»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.
jnthn is safely home from GPW :) 00:04
phenny jnthn: 07 Mar 21:17Z <tadzik> ask jnthn how hard would it be to make S14-roles/attributes.t run on rakudo?
jnthn phenny: tell tadzik I think that test may be out of date...will try and take a look in the next couple of days 00:05
phenny jnthn: I'll pass that on when tadzik is around.
00:06 att left
jnthn phenny: tell colomon on Win64, Parrot needs --parrot-option=--intval="long long" 00:06
phenny jnthn: I'll pass that on when colomon is around.
colomon jnthn: yes, working on getting a build right now 00:07
phenny colomon: 00:06Z <jnthn> tell colomon on Win64, Parrot needs --parrot-option=--intval="long long"
00:08 sivoais joined
sorear good * #perl6 00:09
pmurias: I am unqualified to answer that question. 00:10
00:13 mtk left 00:16 whiteknight joined
colomon o/ 00:16
00:34 pmurias left 00:56 thou left 01:02 att joined 01:07 scott__ joined 01:11 cbk1090 joined
cbk1090 Hello #perl6 01:12
01:25 sporous left 01:27 sporous joined
colomon o/ 01:27
01:33 havenn joined 01:34 alc joined 01:37 overrosy left 01:38 overrosy joined, {}o_0{} joined
{}o_0{} hi! I'm a snow crab! 01:39
01:39 havenn left 01:48 alc left
dalek p/new_make: 23eaebb | bacek++ | tools/build/Makefile.in:
Change PMC building rules to enable parallel build.
 01:58
02:02 alc joined
dalek p/new_make: ece5fbf | bacek++ | tools/build/Makefile.in:
Add dependencies for PMC building.
 02:08
02:09 tokuhirom left
[Coke] pugs: 3.Int.say 02:11
p6eval pugs b927740: OUTPUT«*** No such method in class Int: "&Int"␤ at /tmp/sgBWnhp8va line 1, column 1 - line 2, column 1␤»
dalek p/new_make: 03ac24e | bacek++ | tools/build/Makefile.in:
Finish PMC building rules.
 02:20
02:27 isBEKaml left 02:29 wolfman2000 joined 02:31 jferrero left 02:35 jferrero joined 02:51 havenn joined, xinming joined
dalek gs.hs: e06a935 | coke++ | Pugs/src/Pugs/Prim.hs:
Add .Bool coercion

Fixes #8		 02:59
gs.hs: 0d7fe82 | coke++ | .gitignore:
ignore some generated files
ast: 7857696 | coke++ | S (11 files):
pugs unfudge
 03:01
03:08 orafu left 03:09 orafu joined
dalek p/new_make: 40bae77 | bacek++ | tools/build/Makefile.in:
Rework bigin ops building
 03:11
03:12 lurkster joined 03:18 alester_ joined
dalek p/new_make: ae037b7 | bacek++ | tools/build/Makefile.in:
Bring all PMC deps into one place
 03:23
p/new_make: 7261b1c | bacek++ | tools/build/Makefile.in:
Don't use bare '.o' for file extension.
 03:37
p/new_make: 7116b98 | bacek++ | tools/build/Makefile.in:
Use variable instead of hardcoded path
p/new_make: 67e4e13 | bacek++ | tools/build/Makefile.in:
Fix previous commit
 03:39
03:44 xinming left 03:45 xenu left
dalek gs.hs: 56c053f | coke++ | t/spectest.data:
run more tests.
 03:47
03:48 alc left 03:50 skipper joined
dalek p/new_make: 7cee7ce | bacek++ | tools/build/Makefile.in:
And again...
 03:50
03:52 havenn left
[Coke] perl6: say Q{hello}; 03:56
p6eval rakudo 182b2f, niecza v15-4-g1f35f89: OUTPUT«hello␤»
..pugs b927740: OUTPUT«*** No such subroutine: "&Q"␤ at /tmp/Ahg7zmJctu line 1, column 5-13␤»
[Coke] n: say 18*24 04:05
p6eval niecza v15-4-g1f35f89: OUTPUT«432␤»
04:05 marmalade left
sorear o/ [Coke] 04:06
colomon \o 04:08
[Coke] perl6: say 35.unpolar(0.8) 04:12
p6eval niecza v15-4-g1f35f89: OUTPUT«24.384734827150787+25.107463181483297i␤»
..rakudo 182b2f: OUTPUT«24.3847348271508+25.1074631814833i␤»
..pugs b927740: OUTPUT«*** No such method in class Int: "&unpolar"␤ at /tmp/3SDyM2vmz_ line 1, column 5 - line 2, column 1␤»
[Coke] sorear, colomon: o/
colomon bedtime here, actually 04:15
04:15 alvis left
dalek ast: 062d17a | coke++ | S32-num/complex.t:
(slightly complex) pugs fudge
 04:20
[Coke] perl6: say 1/2R 04:21
p6eval rakudo 182b2f: OUTPUT«===SORRY!===␤Confused␤at /tmp/RY6JLoENA1:1␤»
..pugs b927740: OUTPUT«*** ␤ Unexpected "R"␤ expecting "_", fraction, exponent, term postfix, operator, ":" or ","␤ at /tmp/wqyznLHWuC line 1, column 8␤»
..niecza v15-4-g1f35f89: OUTPUT«===SORRY!===␤␤Whitespace is required between alphanumeric tokens at /tmp/W2tQJJgtxL line 1:␤------> say 1/2⏏R␤␤Two terms in a row at /tmp/W2tQJJgtxL line 1:␤------> say 1/2⏏R␤␤Parse failed␤␤»…
04:25 lurkster left
timotimo oh, cute, with something like postfix:<kg>(Numeric $foo) you could come up with units and stuff. that's neat 04:31
04:32 alvis joined
timotimo perl6: sub postfix:<kg>(Numeric $val) { say "$val kilograms"; }; 9001kg; 04:32
p6eval pugs b927740, rakudo 182b2f, niecza v15-4-g1f35f89: OUTPUT«9001 kilograms␤»
dalek p/new_make: e7ff7f5 | bacek++ | tools/build/Makefile.in:
DRY: use autovariable.
04:33 alc joined
sorear timotimo: I think it exists for CORE::postfix:<i> 04:34
as in 3 + 4i
04:35 alc left
bacek_at_work ~~ 04:38
phenny, tell moritz Can you try nqp new_make branch? If you have something like Win32 it will be very helpful
phenny bacek_at_work: I'll pass that on when moritz is around.
bacek_at_work phenny, tell jnthn Can you try nqp new_make branch? If you have something like Win32 it will be very helpful
phenny bacek_at_work: I'll pass that on when jnthn is around.
04:46 birdwindupbird joined 04:48 snearch joined
dalek p/new_make: 1971fe7 | bacek++ | src/6model/Defines.mak:
Add more dependencies to enable parallel build.
 04:55
05:00 skipper left 05:04 skids left 05:06 kaleem joined 05:16 lestrrat left 05:17 lestrrat joined 05:18 xinming joined 05:21 thou joined 05:33 envi_ left 05:38 xinming_ joined, xinming left 05:40 alester_ left 05:53 birdwindupbird left 06:13 xinming_ left 06:14 xinming joined, NamelessTee joined 06:23 xinming left 06:49 xinming joined 06:55 snearch left 06:57 NamelessTee left 07:12 preflex_ joined 07:13 preflex left, preflex_ is now known as preflex
tadzik good morning 07:15
phenny tadzik: 00:05Z <jnthn> tell tadzik I think that test may be out of date...will try and take a look in the next couple of days
07:17 key__ joined, key__ is now known as key_
moritz \o 07:17
phenny moritz: 04:38Z <bacek_at_work> tell moritz Can you try nqp new_make branch? If you have something like Win32 it will be very helpful
dalek p: 2195abb | moritz++ | / (13 files):
Merge remote branch 'origin/new_getprop'
 07:19
07:21 thou left
sorear o/ tadzik , moritz 07:26
nebuchadnezzar morning 07:27
07:29 woosley joined
moritz phenny: tell jnthn the nqp branch new_make works fine for me on linux. If it also works for you, please merge it 07:30
phenny moritz: I'll pass that on when jnthn is around.
07:50 bynari left 07:57 birdwindupbird joined
bacek moritz, there is some problem with parallel build of dyncall in nqp... 08:07
I couldn't make it to behave properly.
moritz that's knowin-ish 08:14
*known-ish
08:16 birdwindupbird left 08:28 birdwindupbird joined 08:37 wtw joined 08:38 xinming left 08:41 xinming joined 08:46 sisar joined 08:47 woosley left 09:07 mj41 joined 09:12 birdwindupbird left 09:14 birdwindupbird joined 09:17 key_ left 09:19 birdwindupbird left
masak morning, #perl6 09:21
GPW++ 09:22
moritz++ # GPW
eiro hi all
moritz masak: did the journey home work out well? 09:23
eiro journeesperl.fr/fpw2012/newtalk < masak, tadzik :)
tadzik, can you talk about bailador ? 09:24
(if you concider joining the event)
moritz hm, strasbourg is about 5..6 hours by train 09:25
masak moritz: yes, hitchlessly. 09:26
09:27 am0c joined
moritz nom: class A { has package B { }} 09:27
p6eval rakudo 182b2f: OUTPUT«===SORRY!===␤Cannot use has scope with package at line 1, near "{ }}"␤»
tadzik eiro: I will consider that, although I might have some exams at the very end of June. As soon as I get to know that I don't have any I'd gladly join you on the workshop 09:31
eiro moritz, masak -> FKB 09:32
moritz FKB?
eiro moritz, the lowcost airport of baden baden
tadzik, cool
moritz eiro: if I come, it'll be by train, not flight
eiro ok ... 5 hours of train... you are a warrior 09:33
tadzik althogh there's always a chance that I either do my exams on the first attempt and/or procrastinate some of them for September :)
eiro tadzik, don't! exams come first!
moritz because going to the airport, checking in, the flight, waiting for the baggage etc. and all the time safety margins make the flight not much faster
tadzik heh, well
moritz and for me trains are much less stress than flights 09:34
eiro moritz, ok then
DrEeevil moritz: bern-berlin, train is at least as fast and more comfy
eiro btw: please make some sound about fpw in germany: we're closer to you than ever :) 09:35
also i have to contact rafl :)
moritz eiro: if you had said that one day earlier, I could have announced that in the closing talk of the GPW
eiro moritz, i was at GPW '12 09:36
s/12/11
gpw12 still happenned ?
tadzik yeah, ended yesterday :)
moritz from Monday to yesterday 09:37
eiro DAMNED!
moritz eiro: around March is the usual time for it; last year it was unusually late
eiro ok 09:38
09:47 {}o_0{} left
dalek p/new_make: 96c5cef | bacek++ | / (2 files):
Hackish workaround to build dyncall in single job to enable build of rest of nqp in parallel
 09:50
bacek moritz, ^^^ :)
moritz bacek: does that work on non-GNU makes? 09:51
09:53 mikemol left
jnthn mornin' 09:54
phenny jnthn: 04:38Z <bacek_at_work> tell jnthn Can you try nqp new_make branch? If you have something like Win32 it will be very helpful
jnthn: 07:30Z <moritz> tell jnthn the nqp branch new_make works fine for me on linux. If it also works for you, please merge it
jnthn Microsoft (R) Program Maintenance Utility Version 10.00.30319.01 09:56
Copyright (C) Microsoft Corporation. All rights reserved.
makefile(474) : fatal error U1001: syntax error : illegal character '^' in macro
eiro moritz, was rene there ?
jnthn No, it doesn't work.
bacek jnthn, sigh...
moritz eiro: yes
eiro renee beck
moritz baecker iirc 09:57
jnthn bacek: I...assume the only thing you're trying to make parallel is the C bits?
eiro and he said nothing ? he was aware of FPW!
moritz eiro: he said lots of things :-)
jnthn bacek: Well, and the stage 0 PIR can be too I guess
dalek p/new_make: cdbb25d | bacek++ | tools/build/Makefile.in:
Revert "DRY: use autovariable."

It doesn't work with nmake.
tadzik jnthn: the spec says "the private attr of a role is private to a class, not to a role"
09:57 key_ joined
jnthn Stage 1 and stage 2 offer just about no chance of parallelism 09:57
moritz like, gave a 40 minutes talk, and telling people about YAPC::EU, and about the next GPW and so on :-)
he might have simply forgotten about it 09:58
bacek jnthn, I'm trying to "streamline" build process to simplify migration of parrot to 6model :)
jnthn, can you try again?
jnthn bacek: Just did. Now:
LINK : fatal error LNK1181: cannot open input file 'src\pmc\nqp_group.obj'
tadzik I was trying to refactor Set stuff
jnthn Also 09:59
eiro twitter.com/#!/marcchantreux/statu...9190993920 < he knew ! :)
bacek jnthn, strange...
jnthn cl : Command line warning D9035 : option 'o' has been deprecated and will be rem
oved in a future release
bacek jnthn, for which target?
jnthn, can you create gist with make log by any chance? 10:00
eiro moritz, sure. and i think it's still time to spread information in germany 10:01
moritz sure
jnthn bacek: gist.github.com/2000073 10:02
10:02 dakkar joined
jnthn eiro: I've reserved space in my schedule for FPW, and hope to be able to make it. 10:03
eiro jnthn, thanks!
please everyone fill the newtalk page journeesperl.fr/fpw2012/newtalk soon even if you're not sure to be able 10:05
(to come)
also: we found a very cheap but cosy accomodation (35,70€ B&B) 10:06
dalek p/new_make: cd61e1d | bacek++ | tools/build/Makefile.in:
Use proper macro instead of '-o'. jnthn++
cognominal eiro \o/
bacek jnthn, oookey. Warning should gone. I don't quite understand why nqp_group.obj is missing. It was definitely compiled according to log. 10:07
jnthn gist.github.com/2000101 10:08
That one looks odd.
eiro cognominal, \o/ 10:09
cognominal, we also have to add cat in this loop
cognominal cat? nyan cat? 10:10
bacek jnthn, looks like a bug in opsc. I have to talk to opsc owner. Wait... It's me.
10:10 Chillance joined
jnthn bacek: Talking to yourself considered normal. :) 10:10
bacek jnthn, www.fmylife.com/miscellaneous/19236548 :)
10:11 scott__ left
cognominal jnthn, with all the stuff that is going in rakudo, you should write a teaser for Perl 6 that eiro could add to the fpw site. 10:11
… pointing to your next blog entry :) 10:12
moritz bacek: you might enjoy github.com/masak/gpw-talks/blob/ma...f?raw=true
10:12 am0c left
moritz that's the slides of a lightning talk masak++ gave on Tuesday 10:13
eiro cognominal, cat -> Christian (another perl6 fan) 10:16
timotimo the zoom at the end was amazing
bacek jnthn, can you try it with bleeding edge parrot? :) 10:17
masak timotimo: thanks :) 10:18
jnthn bacek: Yeah, but I gotta do $dayjob stuff now, having ignored it for the last three days to do German Perl Workshop. Will give it a go later today. 10:19
bacek moritz++ masak++ nice one!
jnthn, ok, thanks
masak timotimo: I feel I'm only discovering what can be done with this format. I definitely want to play around more with it.
moritz masak: having seen you writing the slides, I think you should optimize the source format a bit for less copy and paste 10:20
masak I see what you mean. 10:25
that part doesn't bother me so much. the thought crossed my mind to be able to lay out the scene frames in Inkscape, however. something like export-to-svg, move boxes in Inkscape, import-back-to-slides. 10:26
10:26 ascrazy joined
masak positioning is probably best done in Inkscape. 10:27
cognominal jnthn, are you gpw slides posted somewhere?
moritz cognominal: it's in the backlog somewhere in the last three days. search for .pdf :-)
cognominal ok
thx, moritz 10:28
masak cognominal: irclog.perlgeek.de/perl6/2012-03-06#i_5251854
moritz: my idea is to prototype Niobium (the diagramming tool) inside of Gryd (the presentation tool). 10:29
cognominal thx masak
10:29 ascrazy left
cognominal masak: have you watched kineticjs? it is almost there. That's html5 canvases with scene memorization. It lacks css selectors to compete with SVG. 10:30
it seemas that html5 canvases support is better than SVG support theses days. 10:31
kineticjs.com/
10:34 envi_ joined, sisar left
dalek kudo/nom: 2407a5b | moritz++ | src/ (4 files):
Merge branch 'world-API' into nom
 10:40
kudo/nom: e09d7f8 | moritz++ | src/ (3 files):
generalize X::Sub::Scope to X::Declaration::Scope

and throw it in one more place. Also throw another instance of X::Signature::Placeholder
moritz nom: role A[$x] { }; class B does A[1 + 2] { } 10:41
p6eval rakudo 182b2f: OUTPUT«===SORRY!===␤Cannot use '1 + 2' as an argument to a parametric role as its value is not known at compile time at line 1, near " { }"␤»
moritz LTA 10:42
jnthn Well, bs was a blocker for the constant folder before now. 10:45
moritz I know
should it treat the argument the same way as the RHS of a constant declaration?
jnthn Very unlikely. 10:46
That just puts the RHS as a BEGIN time thing
oh
hm
Maybe that *would* work out... ;)
moritz which is exactly what you do when you want a value at compile time :-)
10:47 whiteknight left
moritz goes through panics in Perl6::World 10:47
jnthn We probably need to try and map out "how hard to try" in a bunch of cases.
10:47 key_ left
jnthn Like, run the thing as BEGIN, run the constant folder, etc. 10:48
moritz nom: my $x = 3; constant a = $x ** 2; say a
p6eval rakudo 182b2f: OUTPUT«use of uninitialized variable $x of type Any in numeric context␤0␤»
10:50 xinming left
jnthn That assignment doesn't take place until runtime. 10:50
moritz I know
jnthn Nice that we get the warning. :) 10:51
moritz I was just pondering if such things should generally be forbidden, or trigger a warning at compile time, or simply use the static pad and be fine with it
jnthn It did 2 and 3 :)
Used the static pad, which held an uninitialized value, and we got a warning. :) 10:52
10:55 xinming joined 11:01 Trashlord joined 11:26 NamelessTee joined 11:29 envi_ left 11:52 Patterner left 11:54 Psyche^ joined, Psyche^ is now known as Patterner 12:06 y3llow_ joined 12:08 y3llow left 12:09 y3llow_ is now known as y3llow 12:11 PerlJam left 12:12 bacek left, tadzik left 12:13 Juerd left 12:17 PerlJam joined 12:24 am0c joined
masak is feather under heavy load? things are lagging a bit. 12:32
moritz executing 'uptime' will tell you 12:34
masak 13:34:42 up 139 days, 6:49, 5 users, load average: 0.03, 0.06, 0.05 12:35
those load average numbers never meant much to me.
moritz it means there's nearly now CPU load on feather
masak right.
moritz and suggests that the lag comes from the network instead 12:36
(though other explanations are possible too)
masak ok.
ah, the numbers are for the past 1, 5, and 15 minutes, respectively.
arnsholt The rule of thumb with load averages is that it shouldn't exceed the number of CPUs in the machine 12:37
moritz yes, the load average is basically the number of CPUs you'd need to perfectly do all the calculations that are running (or have been running)
arnsholt And the three numbers are over different time spans. First is "right now", second is "a little while" and last "a long while"
moritz 1, 5 and 15 minutes iirc 12:38
12:38 daxim joined
arnsholt Yeah, sounds about right 12:38
moritz man uptime agrees with me :-) 12:39
arnsholt Even better =) 12:40
masak <masak> ah, the numbers are for the past 1, 5, and 15 minutes, respectively. 12:41
you guys don't read what I'm writing :P
moritz did anybody say something while I was away? 12:44
12:44 tokuhirom joined
masak :P 12:45
12:46 Juerd joined
masak "masak's contributions to #perl6 were much less commented on after he became /ignore'd by the majority of core devs..." 12:46
the most damning comment and the clearest example of bynari-of-le-backlog's point -- that it's hard for any product/organization to go against established perceptions -- is a quote I got relayed to me from a former colleague. 12:49
awwaiid "and the monk-like solitude that ensued led him to deeper insights then could have been previously imagined"
masak something like "They added native types, proper object-orientation, a meta-object protocol, grammars, and metaops... onto *Perl*." 12:50
clearly it was meant as a mean joke.
as if the punchline might equally well have been "... onto *INTERCAL*." 12:51
so that's part of the perception we have against us. 12:52
I remember two years ago at the social on OSDC.fr, when a Ruby-or-whatever programmer came up to me and I introduced myself as a Perl 6 developer. 12:53
he spent the next 15 minutes explaining to me that Perl doesn't exist anymore.
awwaiid I've been lately thinking that bynari has a point in a way -- I think it's all about the "middle class" of programmers 12:54
masak somehow the language barrier, or maybe his zeal, prevented me from interrupting/contradicting him.
awwaiid haha
12:54 JimmyZ joined 12:55 am0c left
geekosaur not sure ruby counts there; ruby devs would like to believe they are the inheritors of Perl and have been ever since the English-speaking world discovered Ruby. 12:55
much the same way they think Rails is still relevant
masak Ruby *is* a descendant of Perl. 12:56
it's interesting when you go look at all the similar non-alphabetic variables and run options.
geekosaur oh, certainly
I'm not denying that at all.
jlaire it's a descendant, but becomes an inheritor only when Perl dies 12:57
masak the Perl roots are all over Ruby, even if Modern Ruby doesn't emphasize it.
geekosaur I;'m saying that your anecdote says more about the Ruby community than the Perl one
12:57 sisar joined
masak aye. 12:57
lumi__ PLT Scheme recently renamed to Racket, I guess to escape the scheme name
masak I can't swear he was a Ruby guy.
geekosaur they often act like they're desperate to be taken seriously
awwaiid what a racket! hehehehe 12:58
JimmyZ always heard somebody said language is not the problem. :)
masak geekosaur: or maybe that's just how they *seem* from the point of view of us balanced Perl programmers :)
geekosaur (seriously? they want to be taken seriously, make it possible to build a serious system that won;t be broken the next time someone installs a random gem)
,uch less a random *point release* of ruby...) 12:59
masak JimmyZ: all programming languages are opinionated. all runtimes need to have a memory model, which in the end influences what the language can and can't do. Turing equivalence still applies, but some languages make it easier to do some things.
awwaiid I'll tell you what -- here in DC the ruby group get's 50 meeting members a month, the python gets 30, and perl gets 10. Inversed to their real-world usage? Indicative of future trends? Indicitive of social-networking skillz?
geekosaur indicative of marketing skills, to a large extent 13:00
but, ruby in particular seems to be in that curious spot where they have the marketing thing working for them, but on the technical side tey're doing the "everyone else is wrong and we can ignore whatthey've come up with and do it right!" and then spent the next year or so rediscovering _The Mythical Man-Month_ etc. 13:02
awwaiid want's to start DC-Dynamic-United
13:02 alvis left
awwaiid is also too lazy to _actually0 start DDU 13:03
er, DC-DU
geekosaur (python mostly avoided that stage, by dint of GvR's steering, as did Perl by Larry's. I still have scars from when Linux went through it) 13:04
13:05 mtk joined
awwaiid anyway. perl is fun! 13:05
I been doing generator-based sound stuff in perl5 -- setting up functions that are generators, spitting out new digital samples on every call. Looks like maybe I could do it now in p6 13:07
perl6: sub f_gen { my $x = 0; -> { $x++ } } my $f = f_gen(); $f.().say ; $f.().say ; $f.().say 13:08
p6eval rakudo e09d7f: OUTPUT«===SORRY!===␤Confused␤at /tmp/J849RrwMys:1␤»
..niecza v15-4-g1f35f89: OUTPUT«===SORRY!===␤␤Strange text after block (missing comma, semicolon, comment marker?) at /tmp/d5CG7kTC9_ line 1:␤------> sub f_gen { my $x = 0; -> { $x++ } }⏏ my $f = f_gen(); $f.().say ; $f.().say ␤␤Parse failed␤␤»…
..pugs b927740: OUTPUT«0␤1␤2␤»
masak awwaiid: semicolon after }
awwaiid doh
perl6: sub f_gen { my $x = 0; -> { $x++ } }; my $f = f_gen(); $f.().say ; $f.().say ; $f.().say
p6eval pugs b927740, rakudo e09d7f, niecza v15-4-g1f35f89: OUTPUT«0␤1␤2␤» 13:09
masak also, '$f.()' better written '$f()' if you ask me.
awwaiid oh, didn't know I could. also why do I need a ; after my sub?
masak the '.' is a p5istic translation of '->'
13:09 kaleem left
moritz awwaiid: because tehre's no \n after the closing curly 13:09
awwaiid what other interpretation could it have? I must've missed this one 13:10
masak awwaiid: the rule is that ';' or "}\n" must go between statements.
well. to a first approximation.
awwaiid that for easy-consistent parsing? 13:11
or is there some other ambiguity that it solves that I'm not seeing? 13:13
13:13 Juerd_ joined, grondilu joined 13:14 Juerd_ left
grondilu Is anybody working on a perl6 grammar for mediawiki? I'd be so happy to have that, but it's kind of tough for me to write it. 13:14
masak awwaiid: basically, it's very important for the parser to be able to recognize and flag up two terms in a row. 13:17
grondilu: I did, many years ago, for November the wiki engine. I haven't heard about anyone doing it since then.
13:18 alvis joined
moritz awwaiid: how much perl 5 do you write? 13:18
grondilu masak: where is it? I'd like to have a look. 13:19
moritz in perl 5 there's this annoying thing that if you write 'somecallback sub { .... }'
then the next line triggers a syntax error 13:20
because you didn't write a ; after the }
masak grondilu: seems to be at github.com/viklund/november/blob/m...diaWiki.pm
moritz (which is especially annoying if the sub is several lines long, or even more than a vertical screen length)
grondilu masak: great. Thanks 13:22
moritz has anybody ported november to nom?
awwaiid masak, right, but the newline doesn't end it either. I guess I think of a named sub an an anon sub as two different things -- the named sub is a block construct that needs no ; after it. But I just tried this with if blocks in p6 and it works the same, so at least it's consistent :) 13:23
masak moritz: nope. 13:25
[Coke] (user groups attendences: perl here gets 0. cold fusion gets dozens.)
13:26 kaleem joined
masak r: (sub { $^f("OH HAI") })(sub { say $^m }) 13:27
p6eval rakudo e09d7f: OUTPUT«OH HAI␤»
sisar sorear: warnings seen when make-ing Niecza: gist.github.com/2000992 , line 4, 18 13:32
masak sisar: that's fairly normal, I think. par for the course. 13:33
[Coke] I think he eliminates them as possible, but the fact that we're bootstrapped means those are going to show up occasionally.
sisar masak: ok 13:34
moritz: did you figure out my yesterday's gist : gist.github.com/1986576
[Coke] pugs: say 3.Int
p6eval pugs b927740: OUTPUT«*** No such method in class Int: "&Int"␤ at /tmp/xnrgsNUN0u line 1, column 5 - line 2, column 1␤»
sisar me thinks that i might have some insight... 13:35
moritz sisar: no
awwaiid masak, anyway -- I don't mind offhand about the required [\n;], I was just surprised. For some reason (;|\n) strikes me as the path to the dark side... but I'm not afraid of no dark side 13:36
sisar if you copy-paste any multiline code on the rakudo REPL, then regardless of your next sentence, rakudo responds with 'Confused'
moritz oh
sisar moritz: are you able to reproduce it ? 13:37
[Coke] moritz: quiet ping on the pugs-on-evalbot-update question. 13:38
moritz sisar: dunno, currently I don't have a rakudo runninig $here
[Coke]: pong
13:38 lateau joined
moritz [Coke]: iirc pugs isn't currently updated on p6eval at all 13:39
and from where I am right now, I don't have access to its host :/
sisar moritz: oh, ok.
[Coke] can you try to reproduce the problem ? 13:40
just copy-paste any multiline code, let rakudo complain, then give anyother valid statement
s/statement/ one line code 13:41
[Coke] sisar: no: gist.github.com/2001030 13:42
do you have a cut and paste that shows the problem?
masak JimmyZ: perhaps more importantly, programming languages are tools for doing interesting things. any discussion on which tool is better than the other tools and why is time spent not doing interesting things. :)
sisar [Coke] one sec... 13:44
PerlJam good * #perl6
sisar [Coke]: here: gist.github.com/2001038. Copying lines 1 to 7 should trigger it... 13:45
masak PerlJam! \o/ 13:46
moritz \*/ 13:47
13:47 kaleem left
masak .oO( moritz has whatever head? ) 13:47
or maybe it's a moritz Star.
13:48 kaleem joined
moritz my head just sees stars everywhere 13:48
sisar masak: on strangelyconsistent.org/blog/june-2...at-strings "print the value using three characters", i think 'three' => 'five' 13:49
PerlJam sisar: indeed 13:51
timotimo masak: is there no option to center text? 13:52
masak sisar: good catch. thanks. 13:53
moritz timotimo: that's what Form.pm is for
timotimo oke
masak timotimo: in sprintf? no.
moritz though I have no idea what state it's in
mathw might now
masak I hope the Form.pm we end up with knows how to handle double-width chars.
Form.pm would be a very adequate GSoC project, methinks. 13:54
jnthn Aye
moritz nom: role B { ... } 13:55
p6eval rakudo e09d7f: ( no output )
moritz why doesn't that complain that B is declared but not stubbed?
erm, other way round :-)
jnthn Roles are special cased...it's some trickiness with "which parametric variant did we stub"... 13:57
moritz oh.
jnthn It needs a better solution.
dalek kudo/nom: 4235dfe | moritz++ | src/ (2 files):
X::Package::Stubbed

also automagically promote RPA to Parcel in World.throw		 13:58
ast: 46f9b17 | moritz++ | S32-exceptions/misc.t:
track exception name change, and add another test
 13:59
ast: ddc21bf | moritz++ | S32-exceptions/misc.t:
test X::Package::Stubbed
masak how to compare two strings in a case-insensitive way in Perl 6?
moritz you mean besides $a.lc eq $b.lc? 14:00
masak yeah. I read brian d foy's blog post, and learned from it that .lc isn't enuf.
www.effectiveperlprogramming.com/blog/1507 14:01
I guess what I'm asking is, does Perl 6 have .fc or something equivalent?
if not, this seems to be something that we'd want to steal from 5.16. 14:02
14:02 birdwindupbird joined
geekosaur I thought this had come up before and something got spec-ed 14:03
(not sure it ever got implemented but I know it was discussed) 14:04
14:04 kranius joined
kranius hello #perl6 14:05
14:06 benabik left
masak kranius: greetings, kranius. 14:06
awwaiid sisar, it's not just copy-paste -- just typing "for 2..100 {<enter>" triggers that
masak kranius: welcome back.
14:09 tadzik joined
kranius would jumping straight to perl6 be good for a perl newcomer ? 14:11
PerlJam kranius: probably not.
kranius: (but that's just my opinion) 14:12
kranius why so ?
14:12 JimmyZ_ joined 14:14 JimmyZ left, JimmyZ_ is now known as JimmyZ
masak I'm less hesitant than PerlJam. Perl 6 might work fine as a "first Perl". 14:14
PerlJam kranius: because there are two Perls now. Perl 5 and Perl 6 and they are different. Perl 5 has a long history and lots of tools and lots of edifice (CPAN, cpan testers, etc.) that Just Works for the programmer. Perl 6 doesn't quite have the same level of maturity with respect to its corresponding tools.
masak but there's... right, tools and documentation and maturity in general. 14:15
timotimo i'ven't done perl5 before and i'm learning perl6 (very) slowly
masak if you choose to go with Perl 6, be prepared for more of a "work in progress" experience than with Perl 5.
timotimo i think it's very nice if you don't expect to become productive immediately
masak timotimo++
kranius yes I am aware of that. But I don't really need production readyness 14:16
PerlJam kranius: then learn Perl 6 by all means
!
kranius nor a big ecosystem
I have some experience with perl 5, but it's mainly maintaining some already existing code 14:17
timotimo you can probably get many libraries using the CLR interface of niecza if you want to do gui or whatever
kranius nothing really fancy 14:18
14:19 bynari joined 14:20 tokuhirom left
PerlJam When people talk about Perl "in the wild", I bet their mental model is a Perl 5 universe. I wonder how we can effectively put the "Perl 6 vision of Perl" into people's heads? 14:20
timotimo .u 哈
phenny U+54C8 CJK UNIFIED IDEOGRAPH-54C8 (哈)
PerlJam Seems like the best way is lots of code, documentation, tutorials, articles, real-world examples, etc.
kranius yes code examples is good to spread the word 14:22
masak PerlJam: tadzik had an interesting talk on GPW about how the Perl 6 module ecosystem toolchain has followed an "evolution" path so far, despite numerous attempts to apply "intelligent design" to it. 14:23
at least that's my take on what he said. ;)
tadzik++
PerlJam cool 14:24
14:24 mikemol joined
PerlJam I feel like Perl 6 is on the cusp of something and it's getting close to the time when we need to "turn up the heat" to make things bounce around a little more :) 14:25
masak aye.
14:25 dudulz joined
kranius most people I know are scared by Perl, they have the "perl golf" in mind, unreadable snippets with cryptic syntax :/ 14:25
timotimo hehe, yes
masak there will always be people who think like that. 14:26
PerlJam kranius: "fear of Perl 5" could be replaced by "wonder of Perl 6" :)
timotimo a friend of mine does a lot of perl5 stuff and always brags about how good cpan is and about the usually very high quality of documentation and tests
masak there will always be people who go "LOL PARENTHESES" when you mention Lisp, too.
kranius true
PerlJam just thinks of fingernail clippings
TimToady++ (for the memorable quote) 14:27
14:27 bluescreen10 joined
masak and both groups, while crude, are right in some sense. 14:27
Lisp code does contain parentheses.
PerlJam masak: and a few non-parentheses too ;) 14:28
masak Perl code does contain punctuation. and you have the rope to shoot yourself in the foot.
that doesn't mean you have to.
jnthn You can...shoot yourself with rope?
masak :P
PerlJam jnthn: sure, with a rope gun
masak "Perl 6 gives you enough rope to shoot yourself in the foot." I thought this meme was well-known enough to be mundane at this point... :) 14:29
apparently not.
14:29 dudulz left
jnthn I guess I never bothered to think it through before :P 14:29
kranius that Parrot vm is sexy 14:30
masak it's a wonderful vision, yes.
PerlJam btw, anyone working on an April Fools Joke this year? 14:31
The Perl community used to do such nice AFJs
14:31 skids joined
masak PerlJam: no, but it's a great idea. 14:32
frettled «Perl 5.17 and Perl 6 to merge codebases» 14:33
masak ...upgrading the version number to Perl 11.17 14:34
PerlJam frettled: I was thinking something more like "p5p to adopt Parrot VM for Perl 5 with plans to merge with Perl 6 by 2013" or something
14:34 PacoAir joined
PerlJam (the nuance of p5p using Parrot is important :) 14:34
masak I would assume that that was an April 1st joke even if it arrived in September.
frettled hee-hee 14:35
PerlJam's suggestion is nice.
It would be even better if the current uhm er buildmaster or whattheycallit for Perl 5 signed onto that.
masak oh yes. 14:36
PerlJam "Larry Wall gives up on Perl and moves to the Himalayas"
masak with a credible enuf timeline up until 2013.
PerlJam masak: exactly
with quotes from notable Perl people (in Perl 5 and Perl 6) in the announcement. 14:37
"RJBS says ... it's about time"
awwaiid "Perl6 joins force with Mayans to end world before Christmas"
jnthn "Don't panic, we don't know which Christmas yet." 14:38
PerlJam "obra reveals that this was really his plan all along when he put forth his proposal for the future of perl 5"
awwaiid++ I like that
frettled: btw, the word you were looking for is "pumpking" 14:39
frettled PerlJam: thanks, I forgot
14:42 GlitchMr joined
JimmyZ masak: I agree with you, though that's always contentious 14:43
PerlJam The trick is getting enough of the right people involved and to make it sound almost plausible such that people actually question the reality of it. 14:44
frettled Also, you need a decoy, an _obvious_ April 1st joke.
14:45 JimmyZ left
frettled I think the Mayan joke suggested by awwaiid works. 14:45
masak :D
14:45 JimmyZ joined
frettled As a bonus, mentioning the P6 hackathon in Oslo in April as a hacking meet to ensure API compatibility or somesuch could be cool. 14:46
PerlJam exactly 14:47
combine enough facts with fiction that people aren't sure if it's a joke
frettled That timing is becoming freakishly good.
frettled really looks forward to that hackathon, BTW, it will be the first Perl meet for me in a looong time. 14:48
14:51 Trashlord left 14:52 Trashlord joined 15:00 benabik joined
PerlJam another good AFJ would be to announce that lwall has decided to rename Perl 6 after much flak from the community. (people would certainly believe that! :) 15:00
masak and they'd lynch us when it turns out to be a joke :P 15:01
JimmyZ rename 6 to 7 :) 15:03
benabik The Perl 6 community has gotten tired of the spec changing, so have kicked out TimToady. Now he's working on Perl 7, which is composed entirely of non-ASCII characters. 15:06
masak "Perl 7 will be my rewrite of the community. (mumble grumble)"
benabik masak++ 15:07
JimmyZ or universe 15:08
geekosaur (as usual, xkcd already covered that) 15:09
15:11 marmalade joined
flussence "Perl 6000" 15:11
cosimo quick raise of hands: who's coming to Oslo? :) 15:21
15:21 JimmyZ left
masak o/ 15:22
so are jnthn, moritz, pmichaud, tadzik, and fsergot.
15:22 JimmyZ joined
JimmyZ firefox--, which always crashed with chatzilla 15:23
cosimo masak: cool!
it's going to be great!
masak yes. yes, it is. 15:26
sjn++
[Coke] sisar: still can't reproduce it: gist.github.com/2001553 15:39
I seem to recover immediately after the confusion, whereas you seem to need one more line. 15:40
moritz maybe there's some platform dependent copy/paste mechanism
thing is, if you paste 5 lines, rakudo should say "Confused" 5 times
[Coke] -1 from me on more april fool's day stuff.
sisar [Coke]: why does it look like that you copy-pasted each line one-by-one ? 15:41
masak maybe it's the terminal window application.
[Coke] sisar;no clue. I did not.
but, in any case: it seems that if you hit return one more time, you're ok.
moritz so if you only get a single "Confused", maybe the line separator is weird or something
sisar fyi, i'm using GNOME terminal
[Coke] I'm ssh'd into feather using putty 15:42
sisar [Coke] you have a local rakudo ?
fwiw, awwaiid above was able to reproduce it 15:43
15:46 grondilu left
awwaiid moritz, sisar, actually mine said confused a bunch of times :) 15:47
15:48 Patterner left 15:49 pernatiy left 15:50 Psyche^ joined, Psyche^ is now known as Patterner, pernatiy joined, att left 15:58 kaare_ joined 15:59 thou joined 16:12 HarryS left, JimmyZ left 16:14 lateau left 16:17 tarch joined 16:31 kaleem_ joined 16:34 kaleem left 16:36 att joined 16:40 GlitchMr left
[Coke] masak++ #autopun pdf 16:45
colomon [Coke]: where is it?
TimToady github.com/masak/gpw-talks/raw/mas...topuns.pdf 16:47
er, stupid lag
.oO(you can say that again)
 16:48
colomon TimToady++
TimToady if I was going to be snide, I'd've said "in the backlog"
all I did was type masak in my browser 16:49
and look at the history
masak maybe it was a mistake to publish that PDF. now I have to find new fresh autopuns for YAPC::EU :) 16:51
maybe I could build an autopun factory.
turn the crank; autopun comes out.
TimToady you need to write an autopun macro 16:52
16:52 mj41 left
TimToady actually, just write an AI that wanders the intertubes finding them; I'm sure you'll get more than you can talk about 16:53
or just teach everyone how to do it, and get it to go viral, and then you don't have to talk about them anymore 16:54
masak it's surprisingly difficult to convey what's an autopun and what isn't. 16:55
not just because it's not 100% clearly delineated, but because the the core "trick" isn't easily put into words.
TimToady indeed, it's almost past the semantic level down in the pragmatic level 16:56
16:56 daxim left
TimToady I keep having this nagging worry that you're gonna think I'm trying to make autopuns when I'm just trying to be funny... :) 16:57
PerlJam masak: got any example autopuns that recurse at least once?
or autopuns that pun in 2 orthogonal directions 16:58
masak TimToady: in a give conversation about autopuns, there always seems to be a general uncertainty about whether one is trying to stick to the topic, or trying to be funny. :)
given*
PerlJam your talk seemed to be about wordy autopuns ... any visual autopuns?
masak PerlJam: interesting notions. no, I haven't seen any examples of those. 16:59
TimToady well "All X are liars" can thought of as a recursive autopun if you're an X
masak I never liked the "All X are liars" as an example of the prototypical paradox. it's not a paradox.
awwaiid pft. YOU'RE not a paradox
masak :) 17:00
geekosaur not by itself, but it's supposed to be shorthand for the syllogism that *is* a paradox when complete
17:00 kaleem_ left
TimToady wonders what a metadox would be 17:00
maybe a metadox is what the all cretans thing is
masak well, now, come on. even in a black-and-white world where everyone either is a liar or a truth-teller, things can be resolved by Epimenides being a liar, and not all Cretans being liars. 17:01
TimToady well, I know how Alexander would have solved it
masak i.e., there has to be at least one Cretan who is not Epimenides and who is not a liar. then Epimenides lies, he's a liar, and no paradox.
TimToady cretan 17:02
masak :P
masak decommutes 17:03
geekosaur *snrk*
TimToady discretely
PerlJam masak: "all X are liars" is more of an idealized theoretical construct. Yours are *real* and thus quite a bit more appealing. :-) 17:04
"all X are liars" makes me think of spherical cows of uniform density.
17:07 MayDaniel joined
TimToady it's difficult to have a visual autopun in the concrete (ouch) sense of visual, since an autopun is defined to have a use/mention, which implies some amount of semiotic indirection 17:08
you need something visual that is a "sign" but that is not language, per se
PerlJam TimToady: difficult doesn't mean impossible. :-) I bet Escher has some visual autopuns 17:09
TimToady a toppled statue of Lenin pointing at the ground comes close
the hands drawing each other can be read that way 17:10
PerlJam ah, exactly 17:11
TimToady I think they'll usually come across as "frame violations" in the visual sphere
it's a pity the painting has to say "this is not a pipe" in language 17:12
PerlJam heh, that's the first thing I thought of when I mentioned "visual" autopuns :-)
17:13 birdwindupbird left
TimToady I suspect Escher's impossible objects can be thought of as autopuns, where the semiotic "frame" lies between 3-D space and 2-D projections of that 3-D space 17:15
and our interpretations of those projections back into 3-D in our minds
something kind of 2-level recursive there 17:16
it's like the automatic translator joke doesn't work unless you translate to Russian and then back to English
PerlJam that's an interesting notion. Also the "location" of the autopun. Where does it happen? Is it an autopun only if we have a shared language? a shared community? a shared geography? does it even require sharing? 17:17
TimToady you need the before/after to put the spotlight on some hidden semiotic process that is not apparent 17:18
17:19 isBEKaml joined
TimToady there's this sense in which an autopun must construct two different realities in our heads that then metaphorically fight somehow 17:19
in the semantic space we tend to call these "paradoxes", but they don't actually rise to the level of autopun unless there are pragmatics involved 17:21
PerlJam An image of a square wheel is a sort of autopun
TimToady indeed, "auto" in Greek just means "he/she/it"
in addition to self
17:22 ggoebel left
TimToady so there has to be a self involved 17:22
all X are liars doesn't feel autopunny because X is too abstract
PerlJam masak: I think you'll have no problems coming up with new autopuns ;)
17:23 HarryS joined 17:26 isBEKaml left 17:27 ggoebel joined, isBEKaml joined 17:29 tarch left 17:33 wolfman2000 left
PerlJam According to wikipedia, the Mayans used puns. awwaiid's suggestion of combining the Mayans with Perl 6 for an AFJ seems even more like a winner :) 17:36
17:37 MayDaniel left
TimToady wonders if "end of the world" sounds like "Perl 6 released" in Mayan... 17:38
17:39 nine joined
nine Is current Rakudo using Parrot's OO stuff (classes and esp. methods) in any way? 17:40
17:41 ggoebel left
jnthn nine: PAST and POST do 17:42
PerlJam nine: why do yo ask?
jnthn But Rakudo itself doesn't.
And PAST is getting replaced once I have the tuits.
I think virtually nothing in the NQP repository uses it either. 17:43
17:46 ggoebel joined
nine So it probably would not hurt if I add code to Parrot_find_method_direct? Just ask because I try hard to not touch any code path which is used in a single threaded program. 17:48
17:50 att left
[Coke] +# 03/08/2012 - rakudo++ ; niecza (96.24%); pugs (35.08%) 17:50
+"niecza", 20306, 1, 757, 1540, 22604, 23759
+"pugs" , 7403, 0, 2887, 850, 11140, 23568
+"rakudo", 21099, 31, 626, 1888, 23478, 24026
jnthn nine: Almost certainly not. 17:51
Though I'm a tad curious what you're adding there. :)
nine jnthn: I use completely separated memory domains for each thread so I have each thread run its own GC. For most parts this was quite simple to achieve, but method calls give me a hard time. I'm currently trying to share VTABLEs and classes between threads and just create a local copy of the method in Parrot_find_method_direct 17:55
jnthn Hm. 17:57
Yeah, if you have no notion of objects shared between threads then...there's not much else for it besides copying.
Guess handling sharing needs a write-barrier on share. 17:58
nine For all other objects I use simple proxies which just forward all VTABLE calls to the proxied object. But calling a sub this way would cause the sub to run in its original interp's context screwing things up. 17:59
18:01 fglock joined, dakkar left 18:03 att joined
eiro fglock, are you concidering giving your astonishing talk at fpw ? 18:05
18:05 thou left
fglock eiro: hi 18:05
eiro hi :)
fglock that's the French Perl Workshop? 18:06
I don't know - looking it up
eiro please! 18:07
fglock 24th and 25th of June - looks good :)
eiro fglock, where have you seen those date ? 18:08
it's actually june 29,30
journeesperl.fr/fpw2012/ 18:09
dalek ecs: efdbdd8 | larry++ | S32-setting-library/Str.pod:
spec fc, remove capitalize (see titlecase)

capitalize was defined in terms of ucfirst, which no longer exists. We also need to support case folding as defined by Unicode.
fglock journeesperl.fr/fpw2011/
eiro fglock, let me check but i think we're in 2012 :)
fglock heh 18:10
eiro fglock, was it the default page for ya
fglock I just got a link from google 18:11
didn't pay much attention to the small detail of dates
perlito5 progress report: 18:12
imported all the CORE prototypes
some prototypes work; some "difficult" parses work now
self-compile time is now 10s in v8
perl5-to-perl6 probably need to reuse the javascript data model 18:13
thinking about a low-level backend (parrot or dot-net maybe) 18:14
there is lots of "low-hanging fruit" accumulating 18:15
eiro fglock, it would be cool to report 18:18
18:23 benabik left 18:24 wtw left
skids ERROR: OK. # short autopun 18:33
18:33 sisar left 18:41 birdwindupbird joined 18:45 am0c joined 18:58 _buno_ joined
TimToady perl6: die "Alive" 19:00
p6eval niecza v15-4-g1f35f89: OUTPUT«Unhandled exception: Alive␤ at /home/p6eval/niecza/lib/CORE.setting line 1362 (die @ 3) ␤ at /tmp/dpIcVRsu2l line 1 (mainline @ 2) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3838 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3839 (module…
..pugs b927740: OUTPUT«*** Alive␤ at /tmp/Cy8yc1UjZf line 1, column 1 - line 2, column 1␤»
..rakudo 4235df: OUTPUT«Alive␤ in block <anon> at /tmp/N6X7hTyGT2:1␤␤»
19:03 _buno_ left 19:07 aindilis left, aindilis joined 19:08 isBEKaml left, isBEKaml joined 19:20 plobsing left 19:22 plobsing joined
ggoebel on the subject of renaming perl6... I don't think that could ever fly without TimToady's blessing and the consent of the folks implementing perl6 19:22
perl6 distributions are at liberty to call themselves whatever
I get the significance of the name change, baggage, etc... 19:24
PerlJam ggoebel: That's why it would make a good AFJ. :) 19:25
ggoebel but can't really accept it either
april fool's joke? 19:26
PerlJam aye
19:27 NamelessTee left
ggoebel AFD is coming up soon 19:27
Akoya oysters are used to produce cultured pearls... Maybe someone can do something with that... 19:29
PerlJam Is it a kind of autopun when someone asks a question that contains the answer within the question? 19:31
(they mention the thing they should use)
skids I just consider those lmgtfy bait. 19:32
"Perl6 renames itself to «insert particularly hard to compose unicode sequence here»" 19:36
masak PerlJam: the only such autopun I know is the one from bash.org: "You should learn how to find things out for yourself." -- "How do I do that?"
19:36 havenn joined
flussence I'm thinking of writing a buggy, incomplete set of perl6 scripts for website templates... 19:37
masak ha, here: bash.org/?3936 :)
flussence I'll call it "Perl 6 Hypertext something something"!
masak flussence: "...Matt's archive" :) 19:38
it'll be an instant success.
en.wikipedia.org/wiki/Matt's_Script_Archive 19:39
flussence ouch... I never knew about that yet I recognise the name "FormMail.pl" 19:40
19:41 Trashlord left
moritz Serialization Error: Unimplemented object type 'CTMThunk' passed to write_ref 19:41
that's what I get when playing with sink
jnthn Urgh 19:42
flussence
.oO( I know! I'll call it Perl 6 HTML Plus. It can be abbreviated to "P6H+", like that HQ9+ language. )
jnthn moritz: Yeah, I guess that "can't happen" situation can happen.
moritz: It's a relatively easy fix.
dalek kudo/sink2: 85c63c5 | moritz++ | / (9 files):
re-apply the interesting parts of the "sink" branch
 19:43
moritz it's surpsing how little change it is alltogether
19:44 Trashlord joined
moritz the sink branch, that is 19:44
19:46 localhost joined
jnthn moritz: nomming now, will fix the serializer after that 19:48
moritz jnthn: ok, no hurry 19:49
19:49 isBEKaml left
moritz tries to remember the diagram we wrote on that blackboard in ER 19:50
20:03 integral left
TimToady wonders why there'd be a blackboard in the emergency room, and how much beer moritz had to drink to end up there... 20:03
20:03 leprevost joined
TimToady that would explain why he can't remember it, though... 20:04
20:09 am0c left
masak TimToady: you've written an editor in Perl 6? 20:10
is it published somewhere? :)
20:11 thou joined
[Coke] phenny: tell au that github.com/perl6/Pugs.hs/issues/6 is another one that might get us a large number of unrelated tests. Any pointers on where to look for the grammar of a sub name? 20:13
phenny [Coke]: I'll pass that on when au is around.
TimToady it's not a text editor
masak ok. 20:14
I guess I assumed it was. I've been thinking of writing one.
TimToady it's a quiz editor for the quizzing my wife coaches, which I scorekeep for at meets 20:15
but it does do single character entry and updates the screen as things change
and works a bit like a specialized spreadsheet, so performance matters 20:16
masak nod.
20:18 havenn left 20:19 snearch joined
[Coke] masak: port padre? 20:20
TimToady possible autopun: twitter.com/#!/BryanRoberts72/stat...28/photo/1 20:21
20:21 am0c joined
masak [Coke]: more like, one warm summer evening I had three distinct ideas for how to extend a vim-like editor to be awesome. 20:22
I only remember two of them now.
TimToady 1st idea, rewrite the whole thing in P6
masak one was to allow for Perl 6 code refactoring by hooking it up with a STD parser somehow.
the other was to scrap the file system and replace it by a git repo, and then follow the logical consequences of that decision. 20:23
that latter idea still excites me sometimes.
20:23 NamelessTee joined
masak imagine seamlessly switching between file edit mode, git blame mode, git log mode, and git branch mode. 20:24
all inside of a vim-like editor.
[Coke] masak: vim++, git++ 20:25
20:25 bacek joined
TimToady track blame on a byte-by-byte basis and you're starting to reimplement Xanadu :) 20:25
masak <skids> ERROR: OK. # short autopun
[Coke] wonders if he can trick timtoady into hacking on pugs. 20:26
masak I'm sorry, I only see a contradiction there. no autopun...
Tene masak: you're familiar with github.com/tpope/vim-fugitive yes?
masak Tene: nope. looking.
TimToady has hacked on pugs before, hence is wary :)
[Coke] TimToady: is github.com/perl6/Pugs.hs/issues/6 something you could point me in a direction on ? 20:27
masak Tene: nice.
TimToady I'm afraid hacking in Haskell will turn me into a genius with no empathy for the common man :)
eiro TimToady, can i quote you ?
TimToady nobody is allowed to quote me ever 20:28
fat load of good that does :)
flussence
.oO( the policy is strictly misquotes only )
Tene en.wikiquote.org/wiki/Larry_Wall ;)
TimToady I bet almost everything there is more than ten years old... 20:29
eiro you said you prefer forgive than forbid
20:29 buubot_backup left
Tene en.wikiquote.org/wiki/Larry_Wall#La....282008.29 20:29
TimToady I said a great many things that I never said, but I'm not as good at it as Yogi Berra... 20:30
Tene eh, you did say "almost"
20:30 integral joined, integral left, integral joined
[Coke] pugs: say 3.sec 20:30
p6eval pugs b927740: OUTPUT«*** No such method in class Int: "&sec"␤ at /tmp/RJ9Aa0mayY line 1, column 5 - line 2, column 1␤»
[Coke] pugs: say sec(3)
p6eval pugs b927740: OUTPUT«*** No such subroutine: "&sec"␤ at /tmp/I8WMRsQeDT line 1, column 5 - line 2, column 1␤»
TimToady I hedge every statement I make. Well, almost every. 20:32
masak :P 20:33
autopuns are funny exactly because the link between use and mention is there, triggers in the listener's brain, and becomes something like a "shared secret" between speaker and listener that both can smile knowingly at. 20:34
TimToady or :P at 20:35
masak actually, much humor seems to involve passing a closure into the listener's brain for them to execute.
TimToady that's what you thunk 20:36
PerlJam masak++ a nice techie analogy :)
masak in the case of autopuns, it's "<mention>. <use>. { now look what I just did, putting those close to each other }"
TimToady autopuns are a bit like quines
flussence wonders if autopuns can be used for effective captchas... 20:37
PerlJam flussence: until humor can be automated :-)
TimToady you think bulgarian grannies are not smart enough to follow them?
PerlJam But then you get the problem of people with no sense of humor
TimToady why is that a problem?
</rimshot> 20:38
well, I suppose it should be <rimshot/> really 20:39
PerlJam masak: Have you heard the song "Say you'll haunt me" by the group Stone Sour?
masak no. finding it.
TimToady
.oO(You're so vain, I'll bet you think this song is about you...)
 20:40
[Coke] ww
[Coke] suggests updating S06-operator-overloading/sub.t 's "correct overloaded method called" test.
PerlJam masak: your comment about a shared secret made me think of that song.
[Coke] rw!
PerlJam masak: "Little souvenirs and secrets shared, Little off guard and unprepared"
flussence
.oO( rw = read-write, ww = write amplification... )
 20:41
TimToady wants :aro = "almost read-only" 20:43
TimToady wonders how that differs from :arw... 20:44
20:44 fsergot joined
PerlJam the former is easier to pronounce :) 20:44
fsergot hi o/ 20:45
masak PerlJam: interesting music video.
fsergot! \o/
[Coke] :raw 20:46
20:50 buubot_backup joined 20:54 wooden_ left
TimToady [Coke]: In regard to helping you with Haskell, either giving you a fish or teaching you how to fish would entail me learning how to fish again first... 20:56
20:56 bluescreen10 left 20:57 fglock left
TimToady and if you try to teach me, we'd have entail recursion... 20:58
masak [Coke]: I'm eager to help. I will not have tuits of that shape until $weekend, though.
21:03 leprevost left 21:04 thou left 21:06 birdwindupbird left
diakopter entrail recursion? 21:21
oh, entail
jnthn o/ diakopter :) 21:23
masak diakopter! \o/ 21:28
21:29 att left 21:36 icwiener joined 21:38 pat_js joined 21:45 havenn joined 21:47 kaare_ left
felher I just wanted to briefly note that there is an incomplete sentence in the specs (starting at S04-control.pod:1057) which i am - sadly - not able to fix. Maybe someone with exceptional perl6 experience can fix it? :) 21:50
21:53 pat_js left 21:54 pat_js joined 21:58 havenn left 21:59 havenn joined 22:03 skids left
masak that one can only be fixed by someone who understands the exception model, I'm afraid ;) 22:03
jnthn I guess I could take a 22:04
masak :P
felher :) 22:06
jnthn Hm. Not immediately obvious how to fix that one. 22:07
masak it's as if the one who wrote that sentence got interrupted by a thought flow escape mechanism of some sort, and didn't handle the interruption properly. 22:12
22:12 am0c left
felher Hihi, indeed :) 22:15
dalek ecs: a0d5a9e | jnthn++ | S04-control.pod:
A couple of fixes to the exceptions spec.
 22:16
masak there should be, like, specific instructions for how to behave in such situations.
jnthn That should do it.
felher \o/ 22:18
jnthn++
eiro thanks masak! (for the talk proposal)
masak you're welcome.
I should probably include some slides about Gödel numbering in my next instance of the autopuns lightning talk. 22:19
felher Since we are currently at exceptions: Maybe one could explain to me why we do "set_outer_caller's_bang(@handled) unless @unhandled;" in the code just above that part? :) 22:20
The outer caller gets all handles exceptions in it's @!, if there aren't any unhandled exceptions?
jnthn felher: Correct
felher: It's an optimization really
felher: And may well be a false one 22:21
The reasoning is that if there are unhandled exceptions, we will unwind the stack beyond this point, so the @! we set in that statement could never be read.
22:21 snearch left, alester left
jnthn Though in today's Rakudo, storing a lexical is probably cheaper than the condition check. :) 22:22
felher jnthn: ah, okay. I think i got it. :) My last question would be: What exactly is the outer-caller: Can i do the following: "sub-that-handles-all-exceptions-in-a-catch-block(foo); for @! { ... #check what exceptions were handled by sub-that...}" 22:28
jnthn: is that the outer-caller that gets its @! set? 22:29
jnthn hm 22:31
ah
nom: try { die "foo" }; say $!
p6eval rakudo 4235df: OUTPUT«foo␤ in block <anon> at /tmp/2D0QePVnpc:1␤␤»
jnthn nom: try { die "foo" }; say $!.message 22:32
p6eval rakudo 4235df: OUTPUT«foo␤»
jnthn I thought it was something to do with the setting of the $! there
nom: { die "foo"; CATCH { default { } } }; say $! 22:33
p6eval rakudo 4235df: OUTPUT«foo␤ in block <anon> at /tmp/ldKu89LpyM:1␤␤»
jnthn nom: { die "foo"; CATCH { default { } } }; say $!.message
p6eval rakudo 4235df: OUTPUT«foo␤»
jnthn Like that with an explicit CATCH
22:35 aindilis left 22:36 aindilis joined, djanatyn left
felher nom: sub exc() { die "blar"; CATCH { default { } } }; exc(); say $!.message; 22:37
p6eval rakudo 4235df: OUTPUT«Method 'message' not found for invocant of class 'Any'␤ in block <anon> at /tmp/9aG6hyJvNe:1␤␤»
22:37 Tene left 22:38 thelazydeveloper joined
felher In this case, isn't the 'say $!.message' part the outer-caller? 22:39
jnthn I'm mostly confused by the term "outer caller" there 22:40
It feels more like outer-outer or some such
"the place we're going to unwind to" 22:41
felher Would the "say $!.message"-part be the place we're going to unwind to? 22:42
jnthn Yes
22:42 djanatyn joined
felher jnthn: okay, great :) 22:43
jnthn: so, thanks for your patience and explanations :) jnthn++
jnthn nom: sub foo() { die "lol"; CATCH { default { } } }; foo(); say $!.message
p6eval rakudo 4235df: OUTPUT«Method 'message' not found for invocant of class 'Any'␤ in block <anon> at /tmp/H_cisZ8Bx7:1␤␤»
22:43 Tene joined, Tene left, Tene joined
jnthn perl6: sub foo() { die "lol"; CATCH { default { } } }; foo(); say $!.message 22:44
p6eval rakudo 4235df: OUTPUT«Method 'message' not found for invocant of class 'Any'␤ in block <anon> at /tmp/rlLn1XYcCM:1␤␤»
..pugs b927740: OUTPUT«*** lol␤ at /tmp/iU1tM_vq1W line 1, column 49-54␤ /tmp/iU1tM_vq1W line 1, column 13-22␤»
..niecza v15-4-g1f35f89: OUTPUT«Unhandled exception: Unable to resolve method message in class Any␤ at /tmp/FqhAvC_nrh line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3838 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3839 (module-CORE @ 65) ␤ at /home…
jnthn If it really meant caller I'd kinda expect that to work.
felher For a moment i thought it works in pugs. But pugs just doesn't catch the exception, i think. 22:49
22:50 mtk left 22:53 aindilis left, aindilis joined
dalek p: f2d0311 | jnthn++ | src/6model/serialization.c:
Avoid CTMThunk write_ref error reported by moritz++.
 22:54
jnthn phenny: tell moritz fixed the issue that sunk your progress earlier
phenny jnthn: I'll pass that on when moritz is around.
23:10 marmalade left 23:12 marmalade joined
masak 'night, #perl6 23:13
felher o/ 23:17
23:19 PacoAir left 23:21 havenn left 23:30 icwiener left 23:35 thelazydeveloper left 23:39 sivoais left 23:51 Trashlord left 23:58 Timbus left, NamelessTee left, Timbus 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!