Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2012-07-27

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:09 tokuhiro_ joined #perl6
01:04 skids r: class C { }; role F [ ::T $t ] { method m { $t.WHAT.say } } ; class D does F[ BEGIN { C.new() } ] { }; my D $d .= new(); $d.m;
01:04 p6eval rakudo f23a45: OUTPUT«C()␤»
01:04 skids r: class C { }; role F [ ::T $t ] { method m { $t.WHAT.say } } ; class D does F[ C.new() ] { }; my D $d .= new(); $d.m;
01:05 p6eval rakudo f23a45: OUTPUT«C()␤»
01:05 skids Hmmm... someone fixed something since the June Star. :-)
01:06 * skids is amazed how well implemented the whole parametric role stuff is already, BTW
01:33 colomon joined #perl6
01:47 colomon o/
01:47 TimToady \o
01:49 TimToady welcome to the graveyard shift, when it's daytime over Guam, and nearby hotbeds of programming activity
01:51 skids r: class C { }; role F [ ::T:U $t ] { method m { ":U" } }; role F [ ::T:D $t ] { method m { ":D" } } ; class D does F[ C ] { };
01:51 p6eval rakudo f23a45: OUTPUT«===SORRY!===␤None of the parametric role variants for 'F' matched the arguments supplied.␤Ambiguous call to ''; these signatures all match:␤:(Mu , Mu $t)␤:(Mu , Mu $t)␤␤»
01:52 * skids wonders if ::T:U is specced.
02:05 orafu joined #perl6
02:05 TimToady std: sub foo (Mu:U ::T $) { say T.WHAT }; foo(Int)
02:05 p6eval std e52e3ca: OUTPUT«ok 00:00 43m␤»
02:05 TimToady rn: sub foo (Mu:U ::T $) { say T.WHAT }; foo(Int)
02:05 p6eval niecza v19-15-g051783d: OUTPUT«[31m===[0mSORRY![31m===[0m��GLOBAL::T does not name any package at /tmp/Hh7UhD4z4m line 1:�------> [32msub foo (Mu:U ::T[33m�[31m $) { say T.WHAT }; foo(Int)[0m��A type must be provided at /tmp/Hh7UhD4z4m line 1:�------> [32msub foo (Mu:U ::T[33m…
02:05 p6eval ..rakudo f23a45: OUTPUT«Int()␤»
02:06 TimToady rakudo++
02:06 * skids had just figured out Mu:U ::T :-)
02:06 TimToady r: sub foo (Any:U ::T $) { say T.WHAT }; foo(Mu)
02:06 p6eval rakudo f23a45: OUTPUT«Nominal type check failed for parameter ''; expected Any but got Mu instead␤  in sub foo at /tmp/SV_pAXJuTX:1␤  in block  at /tmp/SV_pAXJuTX:1␤␤»
02:07 skids I wonder what rakudo is doing with that ":U" in the "::T:U"
02:07 TimToady r: sub foo (Any:U ::T $) { say T.WHAT }; foo(42)
02:07 p6eval rakudo f23a45: OUTPUT«Parameter '' requires a type object, but an object instance was passed␤  in sub foo at /tmp/EK6N3C0Ugw:1␤  in block  at /tmp/EK6N3C0Ugw:1␤␤»
02:07 TimToady r: sub foo (Any ::T:U $) { say T.WHAT }; foo(42)
02:07 p6eval rakudo f23a45: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&T' called (line 1)␤»
02:07 TimToady r: sub foo (Any ::T:U $) { say T:U.WHAT }; foo(42)
02:07 p6eval rakudo f23a45: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&T' called (line 1)␤»
02:08 TimToady beats me
02:14 Coleoid joined #perl6
02:25 quietfanatic isn't WHAT a noop on a type object?
02:29 skids r: class C { }; C.WHAT.say; C:U.WHAT.say;
02:29 p6eval rakudo f23a45: OUTPUT«C()␤C()␤»
02:32 TimToady quietfanatic: sure, I was just...I was just...trying to clarify the intent...yeah, that's the ticket!
02:33 skids Well, one could nitpick at the meaning of "noop". :-)
02:36 woosley joined #perl6
02:40 xinming joined #perl6
02:42 Psyche^ joined #perl6
02:52 xinming joined #perl6
02:57 xinming joined #perl6
03:14 TimToady according to S04:1570, http://irclog.perlgeek.de/​perl6/2012-07-14#i_5814213 should already be considered a block and not a hash; you should not have to put a semicolon
03:14 scott__ joined #perl6
03:15 TimToady if you have to put semicolon, it's a bug
03:17 TimToady perl6: say (map { .key => uc .value }, :foo<bar>).perl
03:17 p6eval niecza v19-15-g051783d: OUTPUT«Unhandled exception: Unable to resolve method key in type Any␤  at /tmp/eO1LCKGz_7 line 1 (mainline @ 6) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3929 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3930 (module-CORE @ 564) ␤  at /home/p6…
03:17 p6eval ..rakudo f23a45: OUTPUT«No such method 'key' for invocant of type 'Any'␤  in block  at /tmp/BdUGNYqcsc:1␤␤»
03:17 TimToady perl6: say (map { .key => uc .value }, (:foo<bar>)).perl
03:17 p6eval niecza v19-15-g051783d: OUTPUT«Unhandled exception: Unable to resolve method key in type Any␤  at /tmp/j41T32p_Es line 1 (mainline @ 6) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3929 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3930 (module-CORE @ 564) ␤  at /home/p6…
03:17 p6eval ..rakudo f23a45: OUTPUT«No such method 'key' for invocant of type 'Any'␤  in block  at /tmp/K2BtfRL2RK:1␤␤»
03:18 TimToady perl6: say (map { $_.key => uc $_.value }, (:foo<bar>)).perl
03:18 p6eval niecza v19-15-g051783d: OUTPUT«Unhandled exception: Unable to resolve method key in type Any␤  at /tmp/HsvK6UJhzw line 1 (mainline @ 6) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3929 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3930 (module-CORE @ 564) ␤  at /home/p6…
03:18 p6eval ..rakudo f23a45: OUTPUT«No such method 'key' for invocant of type 'Any'␤  in block  at /tmp/v_qL0RGVKW:1␤␤»
03:19 TimToady perl6: say (map { $^arg.key => uc $^arg.value }, (:foo<bar>)).perl
03:19 p6eval niecza v19-15-g051783d: OUTPUT«("foo" => "BAR", ).list␤»
03:19 p6eval ..rakudo f23a45: OUTPUT«("foo" => "BAR",).list␤»
03:19 TimToady $_.key and .key should do whatever $^arg does
03:19 benabik joined #perl6
03:31 fhelmberger joined #perl6
03:36 mucker joined #perl6
03:47 crab2313 joined #perl6
04:37 kaare_ joined #perl6
04:57 zhutingting joined #perl6
05:04 nodmonkey joined #perl6
05:04 birdwindupbird joined #perl6
05:41 muixirt joined #perl6
05:48 thou joined #perl6
06:17 wtw joined #perl6
06:32 kaleem joined #perl6
06:37 SamuraiJack joined #perl6
06:49 SamuraiJack_ joined #perl6
06:54 SamuraiJack_ joined #perl6
06:57 dalek star: 041912e | pmichaud++ | Makefile:
06:57 dalek star: Put a VERSION file into the star tarball; we can use it during
06:57 dalek star: building.
06:57 dalek star: review: https://github.com/rakudo/star/commit/041912eb0a
06:57 dalek star: fa9799f | pmichaud++ | skel/tools/build/Makefile.in:
06:57 dalek star: Some targets to help build MSI files for Windows (hopefully).
06:57 dalek star: review: https://github.com/rakudo/star/commit/fa9799fd50
06:57 dalek star: 6066c44 | pmichaud++ | skel/docs/announce/2012.07:
06:57 dalek star: Merge branch 'master' of github.com:rakudo/star
06:57 dalek star: review: https://github.com/rakudo/star/commit/6066c440ae
06:59 dalek star: 80c00b5 | pmichaud++ | skel/tools/build/ (2 files):
06:59 dalek star: Add script to patch wxs file generated by Wix "heat" command.
06:59 dalek star: review: https://github.com/rakudo/star/commit/80c00b5913
07:02 TimToady http://rosettacode.org/wiki​/Averages/Mean_angle#Perl_6
07:07 hoelzro masak: ping
07:14 masak hoelzro: pong.
07:14 masak morning, #perl6
07:14 hoelzro masak: thanks for providing that tarball; it ran my crash script without crashing =)
07:15 masak \o/
07:18 hoelzro I also managed to build my own Rakudo without things crashing
07:18 hoelzro interestingly enough, the crash only seems to happen when I install Rakudo and friends to /usr/
07:18 hoelzro I'll be testing more after work
07:20 cognominal for people doing coffeescript, beware :  console.log "toto" if false for i in [1] is interpreted as   console.log "toto" if (false for i in [1])
07:21 cognominal which prints "toto" . Run with coffee -pe to see the disaster.
07:23 BlueT_ joined #perl6
07:23 felher TimToady++ #mean angle on RC
07:28 dalek star: 68b7211 | pmichaud++ | skel/tools/build/Makefile.in:
07:28 dalek star: I meant to ignore an error result, not suppress the command display.  Now fixed.
07:28 dalek star: review: https://github.com/rakudo/star/commit/68b7211b50
07:31 masak cognominal: that sounds like a misdesign, indeed.
07:39 jnthn morning, #perl6
07:39 hoelzro ahoy jnthn
07:41 jnthn hoelzro: dobry den :)
07:41 jnthn Oooh, pmichaud++ is doing MSI stuff \o/
07:58 dalek roast: 35ecf2e | moritz++ | S03-operators/inplace.t:
07:58 dalek roast: remove wrongly ported test (lcfirst -> tc), rakudo-fudge for missing tc
07:58 dalek roast: review: https://github.com/perl6/roast/commit/35ecf2e879
07:59 dayangkun joined #perl6
08:03 moritz we're getting several spectest fails from the transition to tc, which we don't implement :(
08:07 pmichaud I've created a candidate R* tarball if anyone wants to test it:  http://pmichaud.com/sandbox/​rakudo-star-2012.07b.tar.gz
08:07 pmichaud I've also created a .msi version, at http://pmichaud.com/sandbo​x/rakudo-star-2012.07b.msi
08:07 moritz pmichaud++
08:08 pmichaud comments and suggestions welcomed.  I'm off to bed; will check backscroll when I awaken and then produce the final versions and cut the release (unless there are any showstoppers)
08:09 pmichaud afk  # sleep
08:11 dalek roast: 2db35ca | moritz++ | S (2 files):
08:11 dalek roast: rakudo fudges
08:11 dalek roast: review: https://github.com/perl6/roast/commit/2db35ca1ef
08:13 nodmonkey joined #perl6
08:13 BlueT_ joined #perl6
08:14 dalek roast: ea3913c | moritz++ | S0 (2 files):
08:14 dalek roast: more rakudo fudging, move unrelated tests to use uc instead of tc
08:14 dalek roast: review: https://github.com/perl6/roast/commit/ea3913cc5c
08:15 masak what would the implementation of tc look like? is it expressible in Perl 6?
08:16 moritz it's expressible in Perl 6 if you have access to a Unicode database with casefolding information
08:18 dalek roast: 5e1cfcf | moritz++ | S0 (3 files):
08:18 dalek roast: rakudo unfudges
08:18 dalek roast: review: https://github.com/perl6/roast/commit/5e1cfcfad7
08:20 masak ah, right.
08:20 dalek roast: c04506f | moritz++ | S32-str/ (2 files):
08:20 dalek roast: another rakudo fudge, remove a lcfirst test
08:20 dalek roast: review: https://github.com/perl6/roast/commit/c04506f93d
08:20 masak so, probably better to do in C or something.
08:20 moritz afaict ICU alread has what you need
08:20 moritz it just needs wiring up
08:22 masak oh, of course. nice.
08:37 dbr joined #perl6
08:42 felher pmichaud++ : i tested the tarball and i didn't find any problems at a first glance. If there is anything specific i should test, let me know :)
09:05 moritz felher: try  perl6 -MLWP::Simple -e 'say LWP::Simple.get("http://www.perl6.org");'
09:06 felher *trying*
09:08 moritz pmichaud: https://gist.github.com/3186956 that's the first "Missing or wrong version of dependency 'src/gen/CORE.setting'", I get several more during the module building
09:11 felher ./rakudo-star-2012.07/perl6 -MLWP::Simple -e 'say LWP::Simple.get("http://www.perl6.org");' gives me ===SORRY!=== No STable at index 7
09:11 felher :/
09:15 moritz pmichaud: module-install.pl tries to compile URI.pm before URI::DefaultPort
09:18 bbkr joined #perl6
09:18 moritz erm no, that's not quite accurate
09:21 moritz no, it dies while compiling lib/IETF/RFC_Grammar/URI.pm
09:21 sergot joined #perl6
09:21 sergot hi o/
09:23 masak sergocie! \o/
09:24 sergot masaku \o/
09:24 sergot :)
09:24 moritz maybe it's because I have another perl6 in $PATH
09:28 Blih joined #perl6
09:32 seldon joined #perl6
09:33 felher I took a look at the output of 'make install' and found five messages like "Quantifier quantifies nothing at line 7, near " [ \\, ]  ]"" (in JSON::Tiny, Panda, JSON::RPC::{Client,Server}) and four like "No object at index 265" (all in Bailador). All of them while trying to compile .pm into .pir.
09:34 moritz hm
09:34 bbkr hm
09:34 seldon_ joined #perl6
09:34 moritz the "Quantifier quantifies nothing" was usually caused by the regex syntax change regarding ** and %
09:34 moritz so that would be very *very* outdated version shipped there
09:35 moritz or it could be an unrelated problem.
09:37 UncleFester6 joined #perl6
09:38 bbkr JSON::RPC does not use ** quantifier
09:39 UncleFester6 felher: I also got the no stable at index 7 error for LWP::Simple.  I got around it by just recompling the LWP::Simple module itself ...
09:39 * moritz does a complete new build and logs it all
09:40 UncleFester6 LWP::Simple depends on URI but gets compiled first ...
09:41 felher UncleFester6: k, thnx :)
09:41 moritz http://moritz.faui2k3.org/tmp/build.log that's the build log
09:41 UncleFester6 felher: does that mean recompile of LWP::Simple worked for you?
09:43 dalek star: 2640d6c | moritz++ | skel/tools/build/Makefile.in:
09:43 dalek star: [build] build URI before LWP::Simple
09:43 dalek star: review: https://github.com/rakudo/star/commit/2640d6c88c
09:44 moritz ok, I'm changing the order in my local working dir (the one I untar'ed from pmichaud++'s release candidate), and trying again now
09:45 felher UncleFester6: yes
09:46 UncleFester6 good news for now ...
09:47 moritz ah, I now see how we managed to introduce URI and LWP::Simple in the wrong order
09:47 felher moritz: how so?
09:47 moritz when LWP::Simple started to depend on URI, I added both (in the correct order) to the end of the module list
09:47 moritz but, LWP::Simple was already in there, further up
09:47 moritz and bbkr++ noticed, and removed the second occurence
09:48 felher :)
09:48 moritz and back in the days, compilate order didn't matter much, so nobody noticed
09:49 UncleFester6 I am notoriously bad about finding tickets in RT.  Is there a ticket about this compile in order stuff?  I think it's a bug.
09:49 tadzik no, that's intentional
09:49 tadzik it's just hard to get right and prone to bugs
09:50 tadzik if you precompile stuff in wrong order then it won't work
09:50 UncleFester6 why not?
09:50 dalek star: 3f28926 | moritz++ | skel/tools/build/Makefile.in:
09:50 dalek star: [build] add a comment about preserving order of module list
09:50 dalek star: review: https://github.com/rakudo/star/commit/3f2892689e
09:50 moritz UncleFester6: consider   use A; class B does A { };
09:50 moritz UncleFester6: this does a lookup by name for role A
09:50 moritz UncleFester6: and then applies it at compile-time into class B
09:51 moritz UncleFester6: so the generated bytecode depends on the exact form of A
09:51 daxim joined #perl6
09:51 moritz UncleFester6: which means that if you change a single bit, and then precompile A, things are bound to go BOOM
09:52 UncleFester6 I understand what compile means - reviewing the rest ...
09:52 moritz UncleFester6: which is why it makes sense to only precompile B if A is also precompiled, and the precompiled form of A is available
09:52 moritz afk, bbi20
09:52 felher Oh, the JSON::Tiny was only a local problem
09:53 felher due to some old JSON::Tiny in ~/.perl6/lib... sorry for the noise :/
10:01 zhutingting joined #perl6
10:06 UncleFester6 phenny: tell moritz thanks for his explanation and when I can explain the order independent module compiles more cogently I'll bring it up again.  I actually code some parrot/pir etc.
10:06 phenny UncleFester6: I'll pass that on when moritz is around.
10:08 jnthn UncleFester6: I suggest studying the bounded serialization stuff a bit also.
10:08 jnthn That will explain quite a bit :)
10:09 UncleFester6 thx
10:29 fhelmberger joined #perl6
10:37 moritz jnthn: nonetheless a better error message would be *very* helpful
10:37 phenny moritz: 10:06Z <UncleFester6> tell moritz thanks for his explanation and when I can explain the order independent module compiles more cogently I'll bring it up again.  I actually code some parrot/pir etc.
10:38 moritz r: say @*INC
10:38 p6eval rakudo f23a45: OUTPUT«/home/p6eval/.perl6/lib /home/p6eval/nom-inst1/lib/parrot/​4.5.0-devel/languages/perl6/lib␤»
10:39 moritz ah, that might be a problem
10:39 moritz since the modules in $HOME come before those from the installation location, modules in ~/.perl6/lib can cause trouble
10:39 tadzik indeed
10:40 moritz I have the feeleing that whichever way around you put the entries in @*INC, there's always a reason to do it differently
10:41 daxim tell masak tag, you're it
10:42 tadzik why do we have /home/ stuff in @*INC by default anyway?
10:42 tadzik What was the reason to do it differently from Perl 5?
10:42 tadzik it makes user's life easier in some cases, but causes breakages in others
10:43 moritz tadzik: so that pand^Wproto had a location to install to
10:44 tadzik right
10:44 moritz this is something we should discuss in Perl or Frankfurt
10:45 tadzik in Perl 5 currently (correct me if I'm wrong) cpan(m) tries to install stuff to the Perl installation directory, and if it can't, it then goes for $HOME/perl5 I think
10:45 tadzik agreed
10:45 moritz hopefully with help from some p5 folks
10:45 tadzik if we do this, we could just make panda warn loudly if it installs to a place which is not in @*INC
10:50 masak daxim: hello -- so, I'm it, yes?
10:51 masak I just wanted to say that jnthn and I are doing a ($work) hackathon today involving CQRS. we'll be hanging out in #cqrs-perl6 if anyone wants to talk shop. :)
10:52 jnthn I...I will?
10:53 masak oh, sorry; should've told you ;)
10:53 jnthn :P
10:53 masak the channel is a bit desolate right now. just you and me and a log bot. :)
10:53 jnthn .oO( At least the Nirvana songs aren't piped into the chanel :P )
10:54 jnthn .oO( METALICA!!! :P )
10:55 muixirt cqrs?
10:59 cognominal ...and the usual lurkers
10:59 moritz jnthn: ok, this is really strange. I recompiled star with no modules in ~/.perl6/lib, and the only modules that fail are Bailador modules. So now I'm loading LWP::Simple, and still get "No STable at index 7"
11:00 masak muixirt: CQRS is a kind of separation of concerns. separating commands (writes) from queries (reads) in an application.
11:00 moritz jnthn: and the weird thing is, I run it with RAKUDO_MODULE_DEBUG=1 and *all* of the loaded modules come from the right paths, and they are all .pir (except for BOOTSTRAP.pbc)
11:01 masak muixirt: it's a fascinatingly applicable technique, and one that seems to lead straight into being able to reason better about the domain.
11:06 zhutingting perl6: sub postfix:<!> { [*] 1..$^n }; say 5!;
11:06 p6eval rakudo f23a45, niecza v19-15-g051783d: OUTPUT«120␤»
11:13 jnthn moritz: That's...odd, but in general it's boiled down to compilation order in this cases before.
11:15 zhutingting What's mean?//Larry's First Law of Language Redesign: Everyone wants the colon.
11:16 Timbus you would probably feel bad without a colon
11:17 UncleFester6 moritz: about Bailador - Bailador depends on Template::Mojo which doesn't seem to be on the module list
11:18 Timbus zhutingting, it means everyone wanted to use the colon for something
11:19 zhutingting :)
11:23 nodmonkey joined #perl6
11:25 JimmyZ joined #perl6
11:27 moritz UncleFester6: oh, that explains a lot
11:30 dalek star: a4ab725 | moritz++ | / (3 files):
11:30 dalek star: Add Template::Mojo
11:30 dalek star:
11:30 dalek star: Bailador now depends on it. UncleFester6++
11:30 dalek star: review: https://github.com/rakudo/star/commit/a4ab725f4d
11:32 felher moritz: works here :)
11:33 moritz felher: what works?
11:33 felher moritz: adding Template-Mojo fixes the problem with Bailador
11:34 dayangkun joined #perl6
11:35 moritz felher: great
11:36 * JimmyZ couldn't install rakudo-star.msi
11:37 moritz JimmyZ: it probably helps if you're more verbose about the problems you encountered
11:39 JimmyZ I double click it and it does nothing
11:42 moritz oh my
11:43 moritz I changed Makefile in my star copy
11:43 moritz but I didn't change Makefile.in
11:43 moritz and then I reconfigured
11:43 * moritz facepalms
11:46 MayDaniel joined #perl6
11:48 felher moritz: i guess you are talking about local changes? Your patch just works fine. I checked it out and rebuild everything. No problems there.
11:48 moritz felher: and can you actually load LWP::Simple?
11:48 moritz perl6 -MLWP::Simple -e 1
11:49 felher ./perl6 -MLWP::Simple -e 1 && echo hi
11:49 felher gives me "hi" :)
11:49 moritz \o/
11:49 UncleFester6 jimmyz: I installed the .msi OK - did you get the little popup dialog box for installing rakudo
11:51 JimmyZ I got a process bar and then disappeared on windows xp
11:51 TBA joined #perl6
11:52 TBA hi all
11:52 tadzik hey TBA
11:52 TBA r: for (1, 2, 3)[1..*] -> $i { say $i; }
11:52 p6eval rakudo f23a45: OUTPUT«2␤3␤»
11:53 TBA that fails on Rakudo 2011.07, parrot 3.6.0 - is it safe to assume thats an old bug thats fixed?
11:53 tadzik fails in what way?
11:53 TBA eventually times out with "Command terminated"
11:53 tadzik hmm
11:53 tadzik it'd appear so
11:53 tadzik well, 2011.07 is... quite old ;)
11:53 tadzik b: for (1, 2, 3)[1..*] -> $i { say $i; }
11:53 p6eval b 922500: OUTPUT«(timeout)»
11:53 TBA ubuntu package sadly, had some issues getting the latest releases to build but in work so didn't have time to play lol
11:54 tadzik TBA: yeah, looks like you're correct
11:54 TBA will install the latest version of rakudo and try again then, thanks :)
11:54 tadzik have fun :)
11:54 TBA always do with perl6! discovering how insanely simple some stuff is!
11:55 tadzik glad you like it:)
11:55 moritz now LWP::Simple works for me too \o/
12:03 JimmyZ \o/
12:08 TBA having some issues compiling latest rakudo from git?
12:08 TBA make: *** [src/stage2/NQPCORE.setting.pbc] Error 1 Command failed (status 512): make Command failed (status 512): /usr/bin/perl Configure.pl --with-parrot=/root/rakudo/install/bin/parrot --make-install
12:08 erkan joined #perl6
12:08 erkan joined #perl6
12:09 TBA got 1gb ram (its a vm)
12:11 moritz needs about 1.5G :(
12:12 UncleFester6 TBA: see also https://rt.perl.org/rt3/Tic​ket/Display.html?id=113660
12:14 TBA np, remembered having the prob before which is when I upped to 1gb, got enough physical memory left so will up it to 1.5g, thanks
12:15 jnthn Note that Rakudo HEAD from numbers I've seen currently needs 300MB or so less (on 64-bit) for compiling CORE.setting than the previous compiler release.
12:16 jnthn I'm hoping to pull the memory requirements down a bit more in the next week or so too.
12:17 am0c joined #perl6
12:22 Woodi hi today pppls :)
12:23 Woodi R* tarball not yet available on R. page ?
12:23 tokuhiro_ joined #perl6
12:23 masak hi Woody! \o
12:24 Woodi hallo masak :)
12:24 masak i*
12:25 Woodi I was just thinking v6 is hard to learn but baby-perl chat little helped my hope :)
12:25 jnthn Woodi: Should be up later today.
12:26 Woodi would be nice to have book called "Advanced Perl 6: zig-zags and beyound" :)
12:26 Woodi jnthn: thanx :)
12:28 JimmyZ 600MB here for compiling CORE.setting
12:28 JimmyZ now
12:29 jnthn JimmyZ: Can you remember what it was before?
12:29 nodmonkey joined #perl6
12:30 JimmyZ 1GB before qast, iirc
12:30 jnthn Ah. So, an improvement. :)
12:30 JimmyZ before bs was 1.5GB
12:31 JimmyZ or before 6model ? I can't remember ...
12:31 Woodi big improvement :) for example OpenBSD have hardcoded kernel limit for proces memory usage at 1GB seence many years
12:39 nodmonkey joined #perl6
12:51 woosley left #perl6
12:56 tadzik JimmyZ: 64bit?
12:56 JimmyZ 32bit
12:57 jnthn Those numbers must be 32... :)
13:01 TBA built latest rakudo and the range to whatever thing now works :)
13:01 TBA and now to do some real work!!
13:02 TBA left #perl6
13:06 tadzik I wonder if the real work is done in Perl 6 :)
13:12 skids joined #perl6
13:14 Woodi I tell you all, monitor glass without years old collection of fingerprints is worth a small walk for cleaninng fluid :)
13:15 drbean joined #perl6
13:25 bluescreen10 joined #perl6
13:26 tadzik 'p' is not recognized as an internal or external command,
13:26 tadzik operable program or batch file.
13:26 tadzik where on earth does this come from
13:28 atrodo joined #perl6
13:54 fhelmberger joined #perl6
13:54 zhutingting joined #perl6
13:54 masak tadzik: Windows's command shell.
13:56 tadzik yeah, but 'p'? :F
13:56 tadzik I need a shell accound on a windows machine somewhere
13:59 PacoAir joined #perl6
14:06 mucker joined #perl6
14:10 moritz pmichaud: ok, more star build trouble
14:11 moritz pmichaud: I ran perl Configure.pl --prefix=/opt/rakudo-star-2012.07 --gen-parrot --gen-nqp  && make install
14:11 moritz pmichaud: and it built and installed nqp just fine in the right location
14:11 moritz except that the "make install" part failed
14:12 moritz nopaste forthcoming
14:13 moritz pmichaud: https://gist.github.com/3188234
14:17 moritz when I  cd rakudo-2012.07 && perl Configure.pl --with-nqp=/opt/rakudo-star-2012.07/bin/nqp   it seems to build rakudo
14:30 TBA joined #perl6
14:30 TBA hi again :)
14:31 lateau joined #perl6
14:31 TBA quick question! if I do this the match works:
14:31 TBA r: my @arr = <aAa bBb cCc>; for @arr.grep(/B/) -> $l { say $l; }
14:31 p6eval rakudo f23a45: OUTPUT«bBb␤»
14:31 TBA but I want to match against %*ENV<USER>
14:32 TBA how can I interpolate that in the grep parameter? i.e. i want to do @arr.grep(/%*ENV<USER>/)
14:32 TBA but instead i get "Unrecognized regex metacharacter %"
14:36 kaleem joined #perl6
14:36 PerlJam TBA: @arr.grep(/<{%*ENV<hi>}>/) should work
14:37 PerlJam TBA: there are probably other ways but my brain isn't cooperating with my right now.
14:37 PerlJam s/my/me/
14:37 lateau left #perl6
14:37 jnthn That...probably won't work.
14:37 TBA it doesnt :)
14:37 TBA malformed regex
14:37 jnthn I think TBA wants to interpolate the current user into the regex.
14:37 TBA yes, basically
14:38 TBA i could always assign to another var, but it would be easier to read if I can interpolate directly
14:38 PerlJam TBA: you can always assign to a scalar
14:38 jnthn r: /%*ENV{'USER'}/
14:38 p6eval rakudo f23a45: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter % (must be quoted to match literally) at line 2, near "*ENV{'USER"␤»
14:38 jnthn r: /%ENV{'USER'}/
14:38 p6eval rakudo f23a45: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter % (must be quoted to match literally) at line 2, near "ENV{'USER'"␤»
14:38 jnthn r: /$(%*ENV{'USER'})/
14:38 p6eval rakudo f23a45: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter % (must be quoted to match literally) at line 2, near "*ENV{'USER"␤»
14:38 jnthn Hmm
14:38 jnthn I guess the interpolation rule just expects scalars.
14:38 [Coke] I dispute the claim made in backscroll that I am a real person.
14:39 jnthn std: /%ENV{'USER'}/
14:39 p6eval std e52e3ca: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Variable %ENV is not predeclared at /tmp/b9CSzlqF9s line 1:â�¤------> [32m/[33mâ��[31m%ENV{'USER'}/[0mâ�¤Potential difficulties:â�¤  Apparent subscript will be treated as regex at /tmp/b9CSzlqF9s line 1:â�¤------> [32m/%ENV[33mâ��[31m{'USER'…
14:39 PerlJam r: /<{ %*ENV<USER> }>/  # what does it do?
14:39 p6eval rakudo f23a45:  ( no output )
14:39 jnthn std: /%*ENV{'USER'}/
14:39 p6eval std e52e3ca: OUTPUT«Potential difficulties:â�¤  Apparent subscript will be treated as regex at /tmp/KFMdQ07zk_ line 1:â�¤------> [32m/%*ENV[33mâ��[31m{'USER'}/[0mâ�¤ok 00:00 42mâ�¤Â»
14:39 PerlJam r: say %*ENV<USER>
14:39 p6eval rakudo f23a45: OUTPUT«p6eval␤»
14:39 jnthn Yeah, it doesn't parse that way even in std
14:39 jnthn PerlJam: <{ }> is just an assertion
14:40 jnthn TBA: Anyway, the easy way to do it is to put the user into a scalar first.
14:40 jnthn my $user = %*ENV<USER>; ... /$user/
14:40 TBA jnthn: thanks, will go with that
14:41 jnthn r: say %*ENV<USER>
14:41 p6eval rakudo f23a45: OUTPUT«p6eval␤»
14:41 TBA but, should interpolation of a hash element work in a regex? s05 mentions about hashes, but not specific keys of a hash
14:41 PerlJam r: "blahp6evalblah" ~~ / <{ %*ENV<USER> }> /;
14:42 p6eval rakudo f23a45:  ( no output )
14:42 PerlJam r: say "blahp6evalblah" ~~ / <{ %*ENV<USER> }> /;
14:42 p6eval rakudo f23a45: OUTPUT«q[p6eval]␤␤»
14:42 jnthn TBA: So far as I can tell, it's not specified as working.
14:42 jnthn Oh...
14:42 TBA now that looks good
14:42 jnthn Oh! I was thinking <?{ ... }>
14:42 PerlJam r: .say for <fred barny p6eval wilma>.grep(/ <{ %*ENV<USER> }> /);
14:42 p6eval rakudo f23a45: OUTPUT«p6eval␤»
14:43 jnthn PerlJam++
14:43 TBA r: .say for <fred barny p6eval wilma>.grep(/ <{%*ENV<USER>}> /);
14:43 p6eval rakudo f23a45: OUTPUT«p6eval␤»
14:43 TBA r: .say for <fred barny p6eval wilma>.grep(/<{%*ENV<USER>}>/);
14:43 p6eval rakudo f23a45: OUTPUT«p6eval␤»
14:44 TBA so whats different, im confused!
14:44 PerlJam TBA: um ... nothing?  I don't understand what you're confused about
14:45 TBA me either, thought the earlier examples that failed were the same, they werent, my mistake!
14:46 TBA except now I'm getting malformed regex...
14:46 PerlJam TBA: careful if %*ENV<USER> has regex-special characters in it
14:46 TBA .grep(/ <{ %*ENV<USER> }> /)
14:46 PerlJam (presumably %*ENV<USER> won't, but maybe you're using another environment var)
14:47 TBA r: .say for <p6test a b c>.grep(/ <{ %*ENV<USER> }> /)
14:47 p6eval rakudo f23a45:  ( no output )
14:47 PerlJam r: %*ENV<USER> = "foo(bar";   / <{ %*ENV<USER> }> /;
14:47 p6eval rakudo f23a45:  ( no output )
14:47 TBA r: .say for <p6eval a b c>.grep(/ <{ %*ENV<USER> }> /)
14:47 p6eval rakudo f23a45: OUTPUT«p6eval␤»
14:47 TBA argh
14:47 TBA when i run it locally I get
14:47 UncleFester6 joined #perl6
14:48 TBA ===SORRY!=== Malformed regex at eval_0:1
14:48 PerlJam r: %*ENV<USER> = "foo(bar";   "test" ~~ / <{ %*ENV<USER> }> /;
14:48 p6eval rakudo f23a45: OUTPUT«===SORRY!===␤Malformed regex␤at eval_0:1␤»
14:48 flussence oh crap.
14:48 PerlJam TBA: notice the open paren with no closing paren.
14:48 TBA the user im matching against is 'www-data', atm
14:49 flussence it's interpolating that string as *code*
14:49 TBA could it be the - then?
14:49 PerlJam r: %*ENV<USER> = "www-data";   "test" ~~ / <{ %*ENV<USER> }> /;
14:50 p6eval rakudo f23a45: OUTPUT«===SORRY!===␤Malformed regex␤at eval_0:1␤»
14:50 PerlJam looks that way
14:50 PerlJam r: %*ENV<USER> = "wwwdata";   "test" ~~ / <{ %*ENV<USER> }> /;
14:50 p6eval rakudo f23a45:  ( no output )
14:50 TBA excellent, at least its not just me going crazy :)
14:51 tadzik :)
14:52 ggoebel jnthn++ are you expecting the reduction in memory usage to translate into performance gains as well?
14:53 TBA yey :D my bit of code finally works
14:53 TBA thanks all :)
14:53 jnthn ggoebel: Well, it's compile time performance gains if anything, but I'm hoping for a bit. It's hard to guess how much it'll be.
14:54 moritz not swapping is usually a big performance win :-)
14:54 tadzik TBA: curious, is this some work-related stuff?
14:54 TBA yeah, sort of
14:54 jnthn moritz: Well, there is that... :)
14:54 TBA this is what we had:
14:54 TBA killit('-m'); killit('-s'); killit('-q');  sub killit { my $switch = shift; my $output = `ipcs $switch`; $output =~ s/(0x\w+)\s+(\w+)\s+(\w+)/my ($i,$o) = ($2,$3); system("ipcrm $switch $i") if $ENV{USER} =~ \/^$o\/;/ego; }
14:54 TBA this is what we now have (in p6):
14:55 tadzik "i would prefer to have this (...) stuff in perl6 than using moose" -- heard from a cow-orker today :)
14:55 TBA my $user = %*ENV<USER>; for ('-q','-s','-m') { for qqx{ipcs $_}.trim.lines.grep(/$user/)[0..*] -> $line { say qqx{echo ipcrm $_ $line.trim.split(/\s+/)[1]}.trim; } }
14:55 TBA far neater, shorter and far more readable :)
14:55 TBA (ignore the 0..* range :p forgot to remove it)
15:00 masak tadzik: curious question: are you really collaborating with bovines at your workplace?
15:00 tadzik not really
15:00 tadzik TIL a new word :)
15:01 PerlJam TBA: random commentary:  the first .trim looks superfluous to me.  rather than split(/\s+/), maybe you want to use .words ?
15:01 TBA PerlJam: that was it! .words! couldn't remember what it was called (and was writing this on paper earlier... don't ask :p)
15:01 PerlJam (actually, the fact that you used .trim 3 times sets off my spidey sense)
15:02 TBA the first .trim was to get rid of blank lines, that was before I'd remembered I needed to grep for the user anyway, so it was chopping off the first few lines of ipcs output
15:04 masak PerlJam: what I say three times is true? :)
15:05 TBA interesting...
15:05 TBA oh ignore me
15:06 TBA if I want to use $line.words[1], inside a qqx{}, can I?
15:06 TBA I actually get
15:06 TBA $line output, followed by ".words[1]"
15:06 TBA i.e., with this: for <-q -s -m> { for qqx{ipcs $_}.lines.grep(/$user/) -> $line { say $line.words[1]; say qqx{echo ipcrm $line.words[1]}; } }
15:06 PerlJam TBA: .words.[1]
15:06 masak tadzik: tentative theory about the 'p' command thing. see the error output? it has spaces (or some other blank character) split in everywhere. maybe the command was really 'perl6' (or something) and got similarly split on Windows?
15:07 TBA PerlJam++
15:07 TBA you are genius!
15:07 moritz aye, looks like some UTF-16 messup
15:07 TBA its now this short:
15:07 TBA for <-q -s -m> { for qqx{ipcs $_}.lines.grep(/$user/) -> $line { say qqx{echo ipcrm $line.words.[1]}; } }
15:07 tadzik masak: hmm, possibly
15:08 moritz TBA: instead of "say qqx ..." you can simply do   run("...")
15:08 TBA moritz: i did have a play with run() earlier but something didn't work, cant remember what now though
15:09 moritz TBA: ah my fault, it's shell()
15:09 TBA moritz++ :)
15:09 TBA now slightly shorter again
15:09 TBA for <-q -s -m> { for qqx{ipcs $_}.lines.grep(/$user/) -> $line { shell "echo ipcrm $line.words.[1]"; } }
15:10 TBA thats perfectish, ignoring the fact I got my ipcrm call wrong :p
15:10 TBA for <q s m> { for qqx{ipcs -$_}.lines.grep(/$user/) -> $line { shell "echo ipcrm -$_ $line.words.[1]"; } }
15:16 TBA thanks for all your help :) gonna pick this up again on monday I think!
15:16 TBA left #perl6
15:18 moritz wait, 'shell "echo"'
15:18 moritz then you can just say() it right away, no?
15:18 PerlJam testing probably
15:19 moritz I guess so, yes
15:19 jferrero joined #perl6
15:22 thelazydeveloper joined #perl6
15:35 zb wow, i managed to stop & start playback with XMMS2::Client :D
15:36 tadzik nice :)
15:42 GlitchMr joined #perl6
15:43 benabik joined #perl6
15:57 pmichaud good morning, #perl6
15:57 hoelzro morning pmichaud
15:57 jnthn o/ pmichaud
15:57 [Coke] ho
15:58 jnthn pmichaud: I tried the MSI here. It really needs a completion screen of some kind, I think.
15:59 [Coke] a picture of the rakudo arch saying "done"?
15:59 PerlJam [Coke]: "this is just the beginning..."  :)
16:00 jnthn pmichaud: Mostly the problem is that it can run and install successfully *very* fast :)
16:00 jnthn pmichaud: Which is good but the first time I ran it, I assumed it had failed! :)
16:01 jnthn pmichaud: That detail aside, the install it does looks good to me.
16:01 jnthn pmichaud++
16:02 pmichaud jnthn: I can probably get a completion screen there.
16:03 pmichaud I'm also thinking of a license screen.
16:03 moritz don't
16:03 jnthn *nod*
16:03 masak pmichaud! \o/
16:03 moritz those annoy me to no end
16:03 moritz and on linux you don't have that either
16:04 flussence but this isn't on linux!
16:04 pmichaud okay, I can skip the license screen.
16:04 pmichaud I figure strawberrry includes the screen and they tend to think about such things, so I'd follow their lead :)
16:04 pmichaud but we can skip for a while
16:04 pmichaud anyway, I just need to find out what wix provides without too much difficulty.
16:04 pmichaud should I create desktop and/or program shortcuts?
16:05 jnthn pmichaud: Other things to consider: adding a Start Menu entry to the Rakudo Perl REPL, and perhaps the Perl 6 book.
16:05 flussence shortcuts... to what?
16:05 jnthn (Under a Rakudo folder)
16:05 flussence oh, docs. yeah, that seems like a good idea.
16:05 pmichaud flussence: repl
16:05 jnthn People expect to find *something* after they install.
16:12 pmichaud so, other than the .msi stuff, are there any issues that need to be addressed before cutting a standard star release?
16:12 jnthn pmichaud: Yes, see in the backlog.
16:13 jnthn pmichaud: of note
16:13 jnthn 16:47 < moritz> pmichaud: https://gist.github.com/3188234
16:14 pmichaud I'm thinking it picked up an nqp from somewhere else.
16:17 pmichaud moritz: do you have the Makefile from that nopaste handy?
16:18 pmichaud (the one in the rakudo-2012.07 dir)
16:32 pmichaud or, more likely, the nqp executable was missing entirely
16:34 [Coke] # 07/26/2012 - rakudo++ (22806); niecza (90.4%); pugs (41.04%)
16:35 sorear good * #perl6
16:35 TimToady top o' the mornin' to ya
16:35 masak sorear! \o/
16:36 [Coke] https://gist.github.com/147​6841#file_perl6_pass_rates - niecza is failing 7, pugs 54, and rakudo 195.
16:36 TimToady .oO(feels more like the bottom of the morning to me, actually...)
16:36 masak 195!?
16:36 TimToady tc maybe?
16:36 pmichaud szabgab++ also reports spectest failures on p6c
16:36 jnthn Probably all that strings crap.
16:36 jnthn (tc and frineds)
16:37 jnthn I think moritz++ fudged a bunch of it though maybe it didn't make it into the stats today.
16:37 [Coke] that was from yesterday. didn't realize today's run is still going.
16:37 spider-mario joined #perl6
16:37 jnthn Ah, then that certainly explains it.
16:37 [Coke] will report in again in an hour when rakudo's done. ;)
16:38 dalek star: 9132003 | pmichaud++ | skel/tools/build/Makefile.in:
16:38 dalek star: Break out separate 'wxs' and 'msi' Makefile targets.
16:38 dalek star: review: https://github.com/rakudo/star/commit/9132003e48
16:40 moritz pmichaud: the nqp executable wasn't missing, I pasted its --version output below
16:40 pmichaud okay, yes.
16:40 moritz makefile forthcoming...
16:41 vmspb joined #perl6
16:41 moritz pmichaud: http://moritz.faui2k3.org/tmp/star-Makefile
16:42 MayDaniel joined #perl6
16:42 pmichaud moritz: any particular reason for including --gen-nqp, btw?
16:42 pmichaud using --gen-nqp causes star to clone a new nqp repo.
16:42 pmichaud instead of using the one in the tarball bundle.
16:42 moritz pmichaud: no reason, and I wasn't aware of the distinction
16:43 pmichaud I bet that's the issue.
16:43 moritz might well be
16:43 pmichaud since the current nqp would be the qast-merged one.
16:43 moritz wouldn't it pick up the nqp from NQP_REVISION?
16:43 pmichaud ...what NQP_REVISION?
16:44 pmichaud star doesn't have a NQP_REVISION.  Looking.
16:44 pmichaud oh, wait
16:44 moritz but it's copy of rakudo has one
16:44 pmichaud I did my command wrong; I could be wrong about all of this. checking.
16:45 pmichaud never mind, I was in the wrong dir.
16:45 pmichaud --gen-nqp doesn't clone the repo.
16:45 pmichaud testing now.
16:45 pmichaud sorry for false alarm
16:48 pmichaud moritz: I need the Makefile from the rakudo-2012.07 dir
16:48 pmichaud (not from the top-level star dir)
16:49 moritz pmichaud: sorry, I don't have that anymore; I configured rakudo separately to get installed, so the old makefile is gone
16:49 moritz pmichaud: but I can try to reproduce it
16:49 pmichaud okay.
16:57 pmichaud perl Configure.pl --gen-parrot --gen-nqp --prefix=$HOME/p6/install && make install  # just ran successfully for me
16:58 awwaiid joined #perl6
16:58 moritz woah, rakudo HEAD compiled the setting using just 1.2GB memory. jnthn++ jnthn++
16:59 TimToady I don't suppose anyone here has an Indian visa and would like to go to India after YAPC::EU to give a talk in my place?
16:59 moritz about 800MB after creating the PAST nodes
16:59 moritz the star installer still runs --gen-nqp on the other machine
17:00 * TimToady could not get an Indian visa because he was unwilling to sign a paper saying he would engage in no religious activity
17:00 TimToady all my activity is religious :)
17:01 moritz pmichaud: ok, I can't reproduce my earlier problem here. Call it static noise :-/
17:03 benabik Hm.  Compiling CORE:  'Can only use get_how on a SixModelObject' in NQPClassHow.mixin (src/stage2/gen/nqp-mo.pir:10944) (src/stage2/gen/nqp-mo.pm:1249)
17:03 TimToady and http://www.techniche.org/techniche12/ has already advertised far and wide that I'm coming, and is asking who I can send instead...
17:04 PerlJam What's Damian doing around that time period?  :)
17:04 au TimToady: non-tourist visa not an option?
17:05 TimToady they offered a business visa with abridgement of freedom
17:05 au er, wow.
17:05 TimToady last year I went on a conference visa, but that is not longer an option without letters from three different departments of Indian government
17:05 TimToady that for some reason I have to get, not the conference
17:06 au suddenly the Department of Homeland Security seems rather enlightened by comparison...
17:06 TimToady PerlJam: I've heard TheDamian swear that he was never going back to India.
17:06 [Coke] (no religious activity) clearly india /has/ religious activity already. are they trying to avoid missionaries?
17:07 TimToady [Coke]: I believe that's the motivation, but my principles will not allow me to sign such a document even if I wasn't planning any religious activity by their definition
17:07 pmichaud TimToady++ # principles
17:08 au TimToady++
17:08 au telepresence?
17:08 TimToady and quite apart from the stand-for-freedom aspect, when you believe in a higher authority, you cannot swear to a lesser authority to ignore the higher authority... :/
17:08 au (either via robots or just regular Skype)
17:08 PerlJam au++ good idea
17:09 sorear \o/ au
17:09 pmichaud we had a presentation from Ward Cunningham here via telepresence -- it worked out pretty well.
17:10 au "one cannot swear to a lowercased authority to ignore the titlecased Authority"
17:10 au o/ sorear
17:10 TimToady well, not to ruin your joke, but India is usually titlecased as well
17:11 au so it's a lowercased authority with a titlecased name...
17:11 TimToady works for me :)
17:11 au :)
17:12 moritz p6: say 'foo'.codes
17:12 p6eval rakudo f23a45: OUTPUT«No such method 'codes' for invocant of type 'Str'␤  in block  at /tmp/e0Vt4ywFOm:1␤␤»
17:12 p6eval ..niecza v19-15-g051783d: OUTPUT«3␤»
17:12 sorear well, if you can do your paperwork in a caseless script, it might work out
17:12 sorear I hear India has a few of those
17:14 * [Coke] pounds on au and asks for a pugs patch. ;)
17:14 [Coke] er, pounces!
17:14 au ouch!
17:14 [Coke] pounds sounds much less friendly. stupid fingers with their automatic typeahead.
17:14 TimToady sorear: ॺ
17:14 [Coke] sorry, au!
17:14 sorear .u ॺ
17:14 phenny U+097A (No name found)
17:15 au np at all
17:15 TimToady DEVANAGARI LETTER HEAVY YA
17:15 sorear would dollars and/or newtons be more friendly?
17:15 [Coke] # 07/27/2012 - rakudo++ (22902); niecza (90.02%); pugs (40.86%)
17:15 [Coke] bother, pugs already pushed down below 41%
17:15 TimToady someone should replace phenny with something more modern
17:15 [Coke] fails: n:4,p:55,r:65
17:15 sorear maybe kilograms, depending on how much you like to pretend that they are the same as kiloponds
17:16 Tene TimToady: phenny gets the unicode table from python
17:16 TimToady pugs don't like pounds, and python doesn't like unicode
17:16 sorear so we just need to convince sbp(?) to upgrade python?
17:16 Tene sorear: could be; I don't know if it's any better in newer python.
17:17 moritz this should take care of the majority of the fails
17:17 dalek roast: 37532bc | moritz++ | S32-str/substr.t:
17:17 dalek roast: fudge substr.t for rakudo
17:17 dalek roast: review: https://github.com/perl6/roast/commit/37532bc264
17:17 TimToady python is too busy being enterprise-ready to worry about being Enterprise-ready...
17:18 [Coke] au: how much effort would it be to make the Test functions not autothread?
17:18 [Coke] I tried making the function signatures match those in spec, but it was insufficient to stop the autothreading.
17:18 [Coke] pugs: say 3.FatRat
17:18 p6eval pugs: OUTPUT«*** No such method in class Int: "&FatRat"␤    at /tmp/dsOIYR5pwk line 1, column 5 - line 2, column 1␤»
17:18 * [Coke] sighs.
17:19 moritz pugs: sub f(Object $x) { say $x }; f 1|2|3
17:19 * [Coke] needs keys to the evalbot to fix that, he guesses.
17:19 p6eval pugs: OUTPUT«1␤2␤3␤»
17:19 moritz [Coke]: do you have root on feather?
17:19 au [Coke]: does isa_ok autothread?
17:20 au iirc it had the Any|Junction|Pair signature to explicitly prevent autothreading
17:20 au so perhaps Mu can be made to mean that
17:20 moritz pugs: sub f(Mu $x) { say $x }; f 1|2|3
17:20 p6eval pugs: OUTPUT«1␤3␤2␤»
17:20 moritz pugs: say Pair ~~  Any
17:20 p6eval pugs: OUTPUT«1␤»
17:20 [Coke] Sorry, user coke is not allowed to execute '/bin/bash' as root on feather.
17:20 au pugs: sub f(Any|Junction $x) { say $x }; f 1|2|3
17:20 p6eval pugs: OUTPUT«any(VInt 1,VInt 2,VInt 3)␤»
17:21 [Coke] ah, if I add |Junction temporarily, that'll do it?
17:21 au but as a quickfix, s/Any/Any|Junction/ will likely do
17:21 au worth a try
17:21 moritz [Coke]: I've given you sudo privs on feather
17:21 [Coke] sweet! Thanks.
17:21 moritz [Coke]: you can  sudo su - p6eval
17:22 moritz [Coke]: and set up a working ghc + pugs combo
17:22 au quietfanatic++ # committing 543ec380 and fixing #24
17:22 moritz [Coke]: and then we'll regularly rsync it over to feather3 where p6eval lives
17:22 * TimToady admits to a fondness for quietfanatic
17:22 au phenny: tell quietfanatic thx for the fix! iirc saying "fixes #24" in the commit message would close the ticket on GitHub automagically
17:22 moritz TimToady: you're forgiven :-)
17:22 phenny au: I'll pass that on when quietfanatic is around.
17:24 [Coke] moritz: if I do a single build, how does that help automatic updates?
17:24 [Coke] or is it just then on me to do a rebuild whenever I want one?
17:25 Tene please use sudo -i instead of sudo su; it'll save you trouble someday.
17:28 moritz [Coke]: I just want something that rsyncs nicely
17:28 moritz [Coke]: preferably along with a cron job that rebuilds regularly
17:29 pmichaud arrrgh, too much chaos here at the house to work productively; I'll bbiaw
17:29 TimToady std: / $(%*ENV<foo>) /
17:29 p6eval std e52e3ca: OUTPUT«ok 00:00 42m␤»
17:30 TimToady r: say %*ENV<USER>
17:30 p6eval rakudo f23a45: OUTPUT«p6eval␤»
17:31 TimToady rn: $_ = %*ENV<USER>; say so / $(%*ENV<USER>) /
17:31 p6eval niecza v19-15-g051783d: OUTPUT«Use of uninitialized value in string context␤  at /home/p6eval/niecza/lib/CORE.setting line 1272 (warn @ 5) ␤  at /home/p6eval/niecza/lib/CORE.setting line 268 (Mu.Str @ 15) ␤  at <unknown> line 0 (ExitRunloop @ 0) ␤  at /home/p6eval/niecza/lib/CORE.settin…
17:31 p6eval ..rakudo f23a45: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter % (must be quoted to match literally) at line 2, near "*ENV<USER>"␤»
17:31 TimToady *bug
17:32 [Coke] au: cabal install pugs-compat ->
17:32 TimToady well, not stdbug :)
17:32 [Coke] cabal: Couldn't read cabal file "HUnit/1.2.5.0/HUnit.cabal"
17:32 au [Coke]: which user? on feather?
17:33 [Coke] ah. http://www.haskell.org/pipermail/h​askell-cafe/2012-July/102392.html
17:33 [Coke] it's for the p6eval user on feather.
17:33 [Coke] perhaps I can pick an earlier version of it.
17:35 [Coke] ah. need cabal to be updated.
17:35 [Coke] was cabal on feather installed with a package manager, I wonder?
17:36 au looks like "sudo apt-get install haskell-platform" can fix it.
17:37 [Coke] on it.
17:37 au or, if that takes too long, "sudo apt-get install cabal-install" first would do as well.
17:39 [Coke] haskell-platform fails epically to install here. I get a curses screen telling me things are screwy.
17:40 [Coke] cabal-install fails also:
17:40 [Coke] https://gist.github.com/3189296
17:42 [Coke] that seems like a job for someone who actually manages a linux box to look at.
17:44 [Coke] oooh. I wonder if I can just steal coke's .cabal dir for now.
17:46 [Coke] nope. ah well
17:49 * geekosaur wonders if that's going to cause problems at the next reboot
17:50 au [Coke]: I've dpkg -i'd cabal-install_0.14.0-2_i386.deb from sid's pool
17:50 au [Coke]: so it should work now
17:50 au (where it = feather:/usr/bin/cabal)
17:56 [Coke] au: \o/
17:59 au glad it worked
17:59 au ...& have a good localtime hackers \o
18:22 spider-mario joined #perl6
18:23 spider-mario joined #perl6
18:26 Chillance joined #perl6
18:28 * [Coke] is still cleaning up the cabal stuff for p6eval - looks like it's complaining that some installs of pre-existing packags were borked. reinstalling now.
18:40 [Coke] moritz: p6eval on feather1 now has a working pugs build.
18:41 [Coke] I assume the p6eval bot knows how to deal with the dist/build/pugs... stuff when executing.
18:42 moritz [Coke]: you give it a path to a binary, it executes it
18:42 moritz [Coke]: what neeeds to be rsync'ed to feather3?
18:44 [Coke] I figured you were syncing the whole home dir. ;)
18:44 [Coke] probably ~/.cabal and ~/Pugs.hs
18:46 moritz is Pugs.hs/Pugs/dist/build/pugs/pugs the right binary to execute?
18:48 moritz pugs: say 1.FatRat
18:48 p6eval pugs: OUTPUT«1␤»
18:48 moritz [Coke]: does that look better? :-)
18:49 moritz [Coke]: now please adjust build-scripts/rebuild-pugs.sh in the evalbot repo
18:51 [Coke] they seem reasonable. might need some tweaking for when "cabal configure" reports an error and things need manual upgrading. (that's a problem on my daily smokes, also)
18:51 [Coke] can probably kill the GHCBIN-related settings.
18:51 * [Coke] will poke later.
18:51 [Coke] moritz++
18:51 [Coke] au++
18:52 [Coke] sorear++ # for trying to do this before.
18:52 moritz [Coke]++ # for doing it
19:05 kaare_ joined #perl6
19:19 quietfanatic Ah, cool.  I'll keep that in mind.
19:19 phenny quietfanatic: 17:22Z <au> tell quietfanatic thx for the fix! iirc saying "fixes #24" in the commit message would close the ticket on GitHub automagically
19:34 Tene I'm always entertained by the apparent time-reversal in responding to a message before it being apparently delivered.
19:34 [Coke] I was just thinking that.
19:34 diakopter [Coke]: were you just thinking that?
19:35 Tene :D
19:38 [Coke] What's a way to say "this parameter is defined" in a signature?
19:38 quietfanatic that one was three minutes early even.
19:39 [Coke] :q
19:39 * hugme hugs [Coke], good vi(m) user!
19:39 TimToady what do you mean by "say"?
19:39 TimToady oh, duh
19:39 TimToady :D
19:39 Tene :D
19:39 Tene Oh, heh
19:40 fhelmberger joined #perl6
19:40 * TimToady contemplates kicking [Coke]++ for excessive cleverness :)
19:46 TimToady but [Coke] already self-kicked with :q--and now I'm wondering if I should be glad he didn't say :wq...
19:57 sftp joined #perl6
19:59 fgomez joined #perl6
20:01 birdwindupbird joined #perl6
20:23 masak lol, hi #perl6.
20:23 masak I'll be with you in a minute. socializing with jnthn. :)
20:23 diakopter I always imagine that lol as a cackle
20:24 moritz do you all watch the opening of the Olympic games?
20:27 TimToady not on till this evening here
20:31 [Coke] "GE is a proud sponsor..."
20:34 birdwindupbird joined #perl6
20:42 [Coke] is the old pugs svn repo around somewhere?
20:43 benabik Doesn't the Pugs.hs.git history stretch back into SVN land?
20:43 benabik I guess not.  I thought I saw git-svn-ids in there.
20:44 [Coke] anyone has a copy of this: https://github.com/perl6/Pugs.hs/issues/22 it'd be helpful.
20:45 [Coke] ah, sweet, found a copy of the svn repo at openfoundry
20:47 moritz [Coke]: the mu repo contains everything there ever was in the pugs svn repo
20:48 moritz (maybe except empty directories, which git doesn't track)
20:50 [Coke] mu repo?
20:50 [Coke] oh, on github.
20:56 [Coke] moritz: I don't see pugs in there. ;)
21:01 diakopter [Coke]: see https://github.com/perl6/mu/​commits/master?&amp;page=61 middle of the page for removing pugs from mu
21:02 diakopter are you looking for a particular commit?
21:06 [Coke] diakopter: I'm looking for the a script in the repo that compiled Prelude.pm down to c
21:07 diakopter pugscc ?
21:08 [Coke] au said it was "the pmc2c packer script"
21:09 [Coke] Warning: you are leaving 30213 commits behind, not connected to
21:09 [Coke] any of your branches:
21:10 benabik That's... a lot of commits.
21:11 [Coke] from mu
21:11 moritz [Coke]: pugs has been deleted from mu, but it's all in the history of that repo
21:12 moritz [Coke]: the README says something about when stuff was deleted
21:12 moritz git checkout 6f0203060f  should get you before the big cleanup
21:15 [Coke] moritz++
21:36 telex joined #perl6
21:46 lue hello world o/
21:53 masak lue! \o/
21:53 sorear lue!
21:53 sorear and masak!
21:54 jnthn evenin'
21:54 lue (if my calculations last night were correct, the opening ceremony started at 12:12 my local time, so shh!)
21:55 TimToady we don't to know who wins
21:57 diakopter I think NBC wins the opening ceremonies
21:57 masak jnthn! \o/
21:58 jnthn masak! \o/
21:58 lue the fact that this is the 30th olympiad (if I'm correct) makes putting roman numerals before the show's title on NBC unfortunate...
22:03 TimToady rpn: say [XXX] <O l y m p i c s>
22:03 p6eval rakudo f23a45: OUTPUT«O l y m p i c s␤»
22:03 p6eval ..pugs: OUTPUT«*** ␤    Unexpected end of input␤    expecting "\\", variable name or "p"␤    at /tmp/PiyEhJtA8U line 2, column 1␤»
22:03 p6eval ..niecza v19-15-g051783d: OUTPUT«Unhandled exception: Excess arguments to ANON, used 2 of 8 positionals␤  at /tmp/0eFP9vo_Zx line 0 (ANON @ 1) ␤  at /home/p6eval/niecza/lib/CORE.setting line 2543 (reduceop @ 12) ␤  at /tmp/0eFP9vo_Zx line 1 (mainline @ 3) ␤  at /home/p6eval/niecza/lib/COR…
22:04 TimToady er...huh?
22:05 TimToady pugs: say [XXX] <O l y m p i c s>;
22:05 p6eval pugs: OUTPUT«*** ␤    Unexpected end of input␤    expecting "\\", variable name or "p"␤    at /tmp/BFx5iAM95a line 2, column 1␤»
22:06 TimToady n: say [XXX] <O l>;
22:06 p6eval niecza v19-15-g051783d: OUTPUT«O l␤»
22:06 TimToady n: say [XXX] <O l y>;
22:06 fgomez joined #perl6
22:06 p6eval niecza v19-15-g051783d: OUTPUT«Unhandled exception: Excess arguments to ANON, used 2 of 3 positionals␤  at /tmp/rZU35pWkOz line 0 (ANON @ 1) ␤  at /home/p6eval/niecza/lib/CORE.setting line 2543 (reduceop @ 12) ␤  at /tmp/rZU35pWkOz line 1 (mainline @ 3) ␤  at /home/p6eval/niecza/lib/COR…
22:06 TimToady n say 1 X 2 X 3
22:06 TimToady n: say 1 X 2 X 3
22:06 p6eval niecza v19-15-g051783d: OUTPUT«1 2 3␤»
22:06 TimToady n: say 1 XX 2 XX 3
22:06 p6eval niecza v19-15-g051783d: OUTPUT«1 2 3␤»
22:07 TimToady n: say 1 XXX 2 XXX 3
22:07 p6eval niecza v19-15-g051783d: OUTPUT«Unhandled exception: Excess arguments to ANON, used 2 of 3 positionals␤  at /tmp/BrtPBgCcuP line 0 (ANON @ 1) ␤  at <unknown> line 0 (KERNEL map @ 2) ␤  at /home/p6eval/niecza/lib/CORE.setting line 245 (crossop @ 5) ␤  at /tmp/BrtPBgCcuP line 1 (mainline @…
22:07 quietfanatic all kinds of strangeness
22:08 diakopter n: say [X-X] <O l y>
22:08 p6eval niecza v19-15-g051783d: OUTPUT«[31m===[0mSORRY![3​1m===[0mâ�¤â�¤Unsupported use of y///; in Perl 6 please use tr/// at /tmp/vzZF26NNWC line 1:â�¤------> [32msay [X-X] <O l y[33mâ��[31m>[0mâ�¤â�¤Undeclared name:â�¤  'X-X' used at line 1â�¤â�¤Parse failedâ�¤â�¤Â»
22:09 diakopter heh
22:09 diakopter n: say [X-X] <O l tr>
22:09 quietfanatic parsing it as an array literal.
22:09 p6eval niecza v19-15-g051783d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤â�¤Use of a closing delimiter for an opener is reserved at /tmp/OWtYnR9csl line 1:â�¤------> [32msay [X-X] <O l tr[33mâ��[31m>[0mâ�¤â�¤Undeclared name:â�¤        'X-X' used at line 1â�¤â�¤Parse failedâ�¤â�¤Â»
22:10 diakopter reserve this!
22:10 TimToady n: say [-X] <O l tr<
22:11 p6eval niecza v19-15-g051783d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤â�¤Unable to parse quote-words subscript; couldn't find right angle quote at /tmp/Sb32ssPvla line 1 (EOF):â�¤------> [32msay [-X] <O l tr<[33mâ��[31m<EOL>[0mâ�¤â�¤Undeclared name:â�¤    'X' used at line 1â�¤Undeclared routine:â�¤ 'tr' used …
22:11 quietfanatic reduce ops and rray literals can lead to ambiguity.
22:11 TimToady n: say [X-X] <O l tr<
22:11 p6eval niecza v19-15-g051783d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤â�¤Unable to parse quote-words subscript; couldn't find right angle quote at /tmp/07ms6dkZ0x line 1 (EOF):â�¤------> [32msay [X-X] <O l tr<[33mâ��[31m<EOL>[0mâ�¤â�¤Undeclared name:â�¤   'X-X' used at line 1â�¤Undeclared routine:â�¤       'tr' us…
22:11 TimToady n: say [X-X] <O l tr<foo>
22:12 p6eval niecza v19-15-g051783d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤â�¤No delimiter found at /tmp/XRZnuIz4HO line 1 (EOF):â�¤------> [32msay [X-X] <O l tr<foo>[33mâ��[31m<EOL>[0mâ�¤â�¤Undeclared name:â�¤   'X-X' used at line 1â�¤â�¤Parse failedâ�¤â�¤Â»
22:12 diakopter backtracking too far...
22:12 TimToady yeah
22:12 TimToady n: say [X-X] <O l tr<foo<bar<
22:12 p6eval niecza v19-15-g051783d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤â�¤Unable to parse quote-words subscript; couldn't find right angle quote at /tmp/u52IwTOJPJ line 1:â�¤------> [32msay [X-X] <O l tr<[33mâ��[31mfoo<bar<[0mâ�¤â�¤Undeclared name:â�¤      'X-X' used at line 1â�¤Undeclared routine:â�¤       'tr' used …
22:12 TimToady n: say [X-X] <O l tr/foo/bar/
22:12 p6eval niecza v19-15-g051783d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤â�¤Action method quote:tr not yet implemented at /tmp/0vo0q4C1U_ line 1 (EOF):â�¤------> [32msay [X-X] <O l tr/foo/bar/[33mâ��[31m<EOL>[0mâ�¤â�¤Undeclared names:â�¤      'O' used at line 1â�¤     'X-X' used at line 1â�¤Undeclared routine:â�¤       'l'
22:12 p6eval ..us…
22:13 TimToady n: say s<foo> = 'bar';
22:14 p6eval niecza v19-15-g051783d: OUTPUT«Unhandled exception: Unable to resolve method subst in type Any␤  at /tmp/P4aX2JEIBG line 1 (mainline @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3929 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3930 (module-CORE @ 564) ␤  at /home/…
22:16 TimToady nr: sub tr { { foo => 42 } }; say tr<foo>
22:16 p6eval niecza v19-15-g051783d: OUTPUT«[31m===[0mSORRY![31m===[0m��No delimiter found at /tmp/AiQeVW7lkf line 1 (EOF):�------> [32msub tr { { foo => 42 } }; say tr<foo>[33m�[31m<EOL>[0m��Parse failed��»
22:16 p6eval ..rakudo f23a45: OUTPUT«42␤»
22:18 diakopter my ubuntu vm says Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0.0)
22:18 * diakopter sad
22:19 lue blog post!  https://rdstar.wordpress.com/2012/​07/27/making-an-if-game-in-august/
22:21 * masak reads
22:22 masak hm. I will make some internal refactors to Game::Engine, yes. 'things' and 'rooms' may end up in inner classes.
22:23 masak but the key here is not to expose any inner workings out of Game::Engine.
22:23 masak so it may be implemented using hashes, or inner classes, or bubble gum and wires.
22:24 masak aww, I was hoping to read about what the *game* would be like. :) not just a bunch of tech you plan to put in or around Game::Engine...
22:26 pmichaud hello again
22:26 diakopter o/
22:26 masak \o/
22:26 lue sorry. There will be a hamster wheel room though (like the carousel room in Zork II, only hamster-wheel oriented)
22:27 lue (to the tell the truth, I don't have much of a plan for what the actual game will be... /me plans a game)
22:27 masak you're free to copy/paste or augment Game::Engine, of course... but I would prefer it if you didn't.
22:27 masak I can guarantee there's actually a lot of cool stuff you'll be able to do just using the same principles Crypt::Game uses to build the adventure.
22:28 sergot joined #perl6
22:28 snearch joined #perl6
22:29 jnthn o/ pmichaud
22:29 lue Yeah, I was thinking of how far I'd end up modifying Adventure::Engine, and how I didn't want to go too far.
22:30 masak I'm trying to make it so that it won't need any modifications at all.
22:30 masak hooks are really flexible, for example.
22:31 lue .oO( Maybe if I end up being that stubborn, I could use OOP anyway and then write an interface that takes info out of objects and puts it in Adventure::Engine as necessary :) )
22:32 lue But I'll definitely try my hand at using Adventure::Engine the way it's used in crypt first, so I can get a better feel for it.
22:32 masak I am using OOP. I'm just not modeling things as objects which I'm not exposing to the outer world anyway.
22:33 masak lue: yes, a very good start is reading through the BUILD submethod in Crypt::Game, to see what the possible repertoire of Game::Engine is in terms of world creation.
22:34 TimToady arguably, variables and closures are objects too :)
22:35 * lue is starting to realize perhaps his mind is too stuck on how his implementation of a P6 IF engine would've done things.
22:36 lue masak: Are you sure you don't want to rename it Game::Engine? (or IFGame::Engine) :)
22:36 masak well, Adventure isn't perfect, but it's more exact than Game.
22:36 masak my brain keeps thinking of these as "text adventures", not as "interactive fiction", for some reason.
22:37 lue (just referring to the fact you keep typing Game::Engine :) )
22:37 TimToady adventure used to be called 'advent' on 6-char machines, so you could confuse everyone into thinking it's Christmas
22:38 drbean joined #perl6
22:38 masak lue: heh :)
22:39 sorear mm, ADV
22:40 masak I'm gonna skip doing the writeup of today's crypt hacking tonight. it's already to late and I need to sleep more than I need to be timely with the blog post.
22:40 masak too*
22:40 * TimToady first played adventure on a DECSYSTEM-20 in the lobby of a the hotel of a DECUS convention.
22:41 masak you'll get a crypt blog post tomorrow morning :)
22:41 lue this is an awesome coincidence by the way: https://en.wikipedia.org/w​iki/Cross_product#Mnemonic
22:41 masak lue: yeah, I read that yesterday :)
22:41 masak quietfanatic and I were talking about vector outer products.
22:44 tokuhiro_ joined #perl6
22:46 sorear I still prefer the Levi-Cevita symbol
22:51 masak I never truly grokked tensors.
22:53 * sorear hugs masak
22:53 masak :)
22:54 masak I can see that they are the generalization of scalar/vector/array/...
22:54 masak but I never went much further than that.
22:58 sorear I like to think about the duality between vectors and linear functions
22:59 sorear (in some branches of geometry, vectors are defined as differential operators, which I do not quite grok but it's another interesting perspective)
23:00 sorear a rank-2 tensor is a linear function from vectors to vectors (when I push a spinning wheel in this direction, it responds in that direction)
23:00 sorear an object constructed from two directions, like an area element or an angular momentum
23:01 sorear a bilinear function from vectors to scalars, or a quadratic function of a single vector
23:01 sorear if you look at gravitational potential near a point and Taylor-expand it, you get a scalar potential, a rank 1 (vector) force field, then a rank 2 tensor which reflects the tides
23:02 sorear (let's stop here, because if you aren't planning to dive into GR you will never use a tensor of rank higher than 2)
23:04 masak ok.
23:04 masak I get all of the above on an intuitive level, at least.
23:06 sorear on another level tensors are a kind of strong typing for physics.  they make it hard to mix certain types of values that should never be mixed.
23:07 sorear when you construct an equation using the tensor formalism, you get for free reasonable behavior under changes of coordinates
23:07 sergot good night o/
23:09 masak ok, so it's types with built-in dimensional analysis.
23:16 whiteknight joined #perl6
23:26 masak I just implemented winning the game in crypt.
23:26 masak so now you can play it all the way through. it's still a rough experience, but at least it's possible.
23:26 masak developing an adventure game using tests is wonderful.
23:27 masak latest commits are here: https://github.com/masak/crypt/commits/master
23:27 masak today's blog post will arrive tomorrow :)
23:44 takadonet joined #perl6

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

Perl 6 | Reference Documentation | Rakudo