joined, fasfdsafadsfas joined 15:01 adu left, adu joined 15:02 yarp left 15:03 adu left 15:04 adu joined
felher Is the callsame/nextsame/... stuff for methods borrowed from lisp? 15:05
moritz has no idea
jnthn would have to reach for AMOP to remember how similar their factoring of it is 15:06
15:08 stopbit left
felher Hm, AMOP sounds interesting :) 15:08
moritz would love some more (especially non-lispy) literature on MOPs
jnthn Yeah, somebody should totally write some... :) 15:09
jnthn volunteers the guy who did 6mo...oh, wait...darn :)
felher :P
15:10 stopbit joined
jnthn felher: AMOP is a nice read, if you can pretend to know LISP while you read it :) 15:11
masak 6mo is so metacircular that even the guy who wrote it ends up pointing to himself :P
jnthn I managed to pretend well enough to follow what was going on :)
masak also, this line of jokes is so metacircular it ends up being mostly about itself, too. :P
masak gets sucked into a vortex of self-reference! 15:12
tadzik you put yourself in it yourself!
er, I mean,
jnthn That must drive one loopy...
felher adds AMOP to his to-read-queue 15:13
sirrobert so, I have a bit of code from Config::INI. There's a method that gets run twice. The first time, Config::INI::parse_file(...) works fine. The second time it complains "Could not find symbol 'Config::INI::&parse_file'. What would cause that?
err a bit of code that uses Config::INI
moritz missing 'use' statement?
or the precompilation bug?
sirrobert no, it works the first time (so it's being used and parse_file works)
I'm not precompiling 15:14
tadzik uh-oh, that'
s my module :P
sirrobert let me see if I can put together a gist that reproduces it ...
moritz that would be helpful, yes
sirrobert tadzik: it doesn't "feel" like a problem with Config::INI ... =)
tadzik good :)
tadzik decomutees
15:15 kaare__ left
masak tadzik: the best moment of this summer was when I realized that I could put the helmet in the helmet, and it would just *disappear* from the room, because it was now hidden in itself. 15:16
moritz woah, carving out a piece of reality into a circulary graph!
FROGGS masak: I've got the key from "the lost room" at home, I can put these things there if I dont need them anymore ;o) 15:17
15:17 kaare_ joined
FROGGS but its not that cool anyway 15:18
moritz wow, I've got an email with subject 'Perl6 Help', and before opening it I thought "if you want help, go to the mailing lists or IRC". Then I opened it, and found that it was from somebody who actually wants to help us :-) 15:19
15:19 ivan``_ joined
FROGGS heh 15:19
BANG! surprise! 15:20
sirrobert =)
FROGGS and? where is he or she? 15:21
PerlJam moritz: "Hi, I'm an independantly wealthy person who happens to love the direction Perl 6 is going; how do I contribute some $$$ to the Perl 6 effort?"
</dream>
moritz PerlJam: nah, "just" offering some proof-reading / copy editing of documentation for now 15:22
skids I'd settle for "I'm an independently wealthy person"
15:22 sqirrel joined
PerlJam heh, I was just about to write: more likely, "I can write, Perl 6 needs docs, how can I help?" 15:22
15:26 fasfdsafadsfas left
masak "I took the liberty of finishing Perl 6 last Thursday between laundry and dinner. Merry Christmas!" 15:27
daxim I'm *so* going to quote you out of context
GlitchMr gist.github.com/3737998 15:28
wow, I've made "serious" PHP code
masak daxim: if you quote me, remember to use two layers of quotes. 15:29
daxim »_«
did you design this, GlitchMr? 15:31
GlitchMr perl6: print ｢\ // welcome \\ to // the \\ world /｣
p6eval rakudo 5aa57b: OUTPUT«===SORRY!===␤Confused␤at /tmp/YegGFWLqdf:1␤»
..niecza v21-11-g09989b4: OUTPUT«===SORRY!===␤␤Unsupported use of bare 'print'; in Perl 6 please use .print if you meant $_, or use an explicit invocant or argument at /tmp/I2aaIhloJI line 1:␤------> print⏏ ｢\ // welcome \\ to // the \\ world /｣␤␤Confu…
GlitchMr daxim: yes
PerlJam GlitchMr: next, get rid of the named function requirement :)
GlitchMr std: print ｢\ // welcome \\ to // the \\ world /｣
p6eval std b842bb3: OUTPUT«ok 00:00 41m␤»
daxim I'd spell whatever - any
brrt GlitchMr: yay kill php! 15:32
GlitchMr Actually, 'off-rails.php' code itself is better
15:32 flightrecorder left
GlitchMr That was just example 15:32
gist.github.com/3738030
15:33 FROGGS left
PerlJam GlitchMr: you've got to have the function that handles the dispatch called "crazy_train" 15:34
GlitchMr Hmmm... I've feeling I could make this more awesome 15:37
15:42 rhr left 15:44 M_o_C joined 15:45 Circlepuller joined, Circlepuller_ left
sirrobert ha 15:59
GlitchMr gist.github.com/3738030
sirrobert I isolated the problem... making a gist now 16:00
GlitchMr Now it looks like average PHP code, doesn't it?
sirrobert oohhhh... oneupsgistship, eh?
GlitchMr: Yep, looks phpish =)
GlitchMr I even put "die or die" because normal "exit" is boring :-) 16:01
sirrobert What's interesting about the file?
16:03 wk_ left
GlitchMr Now, I should push this to GitHub 16:03
16:04 brrt left
sirrobert ok! gist.github.com/3738213 shows the (weird) error I'm getting 16:06
I think it's a bug in 'require'
jnthn What's in Loader? 16:07
But yeah, weird.
PerlJam
.oO( something that stomps on Config::INI::parse_file )
 16:08
sirrobert nothing is in Loader 16:09
16:09 JimmyZ left
sirrobert or you can put junk in it... it doesn't really matter 16:09
the point is, "requiring" any file causes Config::INI::parse_file to break
16:10 am0c joined
jnthn oh, hmmmm 16:11
GlitchMr gist.github.com/3738233
I hope this README is realistic
daxim it's terrible 16:16
seriously, this needs so much improvement I don't know where to start the critique
GlitchMr that's the point 16:17
Also, github.com/GlitchMr/php-off-the-rails
tadzik I don't quite see why are you doing this :) 16:18
GlitchMr To get PHP badge on Coderwall... perhaps? 16:19
daxim no, I'm saying your English writing sucks donkey balls and this reflects poorly on your code crafting
do you want help?
tadzik I'd be ashamed of such thing
GlitchMr But it works!
I mean, it works with simple "Hello, world!" programs 16:20
daxim answer me
GlitchMr No 16:21
I know that my English sucks
jnthn sirrobert: golfed it
sirrobert looking (thanks)
jnthn sirrobert: And think I found the fail 16:22
GlitchMr If you would like to see anything more serious I've made, I would look into my JavaScript, Perl 6 or AWK code 16:23
(too bad that GitHub doesn't recognize AWK) 16:24
timotimo GlitchMr: i don't know php any more. what does your thingie do? :\
GlitchMr timotimo: It works like Dancer, except it's worse
timotimo so it's a small web framework? 16:25
never tried dancer before 16:26
GlitchMr Well, yes
I can do stuff like
16:26 benabik left
GlitchMr get('/', function () { echo 'Hello, world!' }) 16:26
16:26 rhr joined
GlitchMr And it shows "Hello, world!" on main page of site 16:26
16:27 benabik joined
sirrobert jnthn: back (impromptu meeting). What's the skinny? 16:28
timotimo mhm, that reminds me of the microframeworks they have in the python world
jnthn sirrobert: Seems some mess in handling of GLOBAL
sirrobert hm ok. Can I help? =) 16:29
jnthn sirrobert: I think I can see what needs fixing.
sirrobert jnthn: cool, thanks-- I have a workaround in place in the meantime... but it's ugly =)
GlitchMr glitchmr.github.com/php-off-the-rails/ 16:30
:-)
jnthn yeah, I can patch the issue this evening, I think
sirrobert (not -- in the decrement sense =)
jnthn sirrobert: Got a fix locally. spectesting 16:33
sirrobert cool, thanks
jnthn++ 16:34
jnthn One of those "how did we not spot this before" things...
16:34 flightrecorder joined, FROGGS joined
FROGGS hi there 16:35
jnthn hi FROGGS 16:36
dalek kudo/nom: ec413ef | jnthn++ | src/Perl6/ModuleLoader.pm:
Fix busted GLOBAL preservation code.

Fixes a bug whereby require would clobber GLOBAL, discovered by
  sirrobert++.		 16:38
ast: efbc8d7 | jnthn++ | S11-modules/require.t:
Test for require / GLOBAL.
16:44 ajoe joined 16:45 ajoe left
daxim rt.perl.org/rt3/Ticket/Display.html?id=114946 16:45
PerlJam jnthn: What's the difference between get_hll_global and get_root_global ?
16:46 ajoe joined, MayDaniel joined, kaare__ joined
jnthn PerlJam: Where it looks 16:46
PerlJam jnthn: I ask because there's also a get_hll_global on line 273 of ModuleLoader.pm that I wonder if it also shouldn't be changed.
jnthn PerlJam: We were preserving GLOBAL in NQP land
16:47 kaare_ left, ajoe is now known as docdoctor
jnthn PerlJam: er...yes 16:49
jnthn updates it and spectests again 16:50
daxim: On that ticket, I'm not sure the point is optimizability so much as it is stopping people changing the arguments and accidentally updating things in the caller, unless that's specifically declared as being the intent. 16:51
daxim the topic is currently under discussion on p5p, you should add your comment to the ticket 16:53
moritz nr: sub f($x) { $x := 5; say $x }; f(42) 16:54
p6eval niecza v21-11-g09989b4: OUTPUT«5␤»
..rakudo 5aa57b: OUTPUT«===SORRY!===␤Cannot use bind operator with this left-hand side␤at /tmp/RzWynDDFS4:1␤»
16:55 dakkar left 16:59 wk_ joined 17:00 kaare_ joined, kaare__ left 17:04 leont joined 17:08 kaare__ joined, daxim left 17:09 kaare_ left 17:14 xinming_ joined 17:17 xinming left
jnthn dalek: done so 17:17
er, daxim^^ 17:18
dinner &
17:23 REPLeffect joined 17:25 thou left
FROGGS are here any perl6 windows ppl? 17:30
is there something I need to care about when installing rakudo?
17:36 tgt joined
[Coke] #perl6 seems to like reddit, neh? 17:38
17:44 thou joined 17:46 fhelmberger left
moritz it's a love/hate relationship, I think :-) 17:48
dalek href="https://glitchmr.github.com:">glitchmr.github.com: 0cd0e94 | GlitchMr++ | _posts/2012-09-16-perl-6-changes-2012W37.md:
The message is actually more detailed - instead of "value" it shows variable name.
 17:53
href="https://glitchmr.github.com:">glitchmr.github.com: 6973c33 | GlitchMr++ | _includes/footer.html:
Add my realname to footer
 17:54
GlitchMr glitchmr.github.com/perl-6-changes-2012W37.html 17:58
I have feeling that Disqus is broken 17:59
It shows "2 comments" but I only see one
Second one is my comment
Perhaps they have delays or something 18:02
18:12 daniel-s__ joined 18:13 kaare__ is now known as kaare_ 18:14 daniel-s_ left, adu left 18:15 adu joined 18:17 kikulli joined
sisar can someone point out what the problem is: gist.github.com/3738874 ? I'm unable to add a path to $PATH using export in Cygwin. 18:18
PerlJam sisar: bogus (invisible) characters at the start of your command? 18:19
GlitchMr eval: print "\302\226" 18:20
buubot_backup GlitchMr: Â1
GlitchMr eval: "\302\226"
buubot_backup GlitchMr: Â
GlitchMr that... makes sense
sisar PerlJam: yes. I had just copy-pasted that from GlitchMr++'s page (glitchmr.github.com/how-i-learned-t...nda.html). Typing by hand solved it. Thanks. 18:21
18:22 leont left
GlitchMr Wait... invisible characters on my page? 18:23
That doesn't sound right
PerlJam GlitchMr: It probably isn't on your page. I just tried to reproduce by cut-n-paste and couldn't 18:24
GlitchMr What is your browser?
PerlJam I'm on chrome
sisar GlitchMr: Firefox, 15.0
PerlJam: i selected that line by triple-click. 18:25
*triple-left-click
GlitchMr Weird, it works for me
Also Firefox 15.0, triple left click 18:26
(well, I use Opera as my main browser, but other browsers are useful for testing stuff)
sisar GlitchMr: maybe Cygwin has some problme ?
GlitchMr Well, invisible characters wouldn't appear from nowhere
PerlJam sisar: maybe you had some invisible stuff on the line prior to pasting? Can you reproduce?
GlitchMr But I'm on Linux, so perhaps it's something specific to Windows version of Firefox 18:27
sisar PerlJam: yes, i can reproduce consistently.
PerlJam could be.
sirrobert exit
woops =)
18:27 sirrobert left
GlitchMr glitchmr@strawberry ~/g/g/_posts> file 2012-09-15-how-i-learned-to-stop-worrying-and-install-the-panda.md 18:29
2012-09-15-how-i-learned-to-stop-worrying-and-install-the-panda.md: ASCII text
Interesting...
It could be postprocessing, except when using hex editor I don't see anything strange near `export` command. 18:30
sisar GlitchMr: after I have exported, how do make the changes to $PATH permanent ? Just opening and closing the Cygwin terminal does not help. 18:33
GlitchMr Modify .bashrc?
I guess 18:34
It should be there
Or... oh, you're on Windows
Enter into Control Panel and modify system env variables
moritz cygwin probably reads a .bashrc or _bashrc
sisar GlitchMr: Cygwin has its own .bashrc. I'll try modifying that.
GlitchMr sisar: or you may try support.microsoft.com/kb/310519 18:35
jnthn back 18:36
GlitchMr glitchmr@strawberry ~/g/g/_site> perl -C -E 'while ($_ = getc) { say ord if $_ gt "\x7E" }' < how-i-learned-to-stop-worrying-and-install-the-panda.html 18:37
960
dalek kudo/nom: 43ceae0 | jnthn++ | src/Perl6/ModuleLoader.pm:
Also fix setting loading; PerlJam++.
GlitchMr That would be two pi symbols
sisar has to go to all this trouble to run perl6 on windows because Ubuntu just won't dual boot on my new Ultrabook :/
jnthn sisar: Did you use the MSI, or are you doing source builds? 18:38
GlitchMr Except... that wouldn't make sense because it also should find "☺"
sisar jnthn: source builds
jnthn OK. Was just asking in case the reaction was, "oh, there's an MSI?" :) 18:39
18:39 stopbit left 18:40 stopbit joined
GlitchMr glitchmr@strawberry ~/g/g/_site> perl -CSDL -0nE 'for (split //) {say ord if $_ gt "\x7E"}' how-i-learned-to-stop-worrying-and-install-the-panda.html 18:40
960
9786
... no invisible characters
(except for spaces and linebreaks)
sisar GlitchMr: its ok. Don't fret over it :-) 18:41
GlitchMr :-)
I've no idea why it happens, but it's not my site 18:42
(I like having static HTML pages)
sisar jnthn: lets say I install Rakudo using .msi. But then i will have to use the Windows Command prompt right ? Which sucks, btw. Or is there a better command shell for Windows ?
GlitchMr sisar: yes, you can download git 18:43
It includes bash
This thing (msysgit.github.com/) includes bash
jnthn sisar: You can use anything that will invoke the perl6.exe :)
sisar jnthn: what do u use ?
cmd.exe ?
GlitchMr I mean, "msysGit", not so called "Git for Windows"
jnthn sisar: yes 18:44
sisar: I don't really recommend it :P
sisar :-)
GlitchMr I have used bash.exe
When I was using Windows
jnthn I can run Rakudo just fine from the bash included with msysGit, anyways. 18:45
GlitchMr dl.dropbox.com/u/63913412/bytheway.png
jnthn
.oO( maybe I should switch to using that more often...if only my fingers can unlearn what they're used to... )
GlitchMr I've found this screenshot
It's rather old, but it shows bash from msysgit 18:46
(just a word of warning - msysgit includes perl.exe - I would remove it and use some real Perl (such as Strawberry Perl))
(well, ok, that Perl is real too, but it's old - 5.8)
18:47 sirrobert joined
GlitchMr o/ 18:47
pmichaud good afternoon, #perl6 18:49
tadzik god evening #perl6
jnthn afternoon, pmichaud :) 18:50
18:50 adu left 18:51 kikulli left
pmichaud does test builds of rakudo+nqp on latest parrot 18:51
GlitchMr glitchmr@strawberry:~$ `echo echo 42` 18:52
42
jnthn ooh, I should re-test the latest Parrot too...
GlitchMr Who needs eval when you have backticks ;-)
pmichaud I'm testing both linux and win32 (strawberry/gmake)
jnthn win32/msvc here
pmichaud we could use a msv.... right :-)
also a 64-bit test.
(my windows is 32 bit) 18:53
jnthn I'm on 64-bit here
yeah, it built here today 18:57
jnthn installs it and builds an NQP and Rakudo against it.
pmichaud: Builds went fine. 19:04
19:04 alester joined 19:05 docdoctor left
pmichaud builds seem to be okay here too... spectesting 19:06
I get one fail under linux
t/spec/S02-types/declare.rakudo (Wstat: 256 Tests: 40 Failed: 0) 19:07
jnthn oh, that's not just me
pmichaud Parse errors: Bad plan. You planned 79 tests but ran 40.
19:07 yarp joined
pmichaud ...we don't have anyone scheduled for this month's release? 19:08
jnthn r: use Test; my KeySet $paxa; isa_ok($paxa,KeySet );
p6eval rakudo ec413e: OUTPUT«Parameter '' requires an instance, but a type object was passed␤ in submethod perl at src/gen/CORE.setting:11396␤ in sub isa_ok at lib/Test.pm:151␤ in block at /tmp/h24rarP9vl:1␤␤»
jnthn wtf 19:09
r: use Test; my KeySet $paxa; say $paxa 19:10
p6eval rakudo ec413e: OUTPUT«Parameter '' requires an instance, but a type object was passed␤ in submethod gist at src/gen/CORE.setting:11395␤ in sub say at src/gen/CORE.setting:7173␤ in block at /tmp/RuoSzyOCb3:1␤␤»
jnthn r: use Test; my KeySet $paxa; say $paxa.WHAT
p6eval rakudo ec413e: OUTPUT«Parameter '' requires an instance, but a type object was passed␤ in submethod gist at src/gen/CORE.setting:11395␤ in sub say at src/gen/CORE.setting:7173␤ in block at /tmp/hJI7SFasBs:1␤␤»
jnthn r: use Test; my KeySet $paxa; say $paxa.^name
p6eval rakudo ec413e: OUTPUT«KeySet␤»
19:10 M_o_C left, docdoctor joined
jnthn oh... 19:12
commit de96bae506667ba3e639dcddbcb81bdb08fda955 Unfudge KeySet test for Rakudo
GlitchMr: ^^ seems to have been a wrong unfudge?
GlitchMr > KeySet.new 19:13
keyset()
huh?
> my KeySet $fail = KeySet.new
masak nr: keyset(); say "alive"
jnthn pmichaud: Looks like an unfudging of a still failing test rather than a regression
GlitchMr keyset()
p6eval niecza v21-11-g09989b4: OUTPUT«===SORRY!===␤␤Undeclared routine:␤ 'keyset' used at line 1␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1435 (die @ 5) ␤ at /home/p6eval/niecza/src/STD.pm6 line 1147 (P6.comp_unit @ 37) ␤ at /h…
..rakudo 43ceae: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&keyset' called (line 1)␤»
jnthn GlitchMr: Try just say KeySet
GlitchMr This appears to be Rakudo bug 19:14
pmichaud jnthn: wfm. I'm getting fails in the windows build.
(doesn't look like anything major.)
jnthn pmichaud: Mostly what's left is line endings issues
pmichaud: I cleared up some of them this week.
GlitchMr KeySet type is implemented, isn't it?
jnthn Yeah
submethod gist(Any:D $ : --> Str) { "keyset({ %!elems.keys».gist.join(', ') })" }
Why is that a submethod rather than a multi method like all the other gist methods? 19:15
dalek nda: 5e4ee61 | (Tobias Leich)++ | bin/panda.bat:
batch wrapper for windows guys
nda: 56149b2 | tadzik++ | bin/panda.bat:
Merge pull request #17 from FROGGS/master

added bin/panda.bat
tadzik FROGGS++
FROGGS a bitch rapper, err batch wrapper
19:16 brrt joined
tadzik :P 19:16
FROGGS but it doesnt fetch 19:17
==> Fetching File::Spec
Too many arguments.
tadzik :|
FROGGS so there will be another pull request in a minute
:o)
tadzik you fix it, I don't have a windows env :P 19:18
okay :)
jnthn pmichaud: Seeing if I can fix the bug rather than re-fudge the test :)
pmichaud jnthn: I'm fine with either solution. :) 19:19
jnthn Got a patch here, spectesting it to make sure nothing else breaks :) 19:20
19:21 brrt left
dalek p/kill-nqpattr: 87eb6c8 | jnthn++ | src/core/NQPRoutine.pm:
Stub in an NQPRegex type.

This will serve as the code object for regexes, rules and tokens, and will come to hold the caps, NFA, etc.		 19:22
p/kill-nqpattr: b24e129 | jnthn++ | src/NQP/World.pm:
Enable create_code to creates code object types besides NQPRoutine.
p/kill-nqpattr: 024c643 | jnthn++ | src/NQP/Actions.pm:
Start creating regexes as NQPRegex.
FROGGS how do I get a .pir out of a .pm again? 19:25
dalek kudo/nom: 690dadc | jnthn++ | src/core/Set.pm:
Fix 'say KeySet'.
jnthn perl6 --target=pir --output=Foo.pir Foo.pm
FROGGS thanks
jnthn pmichaud: That should do it.
pmichaud jnthn++ 19:26
jnthn pmichaud: Planning to eliminate nqpattr, but not before this month's releases. 19:27
pmichaud jnthn: that sounds excellent to me 19:28
jnthn Then we'll be down to just one cheat. :)
pmichaud when we no longer cheat, will we still be able to play this game? ;-) 19:29
jnthn Don't worry, I'm only counting the cheats that are actually in src/cheats :P 19:30
pmichaud so, we don't count the cheats where we cheat on calling them cheats
jnthn Right. We're cheating at cheating. 19:31
Which is fine as we don't get caught not getting caught.
pmichaud I have a fairly long list of win fails
gist.github.com/3739287 19:32
tadzik jnthn: do you need some testing on kill-nqpattr? 19:35
I'm rebuilding the toolchain anyway
19:35 snearch joined
jnthn tadzik: oh, not for a good while yet 19:36
tadzik okay
jnthn tadzik: This branch will probably take some time...I sure won't get through it today
pmichaud: Hmm...that's a couple more than I get
tadzik okay, sure
pmichaud I'm afk, kid pickup 19:37
sorear good * #perl6 19:39
tadzik hi sorear
masak sorear! \o/ 19:40
19:41 docdoctor left
FROGGS damn it, there is no prove on windows 19:50
tadzik ah
GlitchMr FROGGS: do you have Perl 5 installed?
tadzik I was wondering whether it'd be better to use parrot-prove
FROGGS who can I bug to have stuff in rakudo/bin?
jnthn There...there isn't? :)
GlitchMr prove is part of Perl 5
FROGGS GlitchMr: I have several, but not in PATH right now, shouldnt it work without perl5?
GlitchMr Add prove to PATH and it should be there 19:51
FROGGS thats not the point
tadzik FROGGS: obtw, we'll need some Test Harness in Perl 6 sooner or later :)
GlitchMr glitchmr@strawberry ~/g/perl6-Text-Abbrev> prove -eperl6
t/00-basic.t .. ok
All tests successful.
It just works! 19:52
19:52 cognominal left
FROGGS GlitchMr: IMO it would make sense that rakudo comes with its own prove 19:52
masak (Nicholas Clark)++ # rt.perl.org/rt3//Public/Bug/Displa...?id=114946
GlitchMr Perhaps somebody could make App::Prove or something 19:53
FROGGS tadzik: I believe thats to heavy for me to do
GlitchMr But that isn't that important - Perl 5's prove works
jnthn masak: aye, interesting ticket 19:55
Two approaches enter. Which will leave? :)
masak ULTIMATE IMPLEMENTOR SMACKDOWN
jnthn That sounds painful :P 19:56
masak it's for the benefit of the users!
tadzik FROGGS: dang. That means I'll have to ;)
...try
FROGGS meh
can't we get schwern to do it?
tadzik hehe 19:57
pmichaud jnthn: (#114946) .... ummm, I don't remember Rakudo working the way it appears to be working now. 19:59
(snapshot semantics)
jnthn pmichaud: It's been that way ever since nom, when we eliminated chains of containers.
pmichaud I don't remember eliminating chains of containers, either. 20:00
masak ...which was a big improvement, IIRC.
pmichaud I'm guessing that happened during my absence.
jnthn I...thought we discussed the nom container model and agreed on it, way back?
20:00 cognominal joined
jnthn I'm fairly sure there was some discussion on this. 20:01
pmichaud I don't remember eliminating the chaining effect for ::=, no.
jnthn :/ 20:02
Well, clearly the current model is good enough to go a year without anybody noticing anything amiss.
pmichaud we even had spectests for the previous model
what does niecza think? 20:04
20:04 docdoctor joined, GlitchMr left
pmichaud n: my $x = 3; sub abc() { $x = 4; }; sub def($y) { say $y; abc(); say $y; }; def($x); say $x; 20:04
p6eval niecza v21-11-g09989b4: OUTPUT«3␤3␤4␤»
pmichaud okay, niecza agrees with the current model.
jnthn rn: my $x = 3; sub abc() { $x = 4; }; sub def($y) { say $y; abc(); say $y; }; def($x); say $x; 20:05
p6eval rakudo 43ceae, niecza v21-11-g09989b4: OUTPUT«3␤3␤4␤»
pmichaud r: my $x = 3; sub abc() { $x = 4; }; sub def(\y) { say y; abc(); say y; }; def($x); say $x;
p6eval rakudo 43ceae: OUTPUT«3␤4␤4␤»
pmichaud well, if rakudo and niecza both agree, I think I'm okay with the new semantic. I'm just surprised by the change, is all. 20:06
jnthn *nod*
pmichaud the spec might need to be clarified on this point, though.
jnthn will wait and see what TimToady++ has to say 20:07
pmichaud in particular, "Parameters and arguments" in S06 leads off with "By default, all parameters are readonly aliases to their corresponding arguments--the parameter is just another name for the original argument, but the argument can't be modified through it." 20:08
jnthn Yeah, the current model contradicts that.
masak hm, wow. that doesn't sound like pass-by-value to me. 20:09
pmichaud right, it's not.
as I said, that's why the change surprises me. :)
jnthn If we have to do that, I'd rather find a way to track non-modifiability that doesn't involve multi-leve containers.
*level
pmichaud for readonly I don't think we need multi-level.
jnthn For readonly we'd better be able to statically analyze a bunch of cases.
pmichaud i.e., only add a container if the argument is rw
jnthn Well, then we gotta be careful we don't get things bound into $foo that will flatten. 20:10
Unless the model we have there makes that a non-issue.
pmichaud it's still an issue but might end up not being one 20:11
jnthn OK
20:11 cognominal left
pmichaud the other possibility is for readonly parameters to collapse any chains they find 20:11
although that might be tricky as well. 20:12
jnthn Well, yeah
We get a lot of conceptual simplicity out of knowing it's 1-deep
pmichaud anyway, I suspect that technically speaking, RT #114946 is a legitimate bug per the spec, so either the spec has to change or the implementation does. 20:13
jnthn Agree
TimToady still would like to see an ownership or capability model of writing such that passing an item out of a domain automatically defaults to readonly 20:14
dalek nda: 6b188a2 | (Tobias Leich)++ | lib/Panda/Fetcher.pm:
quote dest dir

There can be spaces in path like in the default directory under windows. HOME under windows is `C:\Documents and Settings\...`
nda: 978dec0 | tadzik++ | lib/Panda/Fetcher.pm:
Merge pull request #18 from FROGGS/master

Quoting $dest dir
tadzik FROGGS: are you developing on windowse usually?
jnthn TimToady: How would the case discussed in the ticket work out with that, ooc?
pmichaud TimToady: ah yes, I remember that discussion now. I need to think on that one a bit more. 20:15
TimToady I would need to look at it, but am a bit distracted at the moment...
jnthn no hurry :)
pmichaud yeah, I need to fold the yapc::eu discussions into my mental model a bit.
I remember the outlines of the ownership model we discussed; I just need to follow the ramifications a bit further. 20:16
essentially the idea was to eliminate the rw flag altogether in favor of a "who owns (i.e. can write to) this thing" model. 20:18
sorear this ties into the question of how we're ever going to handle readonly aggregates
jnthn bbi10 20:19
sorear will wait for pmichaud to understand the capability model before doing anything with it
pmichaud sorear: that seems very reasonable; it'll undoubtedly require some negotiated exploring here :)
I'm afk also for a while
20:19 cognominal joined
FROGGS tadzik: last pull request for today ;o) 20:20
pmichaud between europe trip and local stuff, my office is a disaster that needs cleaning + organizing
it's lowering my productivity by a noticeable amount
FROGGS tadzik: I used to, but since a year ore more I'm on linux, but still have a few VMs for testing
tadzik nice 20:22
20:22 marmay left
FROGGS its good to have these things, otherwise I wouldnt be able to test File::Spec right 20:22
I have Mac OSX and several BSD's too
dalek nda: 7fce5bc | (Tobias Leich)++ | bin/panda.bat:
using absolute path to original panda script file
 20:25
nda: 50c4156 | (Tobias Leich)++ | lib/Panda/Fetcher.pm:
Merge branch 'master' of github.com:FROGGS/panda
nda: 2db413d | tadzik++ | bin/panda.bat:
Merge pull request #19 from FROGGS/master

Using absolute path to panda script
tadzik FROGGS++ on hacking spree :)
FROGGS btw, panda should make bat files too when copying stuff to bin on windows 20:27
tadzik hm
FROGGS in perl5 there is a pl2bat
tadzik probably, yes 20:28
20:28 kst joined
FROGGS I mean, its not that important right now since there are not that many modules carrying stuff in bin around 20:28
tadzik we'll need that eventually 20:29
20:37 cognominal left 20:39 benabik left 20:40 benabik joined
tadzik sleeps. Good night #perl6 20:42
20:42 mdupont left
FROGGS gnight tadzik 20:43
felher night tadzik
20:43 sqirrel left 20:46 cognominal joined 20:47 yarp left, thou left, kaare_ left 20:48 orafu joined 20:51 mucker left
masak good knight o tadziku. 20:53
20:54 mucker joined 21:01 snearch left 21:03 docdoctor left 21:05 skids left 21:06 mucker left, flightrecorder left 21:09 sivoais left 21:14 FROGGS left, tgt left 21:21 sivoais joined
pmurias .perl doesn't print out private attributes? 21:29
jnthn No, by design 21:31
(see long p6l discussion for more :))
pmurias how can I print out them? 21:32
jnthn Use the MOP
.^attributes gives you attribute meta-objects, and you can call .get_value and pass in the object
pmurias that's extremly annoying 21:33
jnthn It's meant to be.
pmurias and shouldn't .gist print out them? 21:34
jnthn No.
(We've already been there, and switched to keeping private things private.)
sorear I think Damian even wanted to do away with get_value 21:35
pmurias jnthn: where can I find the perl6 language thread? 21:36
jnthn sorear: I think so
pmurias: I forget the subject, and I don't know off hand where the best place to read the mailing list archive is 21:37
I forget exactly when it was too. Hm
masak sorear: Damian wanted encapsulation at the cost of the MOP. 21:38
jnthn There needs to be an escape valve somewhere. I think we hid it reasonably well. 21:39
It annoys pmurias++ sufficiently, at least. ;-)
pmurias having to use a module for a common things seems annoying 21:40
masak the thread is called "[perl6/specs] 58fe2d: [S12] spec setting and getting values of attribute..." 21:41
doy isn't the point that externally accessing a private attribute *shouldn't* be a common thing? 21:42
jnthn Exactly.
pmurias doy: when debugging? 21:47
masak when debugging, or when serializing, or when instantiating. that's when we typically like to break encapsulation.
but .perl produces "normal" Perl 6 code, in some sense. 21:48
pmurias and what about .gist?
jnthn
.oO( debugger feature idea! :) )
hm, earlyish $dayjob tomorrow I guess 21:50
'night
masak 'night, jnthn 21:54
21:57 thou joined
geekosaur .gist is, as the name suggests, just the gist of it; it should probably print only enough to uniquely identify 21:57
masak 'night 22:12
22:15 am0c left 22:18 whiteknight joined 22:23 tharple joined 22:25 tharple left 22:28 raiph joined 22:29 whiteknight left
raiph hi all 22:30
22:30 PacoAir left
sorear hello raiph 22:31
raiph i'm curious to know who added the read only alias / ownership capability discussion to today's summary
hi sorear
22:34 spider-mario left 22:40 wamba left 22:43 stopbit left 22:48 pmurias left 23:10 skids joined 23:24 Circlepuller_ joined 23:25 sisar left 23:26 Circlepuller left 23:28 fgomez left 23:36 fgomez joined 23:55 dayangkun joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!