Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2013-02-11

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:04 awwaiid joined #perl6
00:05 jeffreykegler joined #perl6
00:06 jeffreykegler left #perl6
00:06 jeffreykegler joined #perl6
01:16 grondilu joined #perl6
01:20 colomon joined #perl6
02:22 FROGGS joined #perl6
02:45 pyrimidine joined #perl6
03:30 uvtc left #perl6
03:35 colomon rn: say FatRat.new(1, 10**100)
03:36 p6eval niecza v24-23-g0520c7c: OUTPUT«1E-100␤»
03:36 p6eval ..rakudo a3869a: OUTPUT«0.0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000001␤»
03:37 benabik That's a lot of zeros, but I somehow doubt that was 100 of them.
03:38 colomon easy enough to check
03:39 colomon r: say [+] FatRat.new(1, 10**100).gist.comb.map(* == '0' ?? 1 ?? 0)
03:39 p6eval rakudo a3869a: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Found ?? but no !!â�¤at /tmp/4_ixLUJiBa:1â�¤------> [32m0**100).gist.comb.map(* == '0' ?? 1 ?? 0[33mâ��[31m)[0mâ�¤Confusedâ�¤at /tmp/4_ixLUJiBa:1â�¤------> [32m0**100).gist.comb.map(* == '0' ?? 1 ?? 0[33mâ��[31m)[0mâ�¤    expecting any of:â�¤  â€¦
03:39 colomon r: say [+] FatRat.new(1, 10**100).gist.comb.map(* == '0' ?? 1 !! 0)
03:39 p6eval rakudo a3869a: OUTPUT«No such method 'count' for invocant of type 'Int'␤  in method reify at src/gen/CORE.setting:5568␤  in method reify at src/gen/CORE.setting:5491␤  in method reify at src/gen/CORE.setting:5491␤  in method gimme at src/gen/CORE.setting:5881␤  in sub  at src/gen/CORE.s…
03:39 sorear r: say FatRat.new(1, 10**100).subst(rx/0*/, -> $x { $x.chars }, :g)
03:39 p6eval rakudo a3869a: OUTPUT«10.99010␤»
03:39 sorear r: say FatRat.new(1, 10**100).subst(rx/0+/, -> $x { $x.chars }, :g)
03:39 p6eval rakudo a3869a: OUTPUT«1.991␤»
03:39 benabik r: FatRat.new(1,10**100).Str ~~ /0\.(0*)1/; say $1.chars
03:39 p6eval rakudo a3869a: OUTPUT«No such method 'chars' for invocant of type 'Any'␤  in block  at /tmp/t1K_vcMONZ:1␤␤»
03:39 benabik r: FatRat.new(1,10**100).Str ~~ /0\.(0*)1/; say $0.chars
03:39 p6eval rakudo a3869a: OUTPUT«99␤»
03:39 sorear looks good to me
03:39 sorear 99 zeros
03:40 colomon sorear++
03:41 colomon r: say [+] FatRat.new(1, 10**100).gist.comb.map({ $_ == '0' ?? 1 !! 0 })
03:41 benabik Right, zero indexed.  Looked short to me, but I guess my IRC window is wider than I thought.
03:41 p6eval rakudo a3869a: OUTPUT«Cannot convert string to number: radix point must be followed by one or more valid digits in '.⏏' (indicated by ⏏)␤  in method Numeric at src/gen/CORE.setting:10511␤  in sub infix:<==> at src/gen/CORE.setting:2885␤  in sub infix:<==> at src/gen/CORE.setting:2885␤  …
03:41 colomon r: say FatRat.new(1, 10**100).gist.comb
03:41 p6eval rakudo a3869a: OUTPUT«0 . 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1␤»
03:41 colomon r: say FatRat.new(1, 10**100).gist.comb.map({ $_ == '0' ?? 1 !! 0 })
03:41 p6eval rakudo a3869a: OUTPUT«Cannot convert string to number: radix point must be followed by one or more valid digits in '.⏏' (indicated by ⏏)␤  in method Numeric at src/gen/CORE.setting:10511␤  in sub infix:<==> at src/gen/CORE.setting:2885␤  in sub infix:<==> at src/gen/CORE.setting:2885␤  …
03:42 colomon r: say FatRat.new(1, 10**100).gist.comb.map({ $_ eq '0' ?? 1 !! 0 })
03:42 p6eval rakudo a3869a: OUTPUT«1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0␤»
03:42 colomon duh
03:42 colomon r: say [+] FatRat.new(1, 10**100).gist.comb.map({ $_ eq '0' ?? 1 !! 0 })
03:42 p6eval rakudo a3869a: OUTPUT«100␤»
03:45 MikeFair_ joined #perl6
04:21 rindolf joined #perl6
04:56 preflex_ joined #perl6
05:11 SamuraiJack joined #perl6
05:15 alec__ joined #perl6
05:17 xinming_ joined #perl6
05:58 daniel-s_ joined #perl6
06:07 moritz \o
06:10 sorear o/
06:14 sorear phenny: tell masak curl http://masak.org/carl/ # PHP error
06:14 sorear erm
06:14 sorear ...do I have the wrong bot?
06:14 phenny sorear: I'll pass that on when masak is around.
06:16 moritz just ESLOW
06:18 sorear aha
06:23 quester joined #perl6
06:26 arlinius joined #perl6
06:28 [Coke] I got a DB file for aloha if we want to restart him on feather.
06:35 kaleem joined #perl6
07:13 rindolf joined #perl6
07:19 b1rkh0ff joined #perl6
07:24 rindolf joined #perl6
07:31 rindolf joined #perl6
07:46 quester left #perl6
08:09 domidumont joined #perl6
08:23 arnsholt o/
08:25 sjn \o
08:33 FROGGS joined #perl6
08:38 domidumont joined #perl6
08:38 arnsholt sjn: Were you the one trying to wrap libxml with Zavolaj at the Oslo hackathon?
08:41 kivutar joined #perl6
08:53 fhelmberger joined #perl6
08:57 kresike joined #perl6
08:57 kresike hello all you happy perl6 people
08:58 sjn arnsholt: no, that wasn't me
08:58 sjn arnsholt: I was one of the guys just hanging out (and organizing a little :)
08:59 arnsholt Ah, right
09:00 moritz I remeber it wasn't a #perl6 regular
09:00 moritz but I don't know who :(
09:00 arnsholt Oh, well
09:00 arnsholt Maybe I'll get around to it at some point
09:00 FROGGS moritz: ohh, that narrows it down :P
09:02 arnsholt Just wanted to point out that libxml bindings would be a lot more feasible now that the sized stuff has landed
09:03 FROGGS I wish I had time to fiddle with it
09:04 arnsholt If you find the tuits, you know where to find me
09:04 arnsholt OH! And if you want to hack on the SDL stuff again, you can kill the terrible hacks you needed to get sized stuff as well
09:05 cosimo joined #perl6
09:07 FROGGS arnsholt: right, thats exactly the point where I want to test this :o)
09:07 FROGGS these hacks are horrible, and 64bit specific
09:08 arnsholt Excellent. It felt good to be able to close that bug =)
09:08 FROGGS ya
09:09 cosimo joined #perl6
09:09 arnsholt 'lo cosimo o/
09:10 jnthn morning o/
09:10 arnsholt Hopefully this visit will be longer than the last one =)
09:10 FROGGS hi jnthn
09:10 arnsholt G'morning, jnthn
09:14 * arnsholt is contemplating NativeCall string handling
09:14 * jnthn has meetings :)
09:14 arnsholt Lots of time to think of design issues then =D
09:18 moritz arnsholt: please also think of blob handling :-)
09:21 arnsholt Just huge chunks of bytes, you mean?
09:22 sjn arnsholt: If you want some, I have lots of tuits to give away :)
09:22 sjn although they're the wooden flat kind. :D
09:22 moritz arnsholt: yes
09:23 * jnthn guesses we wanna use Buf for those, or something like that
09:23 moritz arnsholt: usually there's an out-of-band signalling of the lenght of the blob somewhere
09:23 marloshouse_ joined #perl6
09:23 moritz jnthn: yes; the question is just how to design the interface to make sense to to the user
09:23 FROGGS sjn: I'd took one :o)
09:24 benabik joined #perl6
09:24 arnsholt At the NativeCall level, I think the person wrapping the API just has to cope with that part of it on their own
09:25 moritz arnsholt: well, but somehow we need the ability to turn tuple of (pointer to blob, length) into a Buf
09:25 moritz and the other way round
09:25 arnsholt Yeah, that's more annoying
09:25 arnsholt This, essentially is what I'm pondering really
09:26 arnsholt Since strings in C are blobs =)
09:26 * FROGGS .oO( Buf.new( pointer => $pointer, length => $bytes ) )
09:26 sjn FROGGS: feel free to get in touch next time you're in Oslo :)
09:27 FROGGS sjn: ohh, Oslo
09:27 FROGGS well, I try not to forget :/
09:27 kst joined #perl6
09:31 nwc10 jnthn: do the meetings have food?
10:02 ObseLeTe joined #perl6
10:17 arlinius joined #perl6
10:19 pete_ joined #perl6
10:19 Guest38296 This is the second time I've set aside a couple of days to really learn Perl6, and been caught by the build tool being unable to handle paths with spaces in
10:20 moritz then please use a path without spaces
10:20 moritz it's a TODO for us, but it really, really shouldn't stop you
10:21 Guest38296 *shrug* my home dir on my dev box has a space in it
10:21 Guest38296 I guess I could build a new VM, or use a different dev box, but most likely, I'll find a different project to hack on
10:21 FROGGS Guest38296: spaces are always a problem when compilers re involved... gcc cant handle that too when passing the path as an argument when invoking commands
10:27 arnsholt Heh. I actually didn't even know that was a problem
10:31 M_o_C joined #perl6
10:35 moritz IIRC parrot has fixed that problem, but nqp and rakudo haven't
10:37 * FROGGS tries
10:39 * moritz re-clones parrot with 5 MB/s and still finds it slow
10:39 moritz if the #parrot folks go through on their plan to rip out much stuff, it might be worth considering whether to restart with a slimmed-down repo
10:40 moritz and keep a full-history archive somewhere else
10:42 FROGGS might be nice to be able to say --gen-parrot-branch=slim or so
10:42 FROGGS to be able to switch and compare
10:43 moritz --gen-parrot=branch already works
10:46 FROGGS perfect
10:46 FROGGS Receiving objects:  27% (113060/407957), 30.81 MiB | 224 KiB/s
10:46 FROGGS -.-
10:59 FROGGS Building NQP ...
10:59 FROGGS /usr/bin/perl Configure.pl --with-parrot=/home/froggs/dev/folder with spaces/install/bin/parrot --make-install
11:00 moritz eeks, this is a nasty problem.
11:01 dalek rakudo/spacey: ebf80f9 | moritz++ | tools/lib/NQP/Configure.pm:
11:01 dalek rakudo/spacey: Small preparation to allow paths with spaces
11:01 dalek rakudo/spacey: review: https://github.com/rakudo/rakudo/commit/ebf80f9424
11:01 FROGGS what if we add quotes to the makefile.in?
11:01 dalek nqp/spacey: 8282ac5 | moritz++ | tools/lib/NQP/Configure.pm:
11:01 dalek nqp/spacey: Small preparation to allow paths with spaces
11:01 dalek nqp/spacey: review: https://github.com/perl6/nqp/commit/8282ac5755
11:02 moritz FROGGS: well, I fear that'll be necessary
11:02 moritz FROGGS: though I'd much rather prefer to add the quotes in Configure.pl
11:05 sqirrel joined #perl6
11:06 FROGGS moritz: I added quotes to lib/NQP/Configure.pm:293 and it looks good so far
11:10 moritz FROGGS: that's not the only place that needs patching :-)
11:10 pjcj joined #perl6
11:10 moritz FROGGS: wait a sec, I'll commit what I have now
11:11 dalek nqp/spacey: ae3b47d | moritz++ | tools/ (2 files):
11:11 dalek nqp/spacey: more fixes to allow spacey paths
11:11 dalek nqp/spacey: review: https://github.com/perl6/nqp/commit/ae3b47dabe
11:11 FROGGS ya, the created makefile is the borke too
11:11 FROGGS this step: /usr/bin/perl /home/froggs/dev/folder with spaces/install/lib/parrot/4.10.0-devel/tools/build/pmc2c.pl
11:12 moritz it now gets further
11:13 moritz up to the linking step
11:13 mtymula joined #perl6
11:13 mtymula hello i have one complaint...;p
11:13 mtymula after ugrating to never version of rakudo
11:13 mtymula i get problem like thjis:
11:13 mtymula Invalid operation on binary string   in method uc at src\gen\CORE.setting:2259   in method dispatch:<.=> at src\gen\CORE.setting:993   in block  at C:\Test_czy_mobilne\mylib\perl6-http-easy-master\perl6-http-easy- master\lib/HTTP/Easy.pm6:93
11:14 mtymula it is like i cannot do this: .=
11:14 mtymula but before i could...
11:14 FROGGS ahh, you get a Buf and it expects a Str there
11:15 mtymula but why?? it works with 2012 verion of rakudo
11:16 FROGGS IO::Socket::INET changed
11:17 mtymula so what can i do to upgrate my rakudo and have working program at the same time
11:17 mtymula ?
11:18 FROGGS mtymula: hold on, I'm inspecting HTTP::Easy right now
11:20 FROGGS hmmm, it is calling .get, that is right, .get should give you a Str IMO
11:27 mtymula FROGGS: so...
11:27 mtymula what can I do??
11:28 FROGGS wait a few minutes? I'm recompiling my rakudo to do some testing
11:41 mtymula ??
11:55 moritz it seems that pbc_to_exe doesn't like spacey paths :(
11:57 FROGGS :/
12:05 moritz but it's not pbc_to_exe's fault
12:05 moritz mlenz@mlenz-workstation:~/p6/with space/rakudo/nqp$ ../install/bin/parrot_config  libparrot_linkflags
12:05 moritz -L"/home/mlenz/p6/with space/rakudo/parrot/blib/lib" -lparrot
12:05 moritz mlenz@mlenz-workstation:~/p6/with space/rakudo/nqp$ ../install/bin/parrot_config  inst_libparrot_linkflags
12:05 moritz -L/home/mlenz/p6/with space/rakudo/install/lib -lparrot
12:06 moritz so the inst_ variant doesn't escape the path
12:06 moritz WTF.
12:08 moritz patched parrot.
12:09 FROGGS I'm still compiling -.-
12:17 pupoque joined #perl6
12:44 ggoebel_ joined #perl6
12:48 drbryane joined #perl6
13:13 FROGGS mtymula: what rakudo version are you using?
13:19 shinobicl joined #perl6
13:27 * [Coke] rips some stuff out of parrot in his "sixparrot" fork - commit messages going to #parrot if anyone wants to follow.
13:27 [Coke] (nothing substanstive yet, just some LHF)
13:28 [Coke] er, *substantive.
13:28 * moritz waits for the commit messages
13:29 moritz fwiw with my last parrot patch, NQP can now build with a space in the prefix path
13:29 moritz (branch 'spacey')
13:29 FROGGS cool
13:29 FROGGS will try in a bit
13:30 moritz rakudo will need some Makefile twiddling too
13:31 arnsholt jnthn: I managed to get a segfault while precompiling Test.pm (using a not entirely latest Rakudo)
13:32 jnthn arnsholt: Are you using a later NQP?
13:32 arnsholt NQP is at HEAD
13:32 jnthn That'd do it if your Rakudo is more than a day or two old.
13:32 jnthn STable layout changed a little.
13:33 arnsholt That's probably it then
13:38 [Coke] Is nqp/rakudo using parrot's calling conventions at all?
13:40 jnthn yes
13:41 [Coke] Danke.
13:41 not_gerd joined #perl6
13:41 not_gerd hello, #perl6
13:42 FROGGS hi not_gerd
13:42 arnsholt jnthn: Yeah, works fine now
13:43 not_gerd [Coke]: if you're looking for stuff to do, you could try resurrecting tools/build/nativecall.pl (or alternatively, port the current nci_thunk_gen.pir to Perl)
13:44 not_gerd nci_thunk_gen.pir pulls in TGE via data_json
13:44 [Coke] not_gerd: I'm working on a slimmed down parrot.
13:45 not_gerd [Coke]: I know
13:45 moritz [Coke]: yes, that would allow to slim down parrot by removing data_json and TGE
13:45 [Coke] Is nativecall being used?
13:45 [Coke] I thought rakudo/nqp rolled their own there.
13:45 moritz maybe for bootstrappy stuff? no idea
13:46 jnthn I'm not aware of it being used
13:46 moritz (I mean, for the parrot build)
13:46 moritz if not, we can rip that out too
13:46 jnthn The NativeCall Perl 6 module is all using stuff inside NQP
13:46 not_gerd afaik it's used for the dynops
13:47 jnthn ah
13:48 not_gerd jnthn: is there anything wrong with pulling down NQP/Rakudo dynops into Parrot core?
13:49 ponbiki joined #perl6
13:50 jnthn not_gerd: That you'd drag an incredible amount of other stuff with them?
13:51 not_gerd jnthn: as long as these are all specific to the Parrot backend, I don't see anything wrong with that
13:52 Rix joined #perl6
13:53 jnthn not_gerd: I don't see what it wins, it just means if I should need to change them they're further away from the things that use them.
13:54 jnthn It's probably *possible* to do technically, I'm just not sure what it wins.
13:56 PacoAir joined #perl6
13:57 bluescreen10 joined #perl6
13:58 not_gerd jnthn: if we take the idea of 'Parrot lite' with the sole purpose of supporting NQP seriously, that's the natural next step after ripping out all parts not used by NQP
13:58 masak is there a RC entry for http://c2.com/cgi/wiki?BagSumInManyProgrammingLanguages ?
13:58 phenny masak: 06:14Z <sorear> tell masak curl http://masak.org/carl/ # PHP error
13:59 masak sorear: I know, it's been that way for a while. :/
13:59 masak sorear: I'm sick of MediaWiki, so I'll probably put a really small PHP script there instead that allows me to edit content in the Simplest Possible way.
13:59 masak sorear: anyway, thanks for the reminder. was there something in particular you were looking for?
13:59 not_gerd jnthn: ie Parrot ends up as /nqp/backends/parrot-lite right next to /nqp/backends/jvm, ...
14:00 nwc10 and upgrade it to dogfood soon after?
14:00 masak nwc10: I'll check whether the web hotel that I'm renting allows Perl 6 at all.
14:01 nwc10 I thought that it would be something like that
14:01 nwc10 they're running Java? Or they consider that more insecure than PHP?
14:02 [Coke] not_gerd: anyway, current plan is to remove unused ops, libs, pmcs. to be done, also need to remove pasm (already have a branch somewhere that does 80% of this), tge, pge, pct, parrot-nqp, winxed, and anything that depends on them (unless we need it.)
14:06 wk joined #perl6
14:12 moritz pbc_to_exe is written in winxed
14:13 [Coke] only recently.
14:13 [Coke] but we can always check in the generated C.
14:14 [Coke] (or use more winxed.)
14:15 * not_gerd posted gist URI of his notes to #parrot
14:17 gtodd joined #perl6
14:18 gtodd hello ...
14:18 pmichaud good morning, #perl6
14:19 gtodd can I build/link parrot rakudo with libedit instead of readline?  is there a --prefer-libedit option or something?
14:19 [Coke] is libedit a dropin replacement for readline?
14:19 nwc10 pmichaud: thanks. I'll have a play. But it might end up being one big messy expression
14:20 gtodd mostly .... heimdal postgres client apps use it ... it's small and sometimes just included with the source I think
14:20 pmichaud is there a log I should read that discusses the reason for eliminating deletepos and existspos ?
14:22 gtodd sqlite will use it too ... I think it's a option for people who want to ship things using a pure BSD license but upstream folks put readline in the build stuff (autoconf cmake whatever) as the default for whatever historical reasons
14:22 pmichaud oops, kid needs unplanned ride to school -- bbi20
14:24 gtodd I don't know much about hacking Configure.pl files but there seems no easy way of pointing rakudo at libedit -  I just want readline or libedit for the REPL
14:24 gtodd ... errm obviously :-[
14:26 jnthn pmichaud: Think it was a day or two ago, maybe a little more..it arose out of reviewing what existspos and deletepos do when tadzik++ was looking at porting them over to the JVM.
14:26 Psyche^ joined #perl6
14:27 jnthn gtodd: I know that we get readline support through calling .readline_interactive, which is provided by Parrot. I'd guess if there's any way to make it happen, it'd be done at the point of configuring Parrot.
14:28 gtodd jnthn: ah ok
14:28 gtodd was looking at the wrong end
14:44 gtodd Just thinking of readline again is there a errm perl6 policy statement or longish term development goal of making perl more umm  "independent" of system libraries? perl5 on my system (5.17) has no dependencies except what I add through modules.  perl6 needs gmp icu readline just to build perl6 argh! :)
14:45 not_gerd r: module Foo { our $*FOO = "foo"; our sub foo { say $*FOO } }; { temp $*Foo::FOO = "bar"; Foo::foo }; Foo::foo
14:45 p6eval rakudo a3869a: OUTPUT«Can only use 'temp' on a container␤  in sub prefix:<temp> at src/gen/CORE.setting:12641␤  in block  at /tmp/ILNYDJYQk2:1␤␤»
14:45 not_gerd ^ what's the right way to do that
14:46 gtodd just realized this is probably a topic that appears ad nauseum  so I hereby retract my last comment/question ... but hope to be able to build small light perl6 with less external library linking one day :)
14:46 gtodd cheers
14:46 moritz not_gerd: there are two ways. The first is to use ordinary variable, and to temp() them
14:46 moritz not_gerd: and the nicer variant is to use $*foo variables (without any namescpaces), and define a new one with 'my $*foo' in an inner scope
14:47 moritz r: module Foo { my $*FOO = 'foo'; our sub foo { say $*FOO } }; { my $*FOO = 42; Foo::foo() }; Foo::foo();
14:47 p6eval rakudo a3869a: OUTPUT«42␤Dynamic variable $*FOO not found␤  in method gist at src/gen/CORE.setting:10513␤  in sub say at src/gen/CORE.setting:7598␤  in sub foo at /tmp/WeXlm4WM0y:1␤  in block  at /tmp/WeXlm4WM0y:1␤␤»
14:48 not_gerd r: 'module Foo { our $FOO = "foo"; our sub foo { say $FOO } }; { temp $Foo::FOO = "bar"; Foo::foo }; Foo::foo
14:48 p6eval rakudo a3869a: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Unable to parse expression in single quotes; couldn't find final "'"â�¤at /tmp/gtlIwJKH5R:1â�¤------> [32m $Foo::FOO = "bar"; Foo::foo }; Foo::foo[33mâ��[31m<EOL>[0mâ�¤    expecting any of:â�¤        statement listâ�¤        prefix or termâ�¤  â€¦
14:48 not_gerd r: module Foo { our $FOO = "foo"; our sub foo { say $FOO } }; { temp $Foo::FOO = "bar"; Foo::foo }; Foo::foo
14:48 p6eval rakudo a3869a: OUTPUT«bar␤foo␤»
14:48 not_gerd ^ I'll go with that, thanks
14:48 not_gerd moritz++
14:48 moritz you're welcome
14:49 pmichaud back again
14:49 not_gerd r: module Foo { our $FOO; our sub foo { say $FOO } }; { temp $Foo::FOO = "bar"; Foo::foo }; Foo::foo
14:49 skids joined #perl6
14:49 p6eval rakudo a3869a: OUTPUT«Any()␤Any()␤»
14:50 not_gerd ^ bug?
14:50 moritz wb, pm
14:50 moritz r: module Foo { our $FOO = Any; our sub foo { say $FOO } }; { temp  $Foo::FOO = "bar"; Foo::foo }; Foo::foo
14:50 p6eval rakudo a3869a: OUTPUT«bar␤Any()␤»
14:51 moritz yes, bug
14:51 * not_gerd needs to leave now
14:51 moritz 'our'-Variables have lots of bugs and limitations. We tend to avoid them whenever possible.
14:52 not_gerd bye, #perl6
14:52 not_gerd left #perl6
14:52 pmichaud jnthn: okay, I'll review the log and comment further if needed :)
14:57 mtymula FROGGS: sorry I was away...
14:57 gtodd
14:58 pmichaud moritz: (slimmed down parrot repo, from log)    yeah, I've often thought we should perhaps trim the repo somehow.
15:00 stopbit joined #perl6
15:00 FROGGS mtymula: no problem, I'm usually patient
15:00 pmurias joined #perl6
15:02 pmurias [Coke]: wouldn't it make sense to just instrument the thing that processes pir to record all the opcodes it uses and measure which ones aren't used by rakudo when running roast?
15:03 [Coke] ah, "just".
15:03 [Coke] Yes, having a list of used vs. unused opcodes would be most helpful. Thank you for volunteering to do that bit. :)
15:03 FROGGS [Coke]++
15:03 pmichaud also need to see which opcodes are used by build process, which might be different from set in roast
15:04 Shozan joined #perl6
15:04 [Coke] pmichaud: and then we have to keep running it once we rip stuff out of the build process.
15:04 sizz_ joined #perl6
15:05 [Coke] but having the list run during spectest would be a great start, yes. It's easy enough to find candidates without that, however.
15:05 pmichaud just doing a grep for pir:: and nqp:: in the rakudo and nqp repos should be a really good start
15:06 p5eval_ joined #perl6
15:06 eviltwin_b joined #perl6
15:06 cxreg2 joined #perl6
15:07 Tene_ joined #perl6
15:07 Tene_ joined #perl6
15:07 moritz (especially if you include generated .pir files)
15:07 diakopte1 joined #perl6
15:07 kresike joined #perl6
15:07 atrodo joined #perl6
15:08 cotto joined #perl6
15:08 raiph__ joined #perl6
15:08 FROGGS I really like the idea of having an rakudo-optimized parrot
15:08 pmichaud FROGGS: I like the idea also.  Should've happened years ago.  :-P
15:09 FROGGS maybe I can hook in at some point
15:09 clkao joined #perl6
15:09 FROGGS pmichaud: ya, would be a shame to not try to "rescue" all the work that has gone into it
15:12 pmichaud FROGGS: ...unless the rescue ends up diverting too many resources from more productive tasks.
15:13 FROGGS ya, you're right
15:14 atrodo So my understanding from the logs is that rakudo has no use for nci?
15:15 pmurias [Coke]: processing all the *.ops file with s/^(inline?\s*op\b.*)$/$1..recoding stmt.../ (or something similiar) should be enough
15:17 FROGGS mtymula: so what is your rakudo version?
15:24 pmurias [Coke]: start of a quick hack for that: http://pastie.org/6117345
15:27 kaare_ joined #perl6
15:30 pmichaud atrodo: it's hard to know exactly what rakudo needs or doesn't need from existing Parrot.  As far as we know, Rakudo doesn't have a direct requirement for Parrot's NCI.  But we might be overlooking something, or there could be an indirect requirement that we've forgotten about
15:31 pmichaud if rakudo or nqp are using Parrot's NCI, we'd probably try to refactor to remove that requirement, though, and I suspect such refactoring wouldn't be hard.
15:35 salv0 joined #perl6
15:36 Liz joined #perl6
15:39 [Coke] pmurias - what is that supposed to get me? a list of parrot ops?
15:49 [Coke] (there is a tools/dev/sear
15:49 [Coke] whoops. search_ops.pl
15:53 SunilJoshi joined #perl6
15:56 thou joined #perl6
16:03 atrodo pmichaud: Okay. [Coke], I'm going to try and rip out nci on your sixparrot branch
16:06 kresike bye folks
16:07 nyuszika7h joined #perl6
16:09 cotto joined #perl6
16:16 Gwyxx joined #perl6
16:20 kaleem joined #perl6
16:22 [Coke] atrodo: want a commit bit?
16:22 [Coke] Or are you going to fork my fork?
16:22 atrodo [Coke]: If I can do this successfully, yea, I'll come asking for a commitbit
16:23 [Coke] hokay.
16:24 * masak .oO( spork me on github )
16:25 kivutar joined #perl6
16:27 jnthn decommute &
16:32 bluescreen10 joined #perl6
16:39 sweet_kid joined #perl6
16:40 rafl joined #perl6
16:40 spider-mario joined #perl6
16:41 bluescreen100 joined #perl6
16:43 SamuraiJack joined #perl6
17:10 Chillance joined #perl6
17:10 FROGGS joined #perl6
17:12 jms53 joined #perl6
17:18 rurban rakudo will build faster by removing pmc's and ops ??? OMG
17:18 [Coke] not that that's a primary goal, but sure.
17:19 * pmichaud wonders if his sarcasm detector is needed here.
17:20 * [Coke] tries to assume a pleasant effect from other senders unless coke himself is in a really crappy mood.
17:20 [Coke] gah. *affect.
17:21 pmichaud yes, I make that assumption also, which is why I sometimes miss sarcasm.  :-)
17:22 FROGGS haha, sheldons wherever I look *g*
17:22 [Coke] rurban: how goes p2? do you think an nqp port will be possible?
17:23 arlinius joined #perl6
17:23 not_gerd joined #perl6
17:23 rurban nqp-p2 is easy, already doable. the compiler and vm is stable.
17:23 FROGGS was p2 the thing which is a she?
17:23 rurban I'm using a new parser and grammar soon (leg => marpa)
17:23 rurban yes
17:24 PerlJam rurban: will you be making an nqp-p2 or leave that to someone else?
17:24 rurban but the AST might need some docs
17:24 rurban PerlJam: I will do, but currently I'm busy with something else
17:24 [Coke] rurban++
17:24 rurban I also have to rewrite my B::C compiler a bit
17:25 not_gerd rurban: removing ops and pmcs is more about making potential refactors easier and moving from Parrot-the-ecosystem to Parrot-the-library that powers NQP
17:25 not_gerd that's my story, and I'm sticking to it ;)
17:26 moritz and removing stuff is easy(ish)
17:26 rurban But it will not be faster. you will do more harm than help
17:26 [Coke] Just an experiment at this point. Much easier for [Coke] to delete someone else's code than write his own.
17:26 moritz so it doesn't need huge benefit to justify it
17:26 [Coke] rurban: "harm" ?
17:26 rurban You should inspect the calling convention and undo the nci method madness
17:26 not_gerd rurban: sure, but that's hard
17:26 rurban ops2c should be reverted to the perl5 lib, so that nqp can go away
17:27 rurban or use nqp-p6 (i have a branch)
17:27 [Coke] rurban: yes, ops2c is already on the todo list.
17:27 Liz joined #perl6
17:27 rurban and use bacek's llvm ops2c
17:27 rurban which jit's the ops.
17:27 rurban Also in some of my branches
17:28 rurban So I guess you'll need nqp for the jit
17:28 rurban I fixed llvm integration already to use static or shared llvm libs
17:29 pmichaud rurban:  just as a general note, it's easy for people to say what should be done.  commits are what count.
17:29 rurban And removing imcc -O1 was also not helpful if you want a faster rakudo
17:30 rurban I'm just warning for the next axe which will destroy months of work, and which will not be undoable
17:30 pmichaud reverts are easy.  it's easy to abandon branches.
17:30 rurban This happened to parrot every two year or so
17:31 PerlJam rurban: I think the next axe to fall will likely kill parrot though ... we'll see what happens.
17:31 [Coke] rurban: all this is happening in a branch as an experiment. That's it.
17:31 [Coke] (not even a branch, a branch on a fork, at the moment.)
17:31 pmichaud if things work as expected, then nqp unmodified should be able to run on both parrot master and on sixparrot
17:31 rurban Perljam: Only parrot can kill parrot, rakudo can just abandon parrot.
17:32 pmichaud so there's no "axe" that is undoable.
17:32 rurban all the previous axes did more harm than benefits
17:32 Liz_ joined #perl6
17:33 [Coke] rurban: what axes?
17:33 rurban only people with a global understanding should axe out stuff.
17:34 rurban your plan to strip parrot to a bare rakudo vm. how should parrot performance be competitive then?
17:34 pmichaud rurban: you're reading too much into "your plan".
17:34 rurban think of rakudo AND performance
17:35 spider-mario moritz: are you sure that qq["$file" --show-config] will work with paths that contain "?
17:35 rurban well, I don't like useless suicidal tendencies our of shere agony and desperation, and false accusations
17:36 spider-mario why not just: open my $CONFIG, '-|', $file, '--show-config'
17:36 rurban out of
17:36 pmichaud rurban: it's an experiment.  If anyone's accusing you of anything, I missed it.
17:36 not_gerd rurban: how does removing unused ops, pmcs and getting rid of nqp-rx hurt performance?
17:36 [Coke] rurban: I'm certainly not falsely accusing anyone here.
17:36 rurban someone says: threads are useless, someone else removes it.
17:36 not_gerd it won't help, but it won't hurt as well
17:36 rurban not you coke, someone else
17:36 [Coke] rurban: no one is removing threads.
17:36 rurban just saying :)
17:36 [Coke] rurban: yes, so go fight with him, please, not us.
17:37 pmichaud rurban: please don't assume that diakopter++ is speaking for the rest of us.
17:37 rurban just an example. nqp e.g. would be needed for jitting
17:37 rurban bacek should be asked
17:37 rurban but maybe the perl5 ops2c would work also.
17:38 [Coke] yes, that conversation already happened on #parrot.
17:38 rurban parrot compiletime does not worry me. parrot runtime worries me.
17:38 [Coke] rurban: yes.
17:38 pmichaud rurban: that point has been made repeatedly, by myself included.
17:38 rurban oh, sorry. I thought we are on #parrot
17:38 [Coke] no one really cares about making compile time faster. side effect only.
17:38 rurban I left #parrot out of anger yesterday
17:39 pmichaud then you're speaking from an incomplete picture of things.
17:39 pmichaud perhaps we should limit discussion to only people with a global understanding of stuff.  :-)
17:40 rurban not_gerd: they were probably needed for some planned subsystem to make things faster, not slower.
17:40 rurban left #perl6
17:40 PerlJam Assuming that Rakudo obtains speeds comparable to other language implementations, what's left to make it "production ready" besides threads?
17:41 pmichaud "It has to run all of CPAN."
17:41 [Coke] that's insane.
17:41 PerlJam ouch.
17:41 [Coke] is p5 compatibility mode a requirement?
17:41 pmichaud I don't know that "threads" are required for "production ready".
17:42 pmichaud that's what I don't like about the "production ready" term.  If we start saying "what does it need to be production ready", it's a slippery-slope to "it has to run all of Cpan"
17:42 [Coke] certainly nice to have. are threads in the spec as first class things, or just side effects about program execution order?
17:42 [Coke] ah, my sarcasm detector was broken.
17:42 timotimo pmichaud: i currently can't properly create a ipython kernel implementation because there are no "real" threads in rakudo yet, just one data point
17:43 pmichaud [Coke]: I tried to put the quotes there as a sarcasm hint :)
17:43 spider-mario could that not be solved with coroutines that would not necessarily run in parallel?
17:43 [Coke] If someone solidifies the spec, I promise to have a go at an implementation with the current parrot threads.
17:43 pmichaud timotimo: I agree that real threads prevent the task you're wanting to do.  But I've written a ton of production code in my lifetime that doesn't require threads.
17:43 [Coke] not going to shake out any issues unless someone starts writing code against them.
17:44 pmichaud if we say "Perl 6 isn't production ready because I need X", it'll never be "production ready"
17:45 pmichaud [Coke]: "solidify the spec" isn't really the model we tend to use.  What typically happens is someone prototypes some code to solve a problem, then we try to adapt it to other problems, then we converge on a solution, then we write a spec.
17:45 pmichaud I.e., we tend to follow an Internet/RFC model (bless the spec after implementation)
17:45 PerlJam pmichaud: To obtain feature parity with other languages in roughly the same niche (if there can be such a thing),  I think we'll need some answer to "threads".   This is how outsiders will judge "production ready".
17:46 pmichaud PerlJam: "feature parity" is a reasonable approximation
17:46 spider-mario do we really want threads as our concurrency model?
17:46 [Coke] pmichaud: ok. then I'm not your guy for the initial implementation, at least not without a lot more tuits.
17:46 pmichaud [Coke]: but I can give a concrete and spec'd example.
17:46 pmichaud it's the same one I gave rurban and have been asking about for several years
17:46 pmichaud implement a threaded version of the hyperops
17:47 pmichaud i.e.,   get    @a >>+<< @b   to run threaded.
17:47 [Coke] Yes, I'm familiar with that request.
17:47 pmichaud that's specced, it's not likely to change, it wants threading.
17:47 pmichaud once we have that example in place, we can look at the underlying threading model and see where the warts are and what else we need to improve threading
17:48 pmichaud there are other places where threading is implicit in Perl 6
17:48 [Coke] "underlying threading model"... which doesn't exist from rakudo's perspective. yes?
17:48 [Coke] just "this would go faster if we used threads".
17:48 [Coke] s/would/might/, even.
17:48 pmichaud a bit more than that
17:49 pmichaud we know that the hyperops are intended to be a way for the compiler to exploit and optimize parallelism.  the implementation doesn't do that yet.
17:50 pmichaud neither rakudo nor the Perl 6 specify exactly how that parallelism is to be exploited... specifically because the language designer(s) don't feel they have a strong enough handle on the problem to say "this is the way it should be done"
17:50 pmichaud s/nor the Perl 6 specify/nor the Perl 6 spec say/
17:51 [Coke] OK. let me stop you there: I'm not your guy.
17:51 pmichaud fair enough.
17:51 [Coke] So, I'll just focus on things I have proper shaped tuits for.
17:51 pmichaud I'm not that guy either.  :)
17:52 pmichaud timotimo: my recommendation for progress in your situation is not to ask "where can I find Perl 6's threading model" or "how will threads work in Rakudo?"  The way forward is to say "here's what I want threads to look like" or "here's how my application should look" and then we start building libraries or code to make it work.
17:54 pmichaud many parts of the parrot implementation suffer (with respect to Perl 6, but other languages also) from saying "we think this is what you'll want this feature to look like, so here it is and hopefully you can adapt to it".  Instead what is needed is for Parrot's features to be driven and derived directly from the needs of the languages.
17:54 pmichaud in the case of Perl 6 threading, it needs to be derived from applications that will use it.
17:55 timotimo pmichaud: this is pretty easy to do actually. all i need is a straight port of this tiny bit of code: https://github.com/ipython/ipython/blob/master/IPython/kernel/zmq/heartbeat.py#L29
17:55 timotimo it'd be totally okay if it looks different, i don't have any preferences
17:56 pmichaud what would you want that code to look like if written in Perl 6?  I suspect you'll want "async" somewhere.
17:56 timotimo the init part is boring, when it comes to threads, it's just the run function that i need
17:56 timotimo actually, i have no idea why this even has a thread running. probably the zmq.device call being blocking
17:57 timotimo so if i could just async zmq.device(zmq::FORWARDER, $!socket, $!socket);  that would be fantastic
17:57 PerlJam timotimo: you have a perl6-zmq ?
17:58 timotimo arnsholt is working on it
17:58 timotimo he asked me to help out, i'll have a look, too, but i don't really know what needs twiddling where and when
18:00 pmichaud it sounds like something we want as a starting point is a simple implementation of async
18:00 pmichaud and by "simple" I mean "we know it won't work in the general case" simple.
18:01 pmichaud just something that gets two threads of control running at the same time -- try not to worry about synchronizing variables or the like just yet
18:02 timotimo yes, that would be enough at this point
18:02 timotimo $!socket would have to be renamed $!socktype, because it isn't even a socket object that's being passed
18:03 timotimo at the same time, these device objects are implemented in pyzmq to use threading, too. but that's no big deal, they don't synchronise data with other threads either
18:03 pmichaud it won't be production ready threads by any stretch of the imagination, but we'll at least have something that we can start building from in Parrot, or at least can start to bridge the gap between what Perl 6 wants ("async") and whatever Parrot provides
18:03 * jnthn home
18:03 arnsholt timotimo: Any kind of hacking goes really
18:03 timotimo hm, except there's one global ZMQ context object that probably needs to be passed, i'll check.
18:04 arnsholt Tests, public perlish interface, and general giving it a spin
18:04 arnsholt I haven't actually tested any of the C paths yet =)
18:11 moritz I would be grateful if somebody could test the 'spacey' branche of nqp on windows
18:12 moritz after a thorough cleaning, in Rakudo   perl Configure.pl --gen-parrot=master --gen-nqp=spacey # in rakudo's 'spacey' branch should be able to build nqp in a prefix with a space in the name
18:12 jnthn moritz: I'll just grab a completely fresh checkout and build that.
18:13 arnsholt timotimo: Oh, and if you have any good ideas for HOW we can test the code, I'm all ears
18:13 arnsholt My current hack of a plan would be something involving Perl 5 code probably =)
18:19 timotimo oof, good question.
18:22 arnsholt Yeah, I know =)
18:24 not_gerd bye, #perl6
18:24 not_gerd left #perl6
18:25 rindolf Hi all.
18:25 rindolf Hi PerlJam , pmichaud
18:26 rindolf Hi jnthn
18:26 * rindolf sprinkles jnthn with some magic vowel dust.
18:26 larks I can't wait until the concurrency stuff gets finished
18:26 jnthn o/ rndlf :P
18:27 rindolf jnthn: heh.
18:27 PerlJam rindolf: greetings
18:27 rindolf jnthn: what's up?
18:27 rindolf PerlJam: what's new?
18:27 PerlJam rindolf: the leading "j" *is* a vowel  ;)
18:27 rindolf inthan.
18:27 PerlJam rindolf: nothing much.   Looking at play-perl.org now.
18:27 rindolf PerlJam: ah.
18:27 jnthn rindolf: Not much. Been doing $dayjob stuff today and now I'm home and pondering food and working on some NQP stuff :)
18:29 PerlJam larks: why?  What do you hope to use it for?
18:29 rindolf PerlJam: ah, I see that it is open source.
18:29 cxreg larks: do you mean the parrot threads support?
18:29 rindolf Or at least has sources on GitHub.
18:30 rindolf jnthn: nice.
18:30 PerlJam rindolf: yep, I just cloned it in anticipation of adding a feature or two :)
18:30 rindolf moritz: hi, how are you?
18:30 rindolf PerlJam: great.
18:32 moritz rindolf: within one standard deviation of my own, peculiar average
18:32 rindolf moritz: OK.
18:32 rindolf moritz: any news?
18:33 rindolf PerlJam: did you also do a GitHub fork?
18:33 PerlJam rindolf: no, not yet.
18:33 rindolf PerlJam: ah.
18:33 rindolf OK.
18:33 larks Well, I just like threads and concurrency =x
18:33 PerlJam I tend to poke around in a clone before I fork for pull requests.
18:33 moritz rindolf: there's always news, depending on how small-scale news you find interesting
18:33 moritz the pope retires!
18:33 PerlJam larks: bummer.  I was hoping that you'd have some specific use-case :)
18:34 moritz nqp can now (in a branch) build in a directory with space in the path
18:34 larks Well, I mainly design hardware so I think in parallel.  How can I break this problem into parallel tasked, how do I need to sync them, etc etc
18:35 rindolf PerlJam: seems like I'll have to buy some http://en.wikipedia.org/wiki/B%C3%B6rek for the Israeli Perl Workshop (also for other people) because I'm not too fond of the Buddha Burgers' Vegan food served there (second year in a row).
18:35 larks sorry for disappointing you =x
18:36 rindolf PerlJam: and I'm gonna give this there - http://www.shlomifish.org/temp-sherlock-holmes-awk/Sherlock-Holmes-about-Awk.xhtml
18:36 rindolf As a 10 minutes talk.
18:36 arnsholt r: .name.say for Grammars.^methods
18:36 p6eval rakudo a3869a: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Undeclared name:â�¤    Grammars used at line 1. Did you mean 'Grammar'?â�¤â�¤Â»
18:36 arnsholt r: .name.say for Grammar.^methods
18:36 p6eval rakudo a3869a: OUTPUT«parse␤parsefile␤MATCH␤MATCH_SAVE␤INTERPOLATE␤OTHERGRAMMAR␤RECURSE␤prior␤orig␤target␤from␤pos␤CAPHASH␤!cursor_init␤!cursor_start␤!cursor_start_subcapture␤!cursor_capture␤!cursor_push_cstack␤!cursor_pass␤!cursor_fail␤!cursor_pos␤!cursor_next␤!cursor_more␤!reduce␤!red…
18:36 rindolf Does anyone here have any plans on coming to the Israeli Perl Workshop in 25-Feb?
18:39 jnthn moritz: Seems the make install-dev in Parrot fails with:
18:39 jnthn Error: couldn't copy installable_ops2c.exe to \ops2c.exe: No such file or direct
18:39 jnthn ory
18:39 jnthn moritz: Or did I need that at master too?
18:39 moritz jnthn: yes, you need parrot/master (which has a fix to pbc_to_exe with spacy paths)
18:40 moritz though it should error out eariler in that case
18:40 moritz erm wait, no
18:43 domidumont joined #perl6
18:43 jnthn moritz: master got further (or different):
18:43 jnthn C:\consulting\omg spaces\rakudo\install\lib\libparrot.lib
18:43 jnthn Error: couldn't copy libparrot.dll to \libparrot.dll: No such file or directory
18:44 M_o_C joined #perl6
19:01 diakopte1 pmichaud: why would you suggest I suggested to remove threads from parrot?
19:01 moritz jnthn: hm, weird
19:02 pmichaud diakopte1: ??
19:08 diakopter pmichaud: nm, you didn't
19:10 aindilis joined #perl6
19:11 FROGGS rindolf: 've got no time and no money... sorry
19:11 pmichaud diakopter: I just know that rurban was using your disagreement with him as a basis for criticizing what other people are doing, and wanted to disabuse him of that.
19:12 rindolf FROGGS: OK, sorry to hear that. :-(
19:13 jcrigby joined #perl6
19:13 FROGGS ya, I'd like be a regular on YAPCs and other workshops and hackathons
19:14 arnsholt FROGGS: Where in the world are you at?
19:18 jnthn dinner, bbl
19:20 thou joined #perl6
19:21 Chillance joined #perl6
19:21 moritz google announced that there'll be a GSoC this year
19:22 moritz I wonder if Rakudo should apply as an org separately
19:23 masak I think so.
19:23 PerlJam definitely.
19:24 masak yet another take on defaults for subtypes: they work as "a constructive proof that the type is inhabited".
19:24 masak phrase borrowed from https://en.wikipedia.org/wiki/Dependent_type but applies in this case, too.
19:30 FROGGS arnsholt: berlin
19:31 FROGGS moritz: I already have two shirt but had no chance to mentor at all
19:36 arnsholt Berlin is workable. I think there might be a workshop in Copenhagen coming up. That might be feasible from Berlin?
19:36 moritz the Nordic Perl Workshop, yes
19:37 FROGGS well, there is the german workshop next month in berlin
19:37 perlhelp joined #perl6
19:37 perlhelp Hi! :)
19:37 FROGGS hi perlhelp
19:37 nwc10 shouldn't everyone just go to http://act.yapc.eu/gpw2013/ and be done? :-)
19:37 FROGGS copenhagen, sounds good
19:38 arnsholt moritz: Is there a website or something for that yet?
19:38 perlhelp This room does perl 6 but not perl 5?
19:38 perlhelp Hi FROGGS
19:38 arnsholt nordic perl workshop on Google only turns up the old versions, but not the upcoming one =(
19:38 nwc10 perlhelp: correct. you'd want #perl for your perl 5 questions, I think
19:38 FROGGS or irc.perl.org
19:39 nwc10 #perl-help on irc.perl.org
19:39 perlhelp May I please kindly ask if there is a big difference between 5 and 6?
19:39 nwc10 on irc.perl.org, #perl is a "social" channel
19:39 moritz perlhelp: there is
19:39 PerlJam perlhelp: yes, there are
19:39 FROGGS channel perl-help on irc.perl.org
19:39 FROGGS ohh, you said that already
19:39 nwc10 there's less difference between Perl 5 and Perl 6 than (say) between Perl and Python
19:39 perlhelp OH, so people cannot discuss perl 5 hee
19:40 nwc10 no, we didn't say *that* :-)
19:40 PerlJam perlhelp: Perl 5 and Perl 6 are different languages in the same family.  If you're interested in doing things with Perl, you likely want Perl 5.  See #perl or the other channels already mentions
19:40 PerlJam er, mentioned
19:40 FROGGS only if you want to compare perl 5 statements and perl 6 ones
19:40 moritz well, we discuss Perl 5 from a design point of view; we don't help with your Perl 5 problems.
19:41 masak <moritz> rindolf: within one standard deviation of my own, peculiar average
19:41 * masak gets a vicarious thrill out of that statement
19:41 rindolf masak: hi.
19:41 moritz masak: :-)
19:41 moritz and you know where I stole that phrase from :-)
19:41 rindolf PerlJam: perlhelp is a troll.
19:42 moritz rindolf: there people who say that about you too. Please moderate yourself with such statements.
19:42 masak perlhelp: hi! 还就不见了!
19:42 PerlJam rindolf: if you say so.  I tend to give the benefit of the doubt.
19:43 masak perlhelp: we love talking about the difference between 5 and 6 here. but mostly because we're very proud of what we have with 6.
19:44 masak perlhelp: I went to FOSDEM the other weekend. to my surprise I had forgotten many of the things that make Perl 6 great. the reaction of my audiences served to remind me.
19:44 masak perlhelp: aww. :)
19:46 masak er. I explained that badly.
19:46 masak I know very well what things about Perl 6 I think are nice.
19:47 masak but I had forgotten how much they matter (as improvements) if all you've seen is Perl 5.
19:47 perlhelp_ joined #perl6
19:47 perlhelp_ Hi!
19:47 perlhelp_ I was disconnected
19:47 perlhelp_ So i missed a few things
19:47 masak perlhelp_: welcome reconnected. :)
19:47 masak perlhelp_: there's a log.
19:47 perlhelp_ But people do not know perl 5 here :)
19:47 perlhelp_ oh ok.
19:47 erry joined #perl6
19:47 FROGGS masak: ya, I notice everyday how powerful p6 is, when hacken p5 or php
19:47 perlhelp_ bless your kind soul msak
19:47 masak Data::Dumper? yes, it's a method.
19:48 masak DateTime? yes, we have it built in.
19:48 masak Test::More? yes, but we call it "Test" :)
19:48 arnsholt XS? No, it's dead and buried
19:48 masak classes? yes, of course.
19:48 FROGGS (I added gist methods to my packages to have a bit of a perl 6 feeling :o)
19:48 masak we got roles, too.
19:48 masak oh, and signatures!
19:49 perlhelp_ I do not want to take up room here. But is thee someone with whom I can discuss perl 5 one-on-one in here?
19:49 masak and grammars! <3
19:49 geekosaur not really
19:49 masak perlhelp_: feel free to lighten your heart.
19:49 FROGGS grammars++
19:49 arnsholt masak: And a type system (which is optional)
19:49 perlhelp_ msask? What do you mean?
19:49 masak we even do multiple inheritance if you're into that thing.
19:50 masak perlhelp_: I mean we trust you not to misbehave and go off topic too much. if you do, we'll tell you.
19:50 masak perlhelp_: this is not a Perl 5 help channel.
19:50 masak most other topics are OK.
19:51 zby_home joined #perl6
19:51 perlhelp_ I need a lot of help with perl and am scared to derail the room.
19:51 masak if you need a lot of help, there are better channels for that.
19:51 moritz perlhelp_: then you should ask for a mentor, but you should ask in Perl 5 room for that
19:51 swarley alright, time to work on something that i'm not sure of and should have written down a few days ago
19:51 perlhelp_ No!  Perl-5 doe not like mentoring.
19:51 masak we do mostly Perl 6 here. many of us are proficient in Perl 5, but that's not why we are here.
19:51 moritz perlhelp_: or maybe get some professional training
19:52 perlhelp_ Maybe I can get a perl 6 mentor then?
19:52 masak sure.
19:52 moritz if you're actually willing to learn Perl 6, sure
19:52 swarley Perl6 would not exactly give you a drop in replacement of what you would use perl5 for though
19:52 swarley At this moment at least
19:52 masak depends on how many wheels you're willing to reinvent yourself :)
19:52 perlhelp_ yes, i would learn perl 6 but i ned omeone to help me
19:53 perlhelp_ i just cannot do it alone.
19:53 jms53 left #perl6
19:53 perlhelp_ i would reinvent with the help
19:53 swarley I would first ask why you are willing to switch from wanting to learn perl5 to perl6 so quickly? The languages are not similar in many aspects
19:53 masak perlhelp_: http://bash.org/?3936
19:53 spider-mario o/ moritz
19:53 spider-mario are you sure that qq["$file" --show-config] will work with paths that contain double quotes?
19:53 spider-mario would that not work instead: open my $CONFIG, '-|', $file, '--show-config'
19:54 spider-mario after all, open is capable of handling the quoting itself
19:54 TimToady r: enum Color (<red green purple> Z=> 1,2,4);
19:54 p6eval rakudo a3869a: OUTPUT«===SORRY!===␤Method 'returns' not found for invocant of class 'List'␤»
19:54 moritz spider-mario: but iirc the open PIPE form isn't implemeneted on windows :(
19:54 perlhelp_ masak: ok, so no mentoring?
19:54 spider-mario oh :(
19:54 masak perlhelp_: I didn't say that :)
19:54 TimToady masak: ^^
19:54 moritz spider-mario: I mean, open pipe LIST
19:54 masak perlhelp_: but you have to take personal responsibility for your learning.
19:54 spider-mario it’s always windows, isn’t it.
19:54 swarley If there was no mentoring for perl6 I would be in trouble ;)
19:54 masak perlhelp_: no-one can put stuff into your head.
19:55 * masak submits TimToady's rakudobug
19:55 perlhelp_ I know.  But I need help on where to start.  Saying "Google" means nothing to me.
19:55 perlhelp_ I don't know what to google.
19:55 swarley But your description of mentoring should be that someone can help you understand something that you tried to learn
19:55 perlhelp_ What to read.
19:55 masak perlhelp_: here's a 5-second mentoring. write a small program in Perl 6. make it either succeed or fail. report your progress here.
19:55 moritz perlhelp_: perl6.org has a pretty good list of links to stuff to read
19:55 swarley http://perl6.org/documentation/
19:55 masak perlhelp_: good luck. ask questions as soon as you get stuck.
19:56 diakopter perlhelp_: what other programming languages have you used?
19:56 perlhelp_ I tried that with perl-5 and they HATE me
19:56 swarley http://en.wikibooks.org/wiki/Perl_6_Programming
19:56 perlhelp_ C++
19:56 perlhelp_ Stata
19:56 perlhelp_ Matlab
19:56 masak perlhelp_: what, all of them HATE you?
19:56 moritz swarley: that's not very up-to-date or complete
19:56 perlhelp_ enough do
19:56 masak perlhelp_: man, what's the statistical likelihood of that? :P
19:57 perlhelp_ i tried to program and told me i was not trying
19:57 swarley moritz; I know, (as I've found) but it's what I've been starting with and it should be enough to get you on the right track shouldn't it?
19:57 perlhelp_ i wrote my own code and failed
19:57 perlhelp_ and said my failures were to rudimentary for their help
19:57 perlhelp_ that is my understanding
19:57 * masak hugs perlhelp_ :)
19:57 swarley Okay, I actually believe you on that
19:57 masak perlhelp_: seems you've had a rough trip.
19:57 perlhelp_ yes!!
19:57 perlhelp_ swarley: you  believe me?
19:57 perlhelp_ I spent 2 weeks doing nothing but lerning perl 5
19:57 masak perlhelp_: we're not like that here. this is the nice corner of the Intertubes.
19:57 swarley Especially from #perl or #c++/##c
19:58 perlhelp_ I'm really exhausted now
19:58 perlhelp_ My back is hurting
19:58 perlhelp_ I' haven't gone to the gym in a long time (or much else for that matter)
19:58 perlhelp_ and I'm kind of at wits ends now
19:58 swarley perlhelp_; what's your motivation for learning?
19:58 perlhelp_ I am working on a manuscript (nothing perl related)
19:58 perlhelp_ And I need to scrape some information off of the internet.
19:58 perlhelp_ I made a lot of headway in perl5
19:59 perlhelp_ But I do not know how to get further help
19:59 swarley What information?
19:59 perlhelp_ and perl 5 people are trying to kill me
19:59 diakopter I doubt that
19:59 perlhelp_ swarley: would you like to see an example?
19:59 swarley Sure
19:59 perlhelp_ diakopter: if you have seen the things they have said to me
19:59 perlhelp_ i almost had  a stroke
19:59 masak perlhelp_: some perl 5 people are very tired of helping people who do not ask questions in their favorite way.
19:59 swarley Perl5 isn't my "mother tongue" but I can read it fairly well
20:00 masak perlhelp_: their behavior often comes off as intolerance.
20:00 perlhelp_ Oh, you want the code or the website ?
20:00 swarley The code would help more in this instance
20:00 masak don't paste anything on-channel.
20:00 masak paste it to a pastebin service.
20:00 swarley www.pastebin.com
20:00 perlhelp_ i know.  i will do a pste
20:00 perlhelp_ i cn't do pastebin
20:00 perlhelp_ paste org ?
20:00 masak any is fine.
20:00 swarley Any paste service
20:00 perlhelp_ patsebin is failing on me for some reasons
20:01 moritz gist.github.com is fine
20:01 perlhelp_ blesss your kind souls
20:01 perlhelp_ ok give me a few seconds to do the pasting
20:01 masak r: say (<red green purple> Z=> 1,2,4).perl
20:01 p6eval rakudo a3869a: OUTPUT«("red" => 1, "green" => 2, "purple" => 4).list␤»
20:01 * geekosaur notes that a number of netnannies block pastebin.com because it has in the past been used to communicate things like stolen credit card numbers
20:01 perlhelp_ I'm sorry if my code does not look pretty
20:01 spider-mario http://paste.awesom.eu/ is good
20:01 perlhelp_ It is the best I can do
20:02 masak perlhelp_: no need to apologise for that.
20:02 geekosaur so I can't reach it from e.g. the local library unless I run tor
20:02 masak perlhelp_: we're all on a learning curve.
20:02 * swarley is learning Go as he implements 6model
20:02 perlhelp_ http://goo.gl/8zTHW
20:02 swarley probably not the smartest choice
20:03 perlhelp_ I'm having some trouble at the url line: 29-31
20:03 moritz that's not Perl 6 code.
20:03 perlhelp_ It only works right now if I manually put it in rather than have perl find it, and I need perl to find it since the address is so aribtrary
20:03 perlhelp_ no it's not.
20:03 perlhelp_ I said it was perl 5!
20:03 moritz perlhelp_: well, we were very clear. If you get Perl help here, it will be Perl 6 help.
20:04 swarley What are you trying to do exactly?
20:04 arnsholt timotimo: *prod?*
20:04 perlhelp_ can i show you the website then?
20:04 swarley This is for a web app?
20:04 perlhelp_ no, i don't think so
20:04 perlhelp_ it is to scrape information off of it
20:05 swarley Why is it that you need to scrape the information? Is there a large amount of data that you need?
20:05 perlhelp_ yes
20:05 spider-mario I love Web::Scraper for doing that
20:05 masak me too.
20:05 perlhelp_ I have a list of 8 million people.  For each one, I know their first name, last name, gender, race, year of birth (within +/- 2), and their U.S. state (or country) of birth.
20:06 swarley 8 million?
20:06 perlhelp_ yes :(
20:06 perlhelp_ maybe less
20:06 masak that's about one Sweden.
20:06 perlhelp_ maybe 4 million
20:06 swarley Well, first of all. You probably will not get that many from the site
20:06 perlhelp_ but a lot
20:06 masak oh. one Norway, then :)
20:06 arnsholt r: my $foo is export; say "alive"
20:06 p6eval rakudo a3869a: OUTPUT«alive␤»
20:06 perlhelp_ ok
20:06 perlhelp_ let's make it 1,000
20:06 arnsholt r: my $foo is export = 4; say "alive"
20:06 p6eval rakudo a3869a: OUTPUT«alive␤»
20:06 perlhelp_ can we got to the next step then?
20:06 diakopter perlhelp_: unfortunately, I don't think we can help you, since we can't do that kind of theng in Perl 6 right now
20:06 perlhelp_ huh?
20:07 swarley You may be able to get 1,000 before your IP is temporarily blocked
20:07 perlhelp_ perl 6 does not scrape?
20:07 perlhelp_ no, it will be ok
20:07 perlhelp_ I've already scraped a lot from the site.
20:07 perlhelp_ It is ok
20:07 perlhelp_ There is no TOS
20:07 perlhelp_ I told them.  They know
20:07 swarley If they know, they should be able to query the database for you.
20:07 perlhelp_ they don't.
20:08 swarley But you just said they know
20:08 masak diakopter: hm, what's missing? we have a LWP::Simple, but that's not the same as LWP::UserAgent, I guess.
20:08 perlhelp_ yes, i told them
20:08 timotimo arnsholt: what's up?
20:08 perlhelp_ can we go on to the next step?
20:09 diakopter masak: most of the things used in the p5 script - CGI, html parser, dom manipulation
20:09 moritz perlhelp_: translating to Perl 6? sure, your call
20:09 perlhelp_ oh that's fine
20:09 swarley I suppose I should make typeObject noew
20:09 swarley now*
20:09 perlhelp_ I meant i thoguht we were stuck on whethe I'm scraping 1,000 times, 5, times, ....
20:09 au perlhelp_: hi. if you are in contact with the admins of that site, have you considered signing up for an API key? https://familysearch.org/developers/docs/getting-started
20:09 perlhelp_ I haven't even described what I'm scraping
20:10 perlhelp_ Yes I signed up
20:10 masak diakopter: CGI doesn't seem to be used in the script.
20:10 perlhelp_ They are still developing it for this part of the site.
20:10 perlhelp_ It is not available yet.
20:10 masak diakopter: was probably a dead end from earlier development.
20:10 perlhelp_ That is how I've been in touch.
20:10 arnsholt timotimo: Are you on a UNIX or Windows?
20:10 perlhelp_ I will be availbale but is not yet
20:10 diakopter masak: ah, good idea
20:10 fgomez joined #perl6
20:11 masak diakopter: and I bet the HTML::DOM module could be worked around in a pinch.
20:12 timotimo i'm on a unix. i do have a windows nearby, but i don't know how hard it is to set it up to compile and run parrot, nqp and rakudo (i guess it's a huge pain)
20:12 arnsholt I'm on a UNIX, so that's not a problem
20:13 arnsholt It's just one less thing to worry about if I can ignore Windows portability for the time being
20:13 perlhelp_ diakopter: I might not be using the best tools.
20:13 perlhelp_ These are my gueses
20:13 perlhelp_ Should I tell you what I'm trying to do?
20:14 masak perlhelp_: it's pretty clear from the script what you're trying to do.
20:14 perlhelp_ oh ok.  i'm glad it is.
20:14 perlhelp_ the script is incomplete though
20:14 perlhelp_ so i did not think it would be.
20:14 masak aye, it's incomplete.
20:14 perlhelp_ it's nearly complete
20:14 masak nice.
20:15 masak best of luck :)
20:15 perlhelp_ there are a few small things I needed to add that I did not yet figure out either.
20:15 perlhelp_ msask: huh?
20:15 masak perlhelp_: this is not a Perl 5 help channel.
20:15 perlhelp_ well, if this can be done in perl 6, i'm all ears!!
20:15 perlhelp_ i'm read to learn perl 6 if so
20:15 masak ;)
20:16 erry left #perl6
20:16 masak general question to channel: can our LWP::Simple do what perlhelp_ is doing with LWP::UserAgent?
20:18 arnsholt star: use NativeCall :EXPORT; say "alive?"
20:18 p6eval star 2012.12: OUTPUT«===SORRY!===␤Error while importing from 'NativeCall': no such tag 'EXPORT'␤»
20:20 arnsholt Oh. I'm going to have to put all of the things into a great big module, aren't I
20:21 masak either way, I guess there's room for LWP::UserAgent in the Perl 6 ecosystem ;)
20:21 perlhelp_ masak: what do you mean?
20:21 masak perlhelp_: well, we don't load CPAN modules from Perl 6 yet.
20:21 masak perlhelp_: so when we want something like LWP::UserAgent, we tend to implement it ourselves.
20:22 perlhelp_ from scratch?
20:22 perlhelp_ WOW
20:22 masak perlhelp_: (this is temporary. with time, we would like to be able to load Perl 5 modules)
20:22 masak perlhelp_: but yeah, this is why no-one stands up and tells you "yes, you can write this script in Perl 6".
20:22 perlhelp_ oh you mean you have to write them?
20:23 perlhelp_ i'm not following
20:23 masak long term, we would like a Perl 5-Perl 6 bridge.
20:23 masak so Perl 6 can load Perl 5 modules.
20:23 masak now, that won't work for all Perl 5 modules.
20:23 benabik joined #perl6
20:23 masak does LWP::UserAgent do XS?
20:23 masak if not, it should be possible to load it in the future.
20:24 masak until then, you'd have to do something in Perl 6 to compensate for the lack of LWP::UserAgent.
20:24 masak perlhelp_: that's why diakopter was less-than-optimistic about you being able to (easily) port that script to Perl 6.
20:24 perlhelp_ oh i see
20:26 [Coke] masak: I wonder if #perl6 could set up a #p5 consultancy to help fund sixdev. :)
20:26 masak [Coke]: hah! :D
20:26 brrt joined #perl6
20:27 masak [Coke]: "Give me your tired, your poor, Your huddled masses yearning to breathe free, The wretched refuse of your teeming shore. Send these, the homeless, tempest-tost to me, I lift my lamp beside the golden door!"
20:27 masak people like perlhelp_++
20:29 perlhelp_ yes, i am an immigrant :)
20:29 masak [Coke]: I'm afraid that if I tell perlhelp_ "can you stick around for a few months? I'm just gonna build LWP::UserAgent", he'll grow impatient and leave. what should I do? :)
20:29 perlhelp_ is that what you mean?
20:29 perlhelp_ i would stick around.
20:29 masak perlhelp_: it was a kind of metaphor.
20:29 perlhelp_ but i think there will be a lot of modules needed
20:29 perlhelp_ it might be many months after thta
20:30 masak perlhelp_: by self-selection, the people on this channel are a patient bunch.
20:31 moritz nah, if I were patient, I wouldn't hack on perl 6, but wait until others have done it.
20:31 uvtc joined #perl6
20:32 perlhelp_ yes i can see!!
20:32 perlhelp_ waiting months to write a module
20:32 perlhelp_ i can see why perl 5 people are bi-otches
20:34 uvtc nwc10: you mentioned #perl just a bit ago. (Hm. I like that word, "justabitigo"). So, #perl at irc.perl.org is the social channel. What's the difference between #perl-help on irc.perl.org and #perl at freenode?
20:34 perlhelp_ #perl is a room of bi-otches
20:34 uvtc perlhelp_: please be kind :)
20:34 swarley Keep in mind, many of us are in that room
20:35 nwc10 uvtc: I don't know what if any differences there are betewen #perl-help on irc.perl.org and #perl on freenode. I've used neither
20:35 nwc10 but I'm told that bother are "help" channels
20:35 uvtc nwc10: thanks.
20:36 masak (then there's #metallica...) :P
20:36 swarley lolol
20:36 brrt left #perl6
20:36 FROGGS I have only one experience with #perl-help from a week ago, and I got help from within 15minutes, and the guy/girl was nice
20:36 perlhelp_ FROGGS: is that the same as #perl ?
20:37 FROGGS no, I meant #perl-help on irc.perl.org
20:38 perlhelp_ oh, i can't get to that site for some reason
20:38 perlhelp_ i go to irc.perl.org and i got a weird pic of an animal
20:38 diakopter it's not a website
20:39 diakopter you have to use an IRC client
20:39 diakopter there isn't a web inyerface like the one you're using for freenode
20:39 diakopter *interface
20:39 diakopter inyerface too
20:39 perlhelp_ xchat?
20:39 perlhelp_ i tried that but failed
20:40 swarley ?
20:40 swarley How did you fail?
20:42 geekosaur xchat doesn't come with irc.perl.org in its default network list; you would need to add it
20:42 perlhelp_ it did not connect
20:42 perlhelp_ i trid to add it
20:42 perlhelp_ it said it did not work
20:43 swarley /newserver irc.perl.org 6667
20:43 geekosaur just worked for me, fwiw
20:44 japhb_ joined #perl6
20:49 swarley I wonder where I should go from here
21:01 kivutar joined #perl6
21:06 shinobicl left #perl6
21:08 perlhelp_ I GOT Some code to work!
21:08 perlhelp_ YAY!!!
21:08 perlhelp_ Take that #perl!!
21:08 perlhelp_ They said I cannot do anything!!
21:10 perlhelp_ SUCK IT #perl :P
21:10 perlhelp_ :)
21:11 thou joined #perl6
21:12 PerlJam perlhelp_: #perl6 is not usually that impolite.  We give hugs to those that need them.
21:13 perlhelp_ i know.  Perl5 tried to kill me this weekend :(
21:13 perlhelp_ #perl6 is amazing!!!!!!!!!!
21:13 perlhelp_ #perl5 sucks!!!!!!!!!
21:13 masak perlhelp_: "you need to learn how to figure out stuff yourself.." :)
21:13 jnthn masak: How does one do that?
21:13 perlhelp_ what's wrong?  some of my code now compiles?
21:13 masak :P
21:14 masak perlhelp_: no, I was just reminding you of the link I sent you earlier today. :)
21:14 PerlJam .oO( jnthn:  Just tell me! )
21:14 masak perlhelp_: it'd still be a good idea for you to find a place where people actually answer Perl 5 questions, I think.
21:15 perlhelp_ yes it would, but i cannot get to that plce
21:15 perlhelp_ #perl5 banned me
21:15 PerlJam There's a #perl5?
21:16 PerlJam or did you mean #perl?
21:17 uvtc perlhelp_: if you want help with Perl 5 specifically, check out either #perl at freenode, or #perl-help at irc.perl.org.
21:17 PerlJam perlhelp_: there's also #perlhelp on efnet
21:17 PerlJam (if anyone still goes to efnet)
21:17 uvtc left #perl6
21:17 perlhelp_ #perl5 here banned me
21:18 TimToady um, there's only one person on that channel
21:18 sorear masak: nah, I just saw a link to it and I couldn't remember what was on that site so I checked
21:19 fgomez joined #perl6
21:19 swarley finished repr.go \o/
21:19 masak sorear: there were many wondrous things on that site. maybe I can rescue some of the content.
21:19 masak swarley++
21:20 swarley And, changes are pushed
21:21 jnthn swarley: öl^Wurl?
21:21 jnthn oh, guessed it :)
21:22 jnthn hm, how to read go code... :)
21:22 perlhelp_ so where else is perl5 help?     /newserver irc.perl.org 6667  ????
21:22 perlhelp_ how do i get there?
21:22 bruges joined #perl6
21:23 swarley It's a little difficult at first :)
21:23 jnthn swarley: On the "// Flattening methods" bit - those exist because on the JVM we handle flattened attributes by including them in a generated JVM type.
21:23 swarley Ah. I only threw them in because I wasn't sure what they were for yet
21:23 jnthn swarley: I don't know go well enough to be sure, but I'd guess it's closer to C than JVM.
21:23 swarley I'm going to tidy up after I get through the code
21:24 jnthn In the C implementation on Parrot we don't do such code-gen; the REPRs are really about computing struct layouts.
21:24 jnthn I dunno what Go does/doesn't let you do in that regard, I'm afraid, so it's hard to give good hints.
21:24 swarley go compiles to machine code, so yeah it's a little more like C
21:24 swarley Well, what would you define to be a flattened attribute?
21:25 jnthn On the third hand, a 6model VM will probably have the REPRs participate in JIT, so I guess they get their code-gen role after all... :)
21:25 swarley because Go doesn't allow for pointer manipulation so that it's a lot harder to get a segfault
21:25 jnthn swarley: Flattened in as in
21:25 jnthn class A { has $!x; has int $!y; }
21:25 jnthn The $!x here is a reference to another object
21:25 swarley Ah
21:26 jnthn The int, however, is not; it's a native interger living in the object body.
21:26 swarley So you're collapsing the reference?
21:26 jnthn Yeah.
21:26 jnthn Well, you could think of it that way :)
21:26 swarley Oh, well I think most of the code I'm going to generate is going to use pointers instead of the actual reference for speed
21:26 swarley err, actual object*
21:27 TimToady everything is really a native, but if you don't specify a native type explicitly, you get an implicit native pointer to a P6 object
21:27 jnthn And there's another way to look at it. :)
21:27 swarley oh, hm.
21:28 swarley I think I'll remove the flattening. Most of the code should just be moving around pointers anyway
21:32 swarley private static HashMap<String, Integer> reprIdMap = new HashMap<String, Integer>();
21:33 swarley I'm assuming it's a map with string keys to int values?
21:36 jnthn Correct.
21:36 swarley Wonderful, I really should learn more Java
21:36 jnthn Looks like you're translating the repr registry...
21:36 jnthn At a language level, there's not a lot of Java to learn. :)
21:38 swarley Yeah, I just need to learn more of the library
21:39 skids ⱺ.ⱺ http://joeyh.name/blog/entry/unicode_eye_chart/
21:40 jnthn Well, the library is a bigger task ;)
21:42 perlhelp_ swarley: How do I get to:    /newserver irc.perl.org 6667  ????
21:42 swarley Open xchat, close the server list.
21:46 perlhelp_ close the sever list?!?
21:46 Ayiko hi #perl6!
21:46 geekosaur you're specifying a server manually so you don't want the server list, just the main window
21:47 Ayiko .u ಅ
21:47 phenny U+0C85 KANNADA LETTER A (ಅ)
21:47 Ayiko skids: the eye chart is missing this besides the beer mug >_>
21:51 swarley jnthn; It's not really my place to suggest this, but wouldn't it be a bit more efficient to directly use REPR.ID and avoid recomputing the size of reprs here? https://github.com/jnthn/nqp-jvm-prep/blob/master/src/org/perl6/nqp/sixmodel/REPRRegistry.java#L27
21:52 jnthn swarley: probably, but given that code is only called from setup() and that's only ever called once at startup... :)
21:52 swarley Oh alright
21:53 jnthn It's kinda bad style though. jnthn--
22:02 dalek nqp/rx-portability: 029bf5c | jnthn++ | src/6model/reprs/NFA.c:
22:02 dalek nqp/rx-portability: NFA REPR serialize/deserialize.
22:02 dalek nqp/rx-portability: review: https://github.com/perl6/nqp/commit/029bf5ca06
22:02 dalek nqp/rx-portability: c56d322 | jnthn++ | src/QRegex/Cursor.nqp:
22:02 dalek nqp/rx-portability: Start !cursor_start refactor.
22:02 dalek nqp/rx-portability:
22:02 dalek nqp/rx-portability: Currently it is the only thing in the NQP or Rakudo codebase that uses
22:02 dalek nqp/rx-portability: Parrot's multiple return values feature. This splits it up into two
22:02 dalek nqp/rx-portability: routines, one for things that really only want the cursor anyway (and
22:02 dalek nqp/rx-portability: so this will be an efficiency win for them) and one for generated code
22:02 dalek nqp/rx-portability: that wants all the things (which we can still do without allocating).
22:02 dalek nqp/rx-portability: review: https://github.com/perl6/nqp/commit/c56d32262d
22:02 dalek nqp/rx-portability: c7f057b | jnthn++ | src/QRegex/Cursor.nqp:
22:02 dalek nqp/rx-portability: First batch of !cursor_start updates.
22:02 dalek nqp/rx-portability: review: https://github.com/perl6/nqp/commit/c7f057b1ed
22:04 slavik1 joined #perl6
22:05 jnthn oops, the commit message is a lie...
22:05 * jnthn figures he'll do the rest and then check if there's actually any noticable difference before worrying...
22:13 swarley reprRegistry.go is done :) (as done as it can be at the moment)
22:13 [Coke] GSOC 2013 announced.
22:14 [Coke] we should try to coordinate with the p5 folks as soon as possible.
22:18 perlhelp_ Coke: why would you wnt to work wth p5?
22:18 perlhelp_ They are not as cool as p6!
22:18 perlhelp_ I'm talking about #perl5
22:18 perlhelp_ #perl6 is the best!!!
22:18 FROGGS perlhelp_: calm down
22:19 MikeFair__ joined #perl6
22:19 perlhelp_ Ok.  I just wnt to expres my love for decent humans who are also developers
22:19 perlhelp_ I know I am getting obnoxious
22:19 perlhelp_ It just seems so rare to find developers who are humans!
22:20 Juerd We need human developers because robots write awful code.e
22:20 Juerd s/e$//
22:21 masak perlhelp_: I suspect many of the other channels are abrasive because the signal-to-noise ratio there is much lower.
22:22 masak it's like in a (physical) room where the only way to communicate is to shout.
22:22 perlhelp_ perl has no signal nor noise
22:22 perlhelp_ they just want to plonk and sleep
22:23 * [Coke] notes that perlhelp_++'s SN ratio is getting lower.
22:23 [Coke] -> homeish
22:24 masak "plonk and sleep" sounds like a good name for a punk band.
22:24 swarley once again, try to remember that practically everyone in here is also in #perl
22:25 * geekosaur doesn't generally hang there, should probably look in at least
22:25 swarley http://www.youtube.com/watch?v=aLwaPP9cxT4
22:25 masak 'night, #perl6
22:25 swarley I find that very interesting
22:25 swarley and unrelated
22:25 swarley night o/
22:26 perlhelp_ swarley: are you here?
22:26 swarley Am I there? No, not usually
22:27 geekosaur ...such timing :/
22:29 perlhelp_ is there nybody in perl #5 with whom i cn please speak?
22:29 perlhelp_ like Bingo?
22:29 geekosaur is there some reason you need to ask in #perl6?
22:31 perlhelp_ yes, i am not allowed in perl#5.  i want to know when i may be allowed back.
22:31 perlhelp_ i am not getting ny clear info on that
22:33 perlhelp_ i just want to know what i can do to be allowed back.  thta is all.  i will do it plus mor.
22:36 swarley hm. Not sure what to do now
22:37 Tene perlhelp_: you keep talking about #perl5, and haven't responded to people who have mentioned #perl
22:37 perlhelp_ huh?
22:37 Tene Two other IRC channels for perl 5 help have been mentioned too.
22:37 perlhelp_ yes, i am confused still on how to get there
22:38 swarley https://help.ubuntu.com/community/XChatHowto
22:39 bluescreen10 joined #perl6
22:43 * swarley ponders what he should do now
22:49 c1sung_ joined #perl6
22:54 mathw joined #perl6
22:55 pmurias joined #perl6
22:56 slavik1 joined #perl6
22:56 Gothmog_ joined #perl6
23:07 swarley Starting on P6Opaque :D I'm more afraid than I look
23:10 thou joined #perl6
23:10 jnthn It's less scary than it used to be...
23:12 swarley oh boy
23:13 dalek nqp/rx-portability: b555f7f | jnthn++ | src/HLL/Grammar.pm:
23:13 dalek nqp/rx-portability: Update HLL::Grammar calls to !cursor_start.
23:13 dalek nqp/rx-portability: review: https://github.com/perl6/nqp/commit/b555f7ff71
23:13 dalek nqp/rx-portability: ff1a336 | jnthn++ | src/QAST/Compiler.nqp:
23:13 dalek nqp/rx-portability: Update call to !cursor_start in code-gen.
23:13 dalek nqp/rx-portability:
23:13 dalek nqp/rx-portability: No measurable performane change.
23:13 dalek nqp/rx-portability: review: https://github.com/perl6/nqp/commit/ff1a336787
23:13 dalek nqp/rx-portability: 67eb623 | jnthn++ | src/stage0/ (9 files):
23:13 dalek nqp/rx-portability: Update bootstrap.
23:13 dalek nqp/rx-portability: review: https://github.com/perl6/nqp/commit/67eb623771
23:13 dalek nqp/rx-portability: a7f47d9 | jnthn++ | src/QRegex/Cursor.nqp:
23:13 dalek nqp/rx-portability: Kill !cursor_start, which used multiple return.
23:13 dalek nqp/rx-portability: review: https://github.com/perl6/nqp/commit/a7f47d9b8e
23:13 dalek nqp/rx-portability: 44f22ef | jnthn++ | src/ops/nqp.ops:
23:13 dalek nqp/rx-portability: Kill dead code.
23:13 dalek nqp/rx-portability: review: https://github.com/perl6/nqp/commit/44f22ef330
23:14 swarley \o/
23:21 japhb_ jnthn++ # Simplifying the porters' jobs
23:21 japhb_ Self-interested altruism FTW
23:25 benabik joined #perl6
23:27 jnthn :)
23:44 dalek Heuristic branch merge: pushed 16 commits to nqp by jnthn
23:44 diakopter jnthn++ thanks :)
23:45 swarley I should probably do more of repr
23:45 dalek rakudo/nom: 7e22a79 | jnthn++ | / (4 files):
23:45 dalek rakudo/nom: Chase !cursor_start changes.
23:45 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/7e22a79da9
23:45 jnthn swarley: P6opaque can actually be stubbed for the most part until you want to implement attributes.
23:46 jnthn swarley: And, moreover, you won't have any way to test it until you've got the KnowHOW bootstrap done.
23:46 jnthn Though maybe you did that already... :)
23:47 swarley Yeah, I'm pretty much stubbing the REPR right now
23:47 swarley I'm just a little exhausted, exams coming up fast
23:48 MikeFair joined #perl6
23:55 skids joined #perl6

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

Perl 6 | Reference Documentation | Rakudo