Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2010-01-04

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:02 * diakopter hears a pin drop
00:02 mberends perl6: ping
00:03 p6eval elf 29447: Undefined subroutine &GLOBAL::ping called at (eval 122) line 3.␤ at ./elf_h line 5881␤
00:03 p6eval ..rakudo db84bc: Could not find non-existent sub ping␤in Main (file src/gen_setting.pm, line 324)␤
00:03 p6eval ..pugs: *** No such subroutine: "&ping"␤    at /tmp/gBhITzikJ4 line 1, column 1 - line 2, column 1␤
00:03 diakopter "everytime you hear a pin drop, an angel^W announcement of a skunkworks project is forthcoming/imminent"
00:04 mberends that'll be the Rekudo ng branch about to become master
00:05 mberends *Rakudo
00:05 ashizawa joined #perl6
00:10 mberends pmurias: when building mildew, I missed a README. You once questioned the usefulness of a README file somewhere, was that for Mildew or for JavaScript::V8? I'm quite happy to create a mildew README if you're not opposed.
00:13 pmurias mberends: having a README for Mildew would be great
00:13 pmurias mberends: i questioned the usefulness of README for CPAN module which are installed using the CPAN client
00:14 mberends pmurias: ok, it will start small, we can expand it gradually
00:32 lumi joined #perl6
00:36 silug joined #perl6
00:41 orafu joined #perl6
00:46 xomas_ joined #perl6
00:56 ewilhelm joined #perl6
01:49 simcop2387 joined #perl6
02:00 agentzh joined #perl6
02:06 lumi joined #perl6
02:40 ewilhelm anybody have input on what exactly PAUSE/CPAN needs to do to support the Perl 6 module system/semantics?
02:42 diakopter ewilhelm: it's a big question to some, a smaller question to others.
02:42 diakopter it depends a lot on the implementation
02:42 ewilhelm it's currently being hashed-out by a bunch of the people who could do something about it but aren't sure what p6 needs
02:43 ewilhelm http://www.nntp.perl.org/group/perl.cpan.workers/2010/01/msg639.html
02:44 diakopter ewilhelm: parrot has its own dependency system/installer
02:44 diakopter http://gitorious.org/parrot-plumage
02:45 ewilhelm on the PAUSE/CPAN side of things, we're only talking about somewhere to upload tarballs and some kind of indexing ala 02packages.details.txt.gz
02:46 diakopter strictly speaking, all of these fields should be present in an index: http://perlcabal.org/syn/S11.html#Versioning
02:47 diakopter the "long name" there defines the namespace
02:48 diakopter the "cpan:" authority is reserved for (of course), cpan
02:48 ewilhelm so, just name+version+author, right?
02:49 diakopter name+version+auth (where "auth" is intentionally ambiguous between "authority" and "author")
02:50 ewilhelm but that's actually just the long index (equivalent to `ls -R`)
02:50 ewilhelm doesn't it needs some kind of short index for "the default for Foo"?
02:51 diakopter that's for the dependency/installer system to work out
02:51 ewilhelm which is more equivalent to the current 02packages.details.txt.gz -- namespaces mapped to author+version
02:51 ewilhelm Really?  There's nothing in the spec for that?
02:52 diakopter yes, it defines the compile-time behavior, but not the install-time
02:53 ewilhelm so what is the wildcard resolution order for "use Dog:auth(Any):ver(Any);" and who decides that?
02:55 diakopter install-time, it's presumably dependent on the installer/packager settings
02:55 ewilhelm "the version lineage that was created first wins" -- by date?  "officially abandoned lineage" -- what's official?
02:56 diakopter ctime on disk, I guess
02:56 ewilhelm wow that's going to give redhat enough rope to hang us all
02:56 diakopter sure, the packagers/distributors get to define their own policies
02:57 diakopter what it does is set cpan's repo (and everyone's github/gitorious/googlecode/sourceforge/codeplex repos) at the same level as the system distributors/packagers
02:58 ewilhelm so does PAUSE need to keep track of namespace ownership at all?
02:58 Tene Man, I haven't seen pm online in ages.  Hope he's having good holidays.  :)
02:58 diakopter ewilhelm: yes, because those represent who owns that package in the "cpan" auth
02:59 diakopter the actual author's name/id isn't in the authority
02:59 ewilhelm ok, so if you ask "a repository" for "the" Dog module, whoever owns that namespace *in that repo* is the one you get?
02:59 diakopter so, an "unauthorized release" would have the auth "cpan_EWILHELM"
03:00 diakopter or something
03:00 diakopter yeah
03:00 diakopter anyone else _please_ correct me where I'm wrong... I haven't seen these issues talked through here for years...
03:01 diakopter ewilhelm: does that seem workable to you?
03:02 ewilhelm I think so.
03:03 diakopter oh wait
03:03 diakopter after reading S11 again...
03:03 diakopter I was mostly right
03:03 ewilhelm sounds like we need 1) upload tool which detects p6 tarball + a place to put it, 2) a p6 "short" index of ownership per namespace (ala 02packages), and 3) a p6 "big" index of everything available
03:04 diakopter it would be auth<cpan:EWILHELM> for every release including unauthorized ones
03:04 diakopter but auth<cpan:$$some_flag_meaning_namespace_owner$$> for the officially owned release
03:05 diakopter I'm curious; why wouldn't the current system "just work"?
03:05 ewilhelm sure.  That's still compatible with 1,2,3 --^, right?
03:06 ewilhelm because ownership per namespace is p5 ownership
03:06 ewilhelm so rjbs owns Config::INI and thus Ovid can't upload his p6 tarball until we settle this
03:06 diakopter hm
03:07 cognominal joined #perl6
03:07 diakopter my first reaction is not to allow the duplicate names
03:07 ewilhelm but, it's looking very much like "just" making an empty copy of the current system and filling it in with p6 data
03:08 diakopter (in order to avoid branching the whole system)
03:08 ewilhelm well, that's going to make lots of people nonhappy
03:08 diakopter well, there aren't going to be distributions that work in *both*
03:09 ewilhelm how is Perl 6 supposed to handle the case of loading a p5/ruby/whatever module where there's multiple instances of the same name installed?
03:09 diakopter so there's either an implied "P6::" in front of that whole namespace, or you have to fork the whole system
03:09 ewilhelm we can't prevent naming overlap in the rest of the language world anyway
03:10 Tene ewilhelm: imported symbols are lexically scoped.
03:10 diakopter to import from other languages requires another qualifier
03:10 Tene So if I import Foo:from<ruby>; then Foo in that lexical scope refers to ruby's Foo.
03:11 ewilhelm Tene, no, oh :from<ruby> is news to me
03:11 Tene Loading p5 modules is :from<perl5>, iirc
03:11 diakopter same from perl5, yes
03:11 ewilhelm diakopter, there you have it
03:12 ewilhelm we have an implied sort of "P5::"/"P6::" everywhere already
03:12 diakopter yes, but, if you placed an invisible P6_flag_something:: in front of all the Perl6 uploads, you could use the current system without branching/duplicating everything
03:13 ewilhelm it's not really "fork the whole system" on the PAUSE pov, more like analogous to "mkdir p6; chdir p6; and start another daemon"
03:13 diakopter (obviously the Perl 6 tools would know not to display that prefix)
03:13 diakopter yeah, but I'm referring to all the other web tools
03:14 ewilhelm that is, the upload tool will (probably) be the same page and just detect p6-ness inside the tarball, then stick it in the p6/ directory on the mirrors and index it accordingly
03:14 ewilhelm but it will be a whole new index
03:15 diakopter if I were you, I'd keep it in the same one with a hidden prefix, since it would require modifying a lot fewer toolchains, imho
03:16 ewilhelm not sure about that
03:16 diakopter b/c the existing cpan could be used to resolve/install dependencies.
03:16 ewilhelm you can't use the CPAN.pm to install p6 code anyway
03:16 diakopter sure you could
03:17 ewilhelm You've read the code?
03:18 ewilhelm so without :from<perl5>, you get "module not found" if there's a p5 Foo.pm and no p6 one?
03:18 agentzh1 joined #perl6
03:18 diakopter "the code"?  I don't see why a perl6 code distribution couldn't use a p6 Makefile.PL
03:18 diakopter p6 Makefile.PL, I mean
03:18 Tene Hmm.  I *do* remember that there's part of the spec about how to distinguish between p5 and p6...
03:19 diakopter omg.
03:19 diakopter I can't type   p5 Makefile.PL
03:19 diakopter ewilhelm: answer to your last question, seems right to me, yes
03:21 diakopter my attitude/goal would be to try to reuse (and continue to use) as much of the existing toolchain as possible
03:21 diakopter but I'm sure others have different priorities.
03:21 ewilhelm re the index: putting p6 tarballs in 02packages is definitely going to explode for people running old CPAN.pm's
03:22 ewilhelm but there's no reason a new CPAN.pm couldn't fetch a second index file for p6 and DTRT with it
03:22 diakopter some folks even talk about having a Perl 6 implementation, an installation of which is not dependent on a Perl 5 installation.  I can't imagine that ever happening, so I focus on reuse.
03:22 diakopter ah
03:26 diakopter fer instance, I imagine a world where all the Perl 6 implementations share a common Perl 5/xs package such that "use v6" in a #!/usr/bin/perl file does Devel::Declare-ish magic and swaps over to a loadable Perl 6
03:32 diakopter maybe I be dreaming.
03:37 diakopter ewilhelm: it is spec'ed that any (well, nearly all) of the pure-Perl distributions on CPAN be accessible from a Perl 6 implementation via :from<perl5>  and use v5;
03:37 diakopter (probably there are some whose Perl usage won't be parseable by STD)
03:37 diakopter *reasonably* parseable, I mean.
03:44 ewilhelm I think that still means there's one index on the mirror for p5 and another for p6
03:44 diakopter ewilhelm: ah
03:45 diakopter I can see how that is functionally equivalent to the hidden prefix I was proposing. :)
03:45 diakopter (now that I read that entire thread you linked)
03:46 diakopter ewilhelm: (with the added benefit that it partitions the index, as adamk said)
03:48 ewilhelm the nntp page is currently missing my last message in reply to adam
03:49 diakopter ewilhelm: oh
03:50 ewilhelm ah, there it is:  http://www.nntp.perl.org/group/perl.cpan.workers/2010/01/msg657.html
03:57 wolf2k_ubuntu joined #perl6
04:43 synth joined #perl6
04:53 douglashunter joined #perl6
04:54 soupdragon joined #perl6
05:08 pugs_svn r29448 | diakopter++ | [sprixel] finish up geitherls and gbothls, afaict.
05:16 dalek rakudo/ng: c2ba04d | tene++ | t/spectest.data:
05:16 dalek rakudo/ng: enable fail tests
05:16 dalek rakudo/ng: review: http://github.com/rakudo/rakudo/commit/c2ba04da02e86e3fd178861f3329c9fb676f1e61
05:17 Tene Oh, this is classy.  Some part of "say" checks defined on an exception before stringifying it.
05:18 diakopter heh
05:18 diakopter makes me smile
05:18 diakopter Tene: where'd everybody go, btw...?  pmichaud, jnthn, moritz_, masak, etc... it's eerily quiet.
05:19 Tene diakopter: I've been wondering that too.  I assume just gone for holidays... I hope.
05:19 diakopter that or conspiratorially sekrit hackathon; either would be ok with me ;)
05:20 Tene Oh, I could throw an exception when it gets set to find out where!
05:22 Tene ... it had better not be doing what I think it's doing...
05:23 pugs_svn r29449 | diakopter++ | [sprixel] 18 lines of explanatory comments/notes.  to myself, I guess. :)
05:23 lumi joined #perl6
05:23 Tene Oh, no, it's not a vivify.  Good.
05:29 pugs_svn r29450 | tene++ | Use an explicit stringify in fail.t
05:32 nihiliad left #perl6
05:52 Tene Okay, I think I just about have 'use fatal' working.
06:17 kst joined #perl6
06:23 pugs_svn r29451 | tene++ | Remove {} from a string and remove fudge marks for fail.t
06:24 Tene We're now passing fail.t unfudged.
06:28 ssm joined #perl6
06:28 Tene {} in a double-quoted string fails badly...
06:29 Tene ng: my $x = "try{}";
06:29 p6eval ng c2ba04: Null PMC access in get_string()␤current instr.: '_block14' pc 29 (EVAL_1:0)␤
06:29 Tene someone should rakudobug that.
06:30 dalek rakudo/ng: aebe139 | tene++ | src/builtins/ (2 files):
06:30 dalek rakudo/ng: Add support for &fail throwing when $*FATAL is set
06:30 dalek rakudo/ng: review: http://github.com/rakudo/rakudo/commit/aebe139297f3542f3d2326b577999c7fd90eb851
06:30 dalek rakudo/ng: 4456d81 | tene++ | src/Perl6/Actions.pm:
06:30 dalek rakudo/ng: Implement "use fatal" a bit hackishly
06:30 dalek rakudo/ng: review: http://github.com/rakudo/rakudo/commit/4456d81dc9c1551a867d58d22c011c977298e537
06:36 lumi joined #perl6
06:37 pugs_svn r29452 | diakopter++ | [sprixel] begin implementing geitherrs, and take ++mberends suggestion to use terminology "deterministic" and "nondeterministic" instead of "stateless" and "stateful".
06:37 Tene Okay, someone should add more tests to fail.t
06:37 diakopter someone has a lot of work to do
06:38 diakopter (good thing I'm not someone!)
06:40 mberends doth someone lurketh here?
06:41 hejki *lurkeths*
06:43 mberends hark! there lurk-ed someone!
06:43 meppl joined #perl6
06:45 * Tene de-lurks
06:45 Tene Okay, fine, I'll read the spec a bit more tomorrow and add more tests.
06:46 * Tene afk # sleeping
06:50 mberends I actually prefer "impure" to "nondeterministic" ;-) # badger #2
06:52 diakopter but.. they're not impure
06:52 diakopter they can have side-effects
06:53 diakopter but are deterministic on their inputs
06:53 mberends arghh, you're right :)
06:53 diakopter the deterministic ones, I mean.  are deterministic on their inputs.  but can have side-effects.
06:55 mberends and are inlinable, as long as they're also nonrecursive
06:55 mberends side-effects are not a blocker, nor is state
06:56 diakopter a deterministic one can't be recursive.  if a rule refers to itself by name, but has no branches/alternations/backtracking, there is no exit from the recursion; it's infinite.
06:57 diakopter it's tricky to keep all the phases distinct in one's brain.
06:57 diakopter there are 4 separate compile-times here :/
07:00 mberends it's hard for me to understand most of this, let alone nit-pick it. I'll just lurk on this for a while :)
07:10 kaare joined #perl6
07:35 Baggio_ joined #perl6
07:40 Su-Shee joined #perl6
07:43 Su-Shee good morning :)
07:51 mberends Su-Shee: still navigating the German vocabulary in http://www.schreibsturm.org/antiverpeil.html (stubbornly refusing to take a machine translation) and it can be summarized in 3 English words: J(F)DI
07:57 kst joined #perl6
08:02 Su-Shee *hihi* essentially, yes :)
08:02 Su-Shee the german vocabulary is heavy slang and relatively long sentences, not very good to translate.
08:03 * diakopter tries it
08:03 diakopter copied, printed, stapled to the wall's placed under pillows or otherwise reused
08:04 diakopter Disclaimer: I am not couch and so I take no patients
08:05 mberends the endless stream of tax returns is familiar :-(
08:07 hejki the endless stream of mandatory taxes to be paid is familiar too
08:08 Su-Shee the sentence would be "I'm not a shrink and I don't take clients" :)
08:08 Su-Shee we just call it "couch" due to Freud's famous one ;)
08:09 renormalist|work joined #perl6
08:18 ashizawa_ joined #perl6
08:19 mberends is there any Perlito documentation? http://wiki.perlito.org/ is empty :(
08:30 xomas_ joined #perl6
08:38 nsh joined #perl6
08:40 nsh joined #perl6
08:47 hejki rakudo: sub foo (Int $x > 2) { say $x } foo(3);
08:47 p6eval rakudo db84bc: Unable to parse multisig; couldn't find final ')' at line 2, near "> 2) { say"␤in Main (file <unknown>, line <unknown>)␤
08:47 hejki rakudo: sub foo (Int $x where > 2) { say $x } foo(3);
08:47 p6eval rakudo db84bc: Unable to parse multisig; couldn't find final ')' at line 2, near "where > 2)"␤in Main (file <unknown>, line <unknown>)␤
08:49 hejki how-to achieve following (warning: perl5 code): sub foo { my $x = shift; return if ($x < 2); } ?
08:49 hejki sub foo (Int $x where 3..Inf) { say $x } foo(4);
08:49 hejki rakudo: sub foo (Int $x where 3..Inf) { say $x } foo(4);
08:49 p6eval rakudo db84bc: Confused at line 2, near "foo(4);"␤in Main (file <unknown>, line <unknown>)␤
08:50 hejki rakudo: sub foo (Int $x where 3..Inf) { say $x }; foo(4);
08:50 p6eval rakudo db84bc: 4␤
08:50 hejki rakudo: sub foo (Int $x where > 2) { say $x }; foo(3);
08:50 p6eval rakudo db84bc: Unable to parse multisig; couldn't find final ')' at line 2, near "where > 2)"␤in Main (file <unknown>, line <unknown>)␤
08:54 spinclad rakudo: sub foo (Int $x where * > 2) { say $x }; foo(3);
08:54 p6eval rakudo db84bc: Unable to parse multisig; couldn't find final ')' at line 2, near "> 2) { say"␤in Main (file <unknown>, line <unknown>)␤
08:54 spinclad ng: sub foo (Int $x where * > 2) { say $x }; foo(3);
08:54 p6eval ng 4456d8: Confused at line 1, near "sub foo (I"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (src/stage0/HLL-s0.pir:328)␤
08:55 spinclad rakudo: sub foo (Int $x where {$^x > 2}) { say $x }; foo(3);
08:55 p6eval rakudo db84bc: 3␤
08:56 * spinclad thinks the * > 2 one should work, eventually
08:56 hejki sure
08:56 hejki tbh i need: PERL5!!: sub foo { my $x = shift; $x = 1 if ($x <= 2); .. }
08:57 spinclad if you need that i'd do the same sort of thing:
08:58 spinclad rakudo: sub foo (Int $x) { $x = 1 if $x <= 2; ... }
08:58 p6eval rakudo db84bc:  ( no output )
08:58 hejki rakudo: sub foo (Int $x where {$^x > 2} or $x = 1) { say $x }; foo(4)
08:58 p6eval rakudo db84bc: Unable to parse multisig; couldn't find final ')' at line 2, near "or $x = 1)"␤in Main (file <unknown>, line <unknown>)␤
08:58 hejki :>
08:58 spinclad i don't recall a perl6 shortcut
08:59 hejki rakudo: sub foo (Int $x = 1 where {$^x > 2}) { say $x }; foo(4)
08:59 p6eval rakudo db84bc: Unable to parse multisig; couldn't find final ')' at line 2, near "where {$^x"␤in Main (file <unknown>, line <unknown>)␤
08:59 spinclad multi sub foo (Int $x) { $x = 1; ... }; multi foo (Int $x where * > 2) { ... }
09:00 spinclad rakudo: multi sub foo (Int $x) { $x = 1; ... }; multi foo (Int $x where * > 2) { ... }
09:00 hejki i was just gonna say i need to do it "haskell"-style :)
09:00 p6eval rakudo db84bc: Unable to parse multisig; couldn't find final ')' at line 2, near "> 2) { ..."␤in Main (file <unknown>, line <unknown>)␤
09:01 * spinclad -> bed &
09:01 iblechbot joined #perl6
09:05 hejki ty for all :>
09:09 cognominal joined #perl6
09:30 meneldor joined #perl6
09:31 meneldor hello all and happy new year!
09:31 mberends likewise meneldor :)
09:36 ejs joined #perl6
09:37 mathw happy
09:37 mathw ha
09:37 * mathw has first-day-back-at-work grumps
09:42 * Su-Shee is looking with a first-day-of-work-feeling at the code which already didn't work last year. ;)
09:43 mathw I'm trying to deal with a bug report that concerns a subsystem I've spent the last two years trying to ignore
09:43 mathw I guess I always suspected that would come back to haunt me
10:25 bryan[c1] joined #perl6
10:44 moritz_ oh hai
10:45 mberends hi moritz_, I wish you an awesome 2010!
10:45 moritz_ mberends: same to you, kind sir!
10:47 pugs_svn r29453 | moritz++ | [t/spec] fix a test and smartlink in list_array_perl.t
10:47 pugs_svn r29453 |
10:47 pugs_svn r29453 | Also fudge the test (instead of TODO), because ng doesn't like the new form
10:47 pugs_svn r29453 | (though I have no idea why)
10:49 moritz_ diakopter++ # perl6.org updates
10:49 pugs_svn r29454 | moritz++ | [S05] small clarification for .match and .subst
10:51 dalek rakudo/ng: 52d140e | moritz++ | src/core/Any-str.pm:
10:51 dalek rakudo/ng: implement :g in .match
10:51 dalek rakudo/ng: review: http://github.com/rakudo/rakudo/commit/52d140efe22b769ed15699527aa0dff3c5181f93
10:51 dalek rakudo/ng: e06af48 | moritz++ | src/core/Any-str.pm:
10:51 dalek rakudo/ng: do not limit Any.subst to regexes, for example strings are fine too
10:51 dalek rakudo/ng: review: http://github.com/rakudo/rakudo/commit/e06af483923dd09fb66698367bc5939ecd7e6942
10:51 dalek rakudo/ng: fe1e749 | moritz++ | src/core/Any-str.pm:
10:51 dalek rakudo/ng: Any.subst() is not global by default, only if :g is present
10:51 dalek rakudo/ng: review: http://github.com/rakudo/rakudo/commit/fe1e749e1a77c2e7fb42ef53a9c40cf0c155dee7
10:51 dalek rakudo/ng: 8817454 | moritz++ | lib/Safe.pm:
10:51 dalek rakudo/ng: [Safe.pm] disallow slurp()
10:51 dalek rakudo/ng: review: http://github.com/rakudo/rakudo/commit/8817454d6b875fdb177ffe79d50c49585ba022f8
11:09 zamolxes joined #perl6
11:25 Helios-- joined #perl6
11:29 SmokeMachine joined #perl6
11:41 colomon joined #perl6
11:42 colomon All sorts of ng patches overnight, \o/ !!
11:42 ab5tract joined #perl6
11:57 Sivasakthikumar joined #perl6
11:57 Sivasakthikumar Hi
11:58 Sivasakthikumar Hi
11:59 colomon hello
12:00 Sivasakthikumar yes .. I'm new in this gropu
12:01 Sivasakthikumar Can u help to me How to Download and install perl 6 and Join this community
12:02 Sivasakthikumar hello
12:02 patspam joined #perl6
12:03 Sivasakthikumar Hello
12:04 colomon http://perl6advent.wordpress.com/2009/12/01/day-1-getting-rakudo/
12:05 Sivasakthikumar Thanks
12:05 soupdragon hi
12:06 soupdragon rakudo: say (10/7).WHAT
12:06 p6eval rakudo db84bc: Rat()␤
12:08 Baggio_ joined #perl6
12:09 payload joined #perl6
12:20 cognominal joined #perl6
12:28 plainhao joined #perl6
12:28 bluescreen joined #perl6
12:29 hejki is there something eqv to bigint?
12:29 colomon hejki: In theory, Int.
12:29 colomon in practice, bot yet in Rakuo.
12:30 colomon not yet in
12:30 colomon Rakudo
12:30 hejki spelling spree ;>
12:31 soupdragon rakudo: say (2**36).WHAT
12:31 p6eval rakudo db84bc: Num()␤
12:31 soupdragon rakudo: say (2**76).WHAT
12:31 p6eval rakudo db84bc: Num()␤
12:31 hejki rakudo: say (2**76).Int.WHAT
12:31 p6eval rakudo db84bc: Int()␤
12:31 hejki :>
12:32 hejki rakudo: say (2**2566).Int
12:32 p6eval rakudo db84bc: -9223372036854775808␤
12:32 hejki rakudo: say Inf.Int
12:32 p6eval rakudo db84bc: -9223372036854775808␤
12:32 hejki :)
12:33 hejki rakudo: say (2**2566)
12:33 p6eval rakudo db84bc: Inf␤
12:33 hejki hmm.. mine returns a sane answer for that
12:33 hejki guess it's the use Safe;
12:33 soupdragon rakudo use Safe.WHAT
12:47 ruoso joined #perl6
12:51 takadonet joined #perl6
12:51 takadonet morning all
12:51 colomon o/
12:52 takadonet colomon: how are u?
12:52 colomon semi-awake
12:52 colomon :)
12:54 cognominal rakudo:  (grammar { token ws { \h* }; rule TOP { ^ <pair>*   }; token pair { \:  \w+  } } ).parse(":a :b"); say @($/<pair>)
12:54 p6eval rakudo db84bc:  ( no output )
12:54 cognominal ng:  (grammar { token ws { \h* }; rule TOP { ^ <pair>*   }; token pair { \:  \w+  } } ).parse(":a :b"); say @($/<pair>)
12:54 p6eval ng 52d140: Method 'isa' not found for invocant of class 'Undef'␤current instr.: 'perl6;Perl6;Actions;_block922' pc 142962 (src/gen/perl6-grammar.pir:29127)␤
12:54 soupdragon rakudo: say (grammar { token ws { \h* }; rule TOP { ^ <pair>*   }; token pair { \:  \w+  } } ).parse(":a :b");
12:54 p6eval rakudo db84bc:  ( no output )
12:54 FCO joined #perl6
12:55 cognominal here it gets me only the first pair :(
13:03 colomon do you need a whitespace rule?
13:04 colomon ie ":a:b" would work?
13:04 cognominal is not what ws is about? I tried using the default, and here, a custom one.
13:05 cognominal colomon, you are right!
13:05 colomon but you never use ws, you just define it
13:05 colomon admittedly I am very new to grammars.  :)
13:05 cognominal TOP being a rule, it should use it...
13:06 cognominal I read a lot about them but did not practice much...
13:07 colomon maybe make it ws { \h+ } and { ^ <pair> ** <ws> }  ?
13:08 colomon that's if you want to require ws between pairs
13:09 colomon or { ^ [ <pair> <ws>? ]* } if the ws is optional ?
13:10 cognominal indeed, quantifiers don't seem to imply  interleaving <ws> in rules
13:10 cognominal I don't know if the specs are explicit about that.
13:11 Kodi joined #perl6
13:11 colomon I've mostly just experimented rather than reading the specs.   ;)
13:11 Kodi So I submitted a tiny Rakudo patch to RT a few months ago (http://rt.perl.org/rt3/Public/Bug/Display.html?id=70105) and it's gotten no responses. Is it less innocuous than I thought, or are the Rakudo folks just busy with other matters?
13:14 colomon Kodi: most of the work has been going on in the ng branch.  but I'll try to see that something happens with your patch in the next day or two.
13:14 Kodi colomon: Cool, thanks.
13:14 cognominal like says TimToady,  I could advocate it both ways.  But knowing him, I would not be surprised to find a third way out. Like a special adverb
13:15 Kodi colomon: So for the time being, patches should be written against ng, not master?
13:15 cognominal <pair> ** <ws>   is top ugly for something that should be pretty common
13:16 colomon Kodi: ng is kind of a minefield at the moment, missing huge hunks of rakudo
13:17 colomon Kodi: in theory, it should get caught up with and become master any week now....
13:17 cognominal In Perl6 grammar, space separated quantified tokens don't happen. But Perl6 is special that way.
13:17 colomon cognominal: but <pair>* and <pair> ** <ws> mean very different things
13:18 cognominal colomon, the good news is with the recent discoveries, we won't need stem cells to clone Jonathan and pmichaud
13:19 ktne joined #perl6
13:19 ktne hello
13:19 colomon but can we have the clones up and running next week?
13:19 colomon ktne: hello
13:19 ktne hi
13:19 ktne is rakudo a little bit slow at startup?
13:20 colomon yes
13:20 ktne will this ever be corrected or it's expected to be like that?
13:20 colomon it's certainly hoped it will be corrected.
13:20 ktne ok then, thanks
13:21 ktne is there any reason why there are so many tests and spectests that are skipped?
13:21 ktne all of them passed except the ones that have been skipped for some unknown reason
13:21 colomon it's already a goods bit better than it was six months ago (at startup)
13:22 Kodi ktne: Rakudo is still a good ways away from being complete, and the specification is still somewhat in flux.
13:22 colomon skips are tests which are expected to fail with the current implementation, ie things that still need to be done.
13:22 ktne ah
13:24 Kodi left #perl6
13:28 Baggio_ joined #perl6
13:33 hejki rakudo: say "sum: " ~ [+] (gather { for 2..20 -> $i { take $i if ([*] (map { $^a % $^b }, (($i) X (2..$i-1)))) != 0 } });
13:33 p6eval rakudo db84bc: sum: 77␤
13:37 * rjbs waits for the first person to report a block not working with { $^one, $^two, $^three }
13:38 moritz_ phenny: tell masak maybe http://perlgeek.de/blog-en/perl-6/perl-6-in-2009.writeback inspires you while writing spec changes summary ... though it's very coarse and selective
13:38 phenny moritz_: I'll pass that on when masak is around.
13:44 colomon moritz_++ for ng commits, yay!
13:45 moritz_ what we have now for Any.subst won't work with closures
13:49 colomon moritz_: yes, I know.
13:49 colomon Wasn't quite sure how to make it better...
13:54 moritz_ for regexes, .match($regex, :g) plus a map and a few substrs should be enough
13:54 moritz_ but I'm hesitant to actually implement it that way
13:55 moritz_ because it means that we need a third alternative for string pattern and replacement closures
13:58 cognominal joined #perl6
13:59 Baggio__ joined #perl6
14:03 PerlJam good morning all
14:03 moritz_ oh hai PerlJam
14:18 colomon o/
14:19 * colomon just remembered it takes over an hour to run spectest on master...
14:19 arnsholt Then you've got time to grab a coffee =)
14:19 moritz_ I've just got an idea
14:19 moritz_ a perl6-websites mailing lists
14:20 moritz_ as a central contact point for perl6.org, dev.perl.org/perl6/ and all the other Perl 6 websites whos maintainers wish to participate
14:23 athomaso1 joined #perl6
14:23 buu joined #perl6
14:23 BooK joined #perl6
14:23 Maddingue joined #perl6
14:23 rjbs joined #perl6
14:23 bloonix joined #perl6
14:25 lestrrat joined #perl6
14:27 meppl joined #perl6
14:28 moritz_ ng doesn't build with latest parrot (due to callsignature/callcontext merge)
14:28 moritz_ looks like a job for jnthn++
14:28 krunen joined #perl6
14:31 Tene Maybe in my licensing email I should have just said "Consider my contributions MIT/X11 licensed, and feel free to relicense in any way you see fit."
14:32 moritz_ and s/whatever/*/ please :-)
14:33 BinGOs joined #perl6
14:33 cognominal joined #perl6
14:33 * mberends tries building the latest ng with Parrot r43076. it works :)
14:38 xinming joined #perl6
14:38 colomon mberends: ship it!  ;()
14:38 colomon ;)
14:39 lumi joined #perl6
14:41 mberends colomon: :)
14:41 scion joined #perl6
14:44 zamolxes joined #perl6
14:51 colomon Just got the following comment for approval on Perl 6 Advent:
14:52 colomon Да уж. По поводу коментариев - навеяла на меня где-то услышанная фраза:
14:52 colomon Только наш клиент в графе “Телефон” мог написать Samsung.
14:52 colomon Babelfish gives that as Yes already. Apropos [komentariev] - blew to me the somewhere heard phrase: Only our client in the graph “telephone” could write Samsung.
14:52 colomon Anyone have an idea if this is a sensible comment or spam?
14:53 Grrrr it's a sensible comment, surprizingly
14:54 moritz_ colomon: we've had a few trackbacks from russian translations of our articles
14:55 Grrrr colomon: a better translation would be:
14:55 Grrrr Well.  About the comments, I was reminded by a phrase I've heard
14:55 Grrrr somewhere: "it is only *our* client who could write 'Samsung' in
14:55 Grrrr the 'Phone' field".
14:56 PerlJam colomon: I tend to go with google translate these days as it does a better job than babelfish.
14:57 mberends sarcasm seems to be a popular style in Russian
15:00 mberends but this  Russian seems to be commenting about another comment(ator), not about a posting
15:02 jaldhar_ joined #perl6
15:04 mberends like writing in an application form, Length of residence in Australia: about 15 metres
15:07 colomon danke!  approved, then.
15:10 jaldhar_ joined #perl6
15:11 gbacon joined #perl6
15:18 Juerd joined #perl6
15:23 jaldhar_ joined #perl6
15:26 sunnavy joined #perl6
15:30 jaldhar_ joined #perl6
15:31 bluescreen joined #perl6
15:36 nihiliad joined #perl6
15:44 am0c joined #perl6
15:46 xinming joined #perl6
15:49 Psyche^ joined #perl6
15:53 alester joined #perl6
15:54 pardunci1 joined #perl6
16:02 astrojp joined #perl6
16:12 somebody_ joined #perl6
16:12 moritz_ $somebody sent me a Perl version of a C++ program I've written and published two years ago
16:13 moritz_ added some fancy OO
16:13 moritz_ and made it a factor 1000 slower :-)
16:14 lichtkind joined #perl6
16:15 soupdragon joined #perl6
16:15 PerlJam moritz_: Perl 5 or Perl 6?
16:15 moritz_ PerlJam: Perl 5
16:16 mj41 joined #perl6
16:18 Su-Shee ugh, what kind of OO did he use?!
16:19 arc_of_descent joined #perl6
16:19 moritz_ 4 method calls + parameter checks with emulated named parameters for things that were a single array access in the C++ version
16:21 moritz_ 4 methods that are called 50k to 100k times per second in the perl version
16:22 lumi joined #perl6
16:23 szbalint looks like premature sugarification.
16:23 Su-Shee *hihi* :)
16:25 orafu joined #perl6
16:25 justatheory joined #perl6
16:26 arc_of_descent left #perl6
16:27 lichtkind hello
16:31 somebody__ joined #perl6
16:32 ash_ joined #perl6
16:32 somebody__ joined #perl6
16:32 _jaldhar_ joined #perl6
16:33 soupdragon joined #perl6
16:44 douglashunter joined #perl6
16:47 lumi joined #perl6
16:48 colomon moritz_: Now you just need to convince them to convert it to Perl 6, achieving both OO brilliance and a total 1,000,000 x slowdown.  :)
16:51 diakopter :/
16:52 obra ...why does dev.pugscode.org now redirect to audrey's github?
16:52 diakopter obra: oh
16:52 obra and how can I get at the trac history browser?
16:52 diakopter I turned it off b/c nothing (including the svn mirror) in it had been updated in a year
16:53 obra ....untrue.
16:53 diakopter oh :( ok.. i'll reactivate it now
16:53 obra I was seeing commits from that day when I looked at it a week ago :)
16:53 obra and audrey's github is....very out of date
16:53 diakopter ?
16:53 diakopter it's the one used to create the hackage releases, I thought
16:54 obra dev.pugscode.org shows audrey's git mirror of pugs from 2008
16:54 obra which isn't what I expected ;)
16:54 obra regardless, looking forward to it being turned back on
16:55 diakopter that should do it
16:56 obra testing
16:56 pmurias joined #perl6
16:56 obra hm. can't get there.
16:56 diakopter 'tis slow
16:56 obra OperationalError: database is locked
16:57 diakopter hm, maybe apache2 needs restarted instead of just reloaded
16:57 kaare_ joined #perl6
16:57 pmurias diakopter: did you get JavaScript::V8 to work?
16:59 obra diakopter: much better. tx
16:59 diakopter well
16:59 diakopter it still throws errors
17:00 diakopter when more than 1 person try to access it simultaneously
17:00 diakopter also, our zend php license has expired, whatever that means. :)
17:01 diakopter obra: it's b/c trac is setup to use the sqlite backend, which... is not a multi-user database (as I'm sure you're well aware)
17:02 diakopter so when the poor contention/wait algorithms time out, errors arise.
17:02 obra actually, it's fine so long as you don't assume write concurrency.
17:02 obra it _is_ multi-reader
17:02 ejs joined #perl6
17:03 diakopter obra: why not use the SVN::Web instance on feather instead of the one in trac?
17:04 diakopter trac is spammed to high heaven.
17:05 diakopter many of the links are to wiki.openqa.org, which is also spammed to high heave.
17:05 diakopter n
17:06 obra ...because it's linked through from the front page of www.pugscode.org
17:08 _jaldhar_ joined #perl6
17:12 pugs_svn r29455 | pmurias++ | [mildew] simplifing method calls
17:12 diakopter pmurias: no; I haven't tried it
17:12 pugs_svn r29456 | pmurias++ | [mildew] simplified for AST::If,AST::Let
17:16 mberends pmurias: did you manage to get that software from the internet to convert your mildew's AST? ;)
17:19 sjohnson rakudo:  my @a = <cow pig chicken>; say @a.grep("i").perl
17:19 pmurias mberends: still working on converting mildew AST to a set of basic nodes which corespond directly to m0ld
17:19 p6eval rakudo db84bc: []␤
17:19 sjohnson rakudo:  my @a = <cow pig chicken>; say @a.grep(/i/).perl
17:19 p6eval rakudo db84bc: ["pig", "chicken"]␤
17:19 mberends pmurias: ok, the README is taking shape here
17:21 pugs_svn r29457 | diakopter++ | [sprixel] finish up geitherrs
17:23 pmurias mberends: after that i'll convert it to SSA, i'll have to code that myself but the algorithm is already on the internet ;)
17:24 mberends pmurias: ok
17:27 * TimToady attempts to wake up after sleeping 13 of the last 18 hours...
17:29 lichtkind oh i sleep also very long lately
17:29 lichtkind thatswhy i said lately
17:30 pmurias TimToady: i removed YAML::Syck from STD.pmc
17:30 ash_ winter makes me want to slep a lot
17:30 TimToady pmurias: thanks
17:32 TimToady with me, it's hard to detangle SAD from sleep deprivation over the last 3 weeks helping take care of Julian
17:32 orafu joined #perl6
17:32 TimToady but the proximate cause was driving straight from Seattle to Mountain View
17:34 takadonet TimToady: how long of a drive is that?
17:34 pugs_svn r29458 | diakopter++ | [sprixel] start and finish gbothrs.  two to go (gbothlrs and geitherlrs).
17:36 pmurias diakopter: what are those strange gbothrs and geitherrs?
17:37 sjohnson rakudo:  my @a = <cow pig i chicken>; say @a.grep("i").perl
17:37 p6eval rakudo db84bc: ["i"]␤
17:38 diakopter pmurias: good question :)
17:38 diakopter gbothlrs stands for grammar both left and right stateful^Wnondeterministic
17:38 diakopter a sequence of two possibly-backtracking pattern nodes
17:38 TimToady takadonet: 'bout 13 hours of driving, plus sotps; 850 miles
17:38 TimToady *stops
17:39 diakopter pmurias: they're specialized for the various permutations to minimize the runtime checks/branches.
17:39 diakopter afk&
17:42 sjohnson this new p6 grep is a big deal
17:42 sjohnson might also be because regex != string in p6
17:42 flip214 joined #perl6
17:43 flip214 rakudo: say [R-] 1,4,9;
17:43 p6eval rakudo db84bc: 6␤
17:43 flip214 rakudo: say [R-] 1,4,9,16;
17:43 p6eval rakudo db84bc: 10␤
17:43 flip214 rakudo: say [R-] 1,2;
17:43 p6eval rakudo db84bc: 1␤
17:43 flip214 rakudo: say [R-] 1,4;
17:43 p6eval rakudo db84bc: 3␤
17:43 flip214 rakudo: say [R-] 1,4,2;
17:43 p6eval rakudo db84bc: -1␤
17:43 flip214 rakudo: say [R-] 1,4,9;
17:43 p6eval rakudo db84bc: 6␤
17:44 flip214 Fine, Left-to-right reduce!
17:44 flip214 No, wrong ...
17:44 flip214 rakudo: say [RR-] 1,4,9,16;
17:44 p6eval rakudo db84bc: Confused at line 2, near "-] 1,4,9,1"␤in Main (file <unknown>, line <unknown>)␤
17:45 flip214 rakudo: say [R/] 1,4,9,16;
17:45 p6eval rakudo db84bc: 7.11111111111111␤
17:45 flip214 rakudo: say [R/] 1,4,9;
17:45 p6eval rakudo db84bc: 2.25␤
17:45 snearch joined #perl6
17:45 cdarroch joined #perl6
17:47 flip214 rakudo: say "Good night everyone"
17:47 p6eval rakudo db84bc: Good night everyone␤
17:49 phenny joined #perl6
17:53 FCO joined #perl6
17:54 phenny joined #perl6
17:57 TimToady hmm, S03:3777 doesn't spec whether R reverses the associativity.
17:58 xinming_ joined #perl6
18:10 cognominal joined #perl6
18:23 xomas_ joined #perl6
18:34 cognominal_ joined #perl6
18:35 pugs_svn r29459 | mberends++ | [mildew/README] new docs, unverified
18:38 Tene TimToady: I'd like some clarification on S04... it says "Any othe ruse of the Failure object"... but doesn't specify what other uses are possible.  Does Failure forward everything to the Exception object it refers to?
18:38 orafu joined #perl6
18:39 Tene I guess I should really make a diff of what I'd propose in that section.
18:40 hejki rakudo: say [*] (1..100)
18:40 p6eval rakudo db84bc: 9.33262154439441e+157␤
18:40 hejki how-to remove e+..-format?
18:41 Tene maybe .fmt?
18:42 hejki rakudo: say ([*] (1..100)).fmt("%.0f")
18:42 p6eval rakudo db84bc: 93326215443944102188325606108575267240944254854960571509166910400407995064242937148632694030450512898042989296944474898258737204311236641477561877016501813248␤
18:42 hejki najs
18:43 mberends but wrong, it should end in a couple of zeroes :(
18:44 stephenlb joined #perl6
18:46 mberends pugs: say ([*] (1..100)).fmt("%.0f")
18:46 p6eval pugs: 93326215443944150000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000␤
18:46 mberends pugs: say ([*] (1..100))
18:46 p6eval pugs: 93326215443944152681699238856266700490715968264381621468592963895217599993229915608941463976156518286253697920827223758251185210916864000000000000000000000000␤
18:46 astrojp joined #perl6
18:47 PerlJam nice.  They differ completely starting from the 16th digit
18:48 Tene I'm kind of curious about the sequence "Largest number that can be produced by p6 code of length n"
18:49 soupdragon is that busy beaver
18:49 Tene Kinda.  Not sure if it's different or not.
18:51 ShaneC joined #perl6
18:55 Su-Shee a book teaching "good perl 6 style" is going to be interesting...
18:55 partisan joined #perl6
18:56 PerlJam Su-Shee: "Perl 6 Best Practices"  :)
18:56 * Su-Shee essentially has already bought "Higher Order Perl 6" :)
18:57 Su-Shee well, I'm prepared to. ;)
18:58 hejki rakudo: my $str = "foobar"; say $str.comb.elems; say "any other ways? :)"
18:58 p6eval rakudo db84bc: 6␤any other ways? :)␤
18:58 Tene any other ways to do what?
19:00 Tene HOP is a very nice book.
19:00 PerlJam hejki: This is Perl.  There are always other ways.
19:03 sjohnson lifetime registration on as many computers that you own so im doing the "right thing"
19:03 Su-Shee rakudo: my $str = "foobar"; say .elems for $str.comb;
19:03 p6eval rakudo db84bc: 1␤1␤1␤1␤1␤1␤
19:04 Tene registration of what?
19:04 Su-Shee rakudo: my $str = "foobar"; say .perl for $str.comb;
19:04 p6eval rakudo db84bc: "f"␤"o"␤"o"␤"b"␤"a"␤"r"␤
19:05 hejki hmm
19:05 hejki my point being
19:05 hejki rakudo: my $str = "foobar"; say $str.length;
19:05 p6eval rakudo db84bc: Method 'length' not found for invocant of class 'Str'␤in Main (file src/gen_setting.pm, line 324)␤
19:05 hejki is there any eqv?
19:05 PerlJam rakudo: my $str = "foobar"; say $str.chars
19:05 orafu joined #perl6
19:05 p6eval rakudo db84bc: 6␤
19:05 hejki ahh
19:05 hejki nice
19:05 Su-Shee rakudo: my $str = "foobar"; say $str.elems;
19:05 hejki PerlJam++
19:05 p6eval rakudo db84bc: 1␤
19:06 Tene hejki: the problem is that "length" is ambiguous between chars, bytes, and graphs.
19:06 Tene so you need to specify.
19:06 PerlJam hejki: "length" is insufficient because you aren't specifying the units of measure
19:06 PerlJam what Tene said
19:06 Tene consider multibyte chars, and combining chars.
19:06 hejki ahh
19:06 hejki true :)
19:06 cognominal_ rakudo: class A; say (A.new x 3) ».» perl
19:06 p6eval rakudo db84bc: Confused at line 2, near "\x{bb}.\x{bb} perl"␤in Main (file <unknown>, line <unknown>)␤
19:07 cognominal_ should method calling be hyperoberable?
19:07 Tene It is.
19:07 Tene rakudo: class A; say (A.new x 3)>>.perl
19:07 cognominal_ rakudo: class A; say ((A.new x 3) ».» perl)
19:07 cognominal_ oops
19:07 Tene rakudo: class A {}; say (A.new x 3)>>.perl
19:07 Tene oright
19:07 Tene ng: class A {}; say (A.new x 3)>>.perl
19:08 p6eval rakudo db84bc:  ( no output )
19:08 p6eval rakudo db84bc: Confused at line 2, near "\x{bb}.\x{bb} perl)"␤in Main (file <unknown>, line <unknown>)␤
19:08 p6eval rakudo db84bc:  ( no output )
19:08 p6eval ng 52d140: Method 'Str' not found for invocant of class 'A'␤current instr.: 'perl6;Mu;' pc -1 ((unknown file):-1)␤
19:08 Tene cognominal_: class declaration is always a timeout in rakudo master for p6eval
19:11 cognominal_ rakudo:  (Int.new x 3) ».» perl
19:11 p6eval rakudo db84bc: Confused at line 2, near "\x{bb}.\x{bb} perl"␤in Main (file <unknown>, line <unknown>)␤
19:11 cognominal_ rakudo:  say (Int.new x 3) >>.perl
19:11 p6eval rakudo db84bc: Confused at line 2, near ">>.perl"␤in Main (file <unknown>, line <unknown>)␤
19:12 cognominal_ rakudo:  say ("f","oo","bar").>>.chars;
19:12 p6eval rakudo db84bc: 123␤
19:13 cognominal_ at last, I got it right
19:13 cognominal_ reading the specs help
19:15 orafu joined #perl6
19:18 Tene you can't have a space between something and its postfix.
19:18 Tene rakudo: say (Int.new x 3)>>.chars
19:18 p6eval rakudo db84bc: 3␤
19:20 Chillance joined #perl6
19:22 colomon rakudo: say (Int.new x 3)>>.perl
19:22 p6eval rakudo db84bc: "000"␤
19:23 Tene I also suspect that you mean 'xx'
19:23 Tene rakudo: say (Int.new xx 3)>>.perl
19:23 p6eval rakudo db84bc: 000␤
19:23 Tene 'x' is str rep, 'xx' is list rep
19:24 rachelBROWN joined #perl6
19:24 cognominal_ yea, I am still falling in many Perl 5 traps.
19:25 sjohnson i do it too :(
19:25 cognominal_ not only falling for them, but taking a while to realize what is happening...
19:25 sjohnson even fell in a bad php trap and used == for strings
19:27 orafu joined #perl6
19:28 cognominal_ a super strict mode should oblige to explicitely numfiy the operands if == is really intended.
19:32 sjohnson please elaborate on the numify thing
19:35 cognominal_ sorry. forget what I said, it would not help on strings.
19:42 nihiliad joined #perl6
19:45 Exodist joined #perl6
19:45 orafu joined #perl6
19:59 orafu joined #perl6
20:00 coke joined #perl6
20:00 coke is RT #71812 due to a missing "begin" ?
20:00 mariano__ joined #perl6
20:01 coke (I would imagine his complaint is correct.)
20:06 ive joined #perl6
20:11 ash_ joined #perl6
20:12 ive joined #perl6
20:13 Su-Shee _HyperWhatever_? :) oh my.. :)
20:13 orafu joined #perl6
20:19 colomon HyperWhatever?
20:20 Su-Shee I'm reading http://perlcabal.org/syn/S02.html
20:21 colomon ouch.  that made my head hurt.
20:23 Su-Shee I'm already crying while hysterically laughing. ;)
20:57 ash_ can rakudo target pbc? instead of pir?
20:58 orafu joined #perl6
20:59 PerlJam ash_: what do you mean?  You can generate a PBC from the PIR.
20:59 ash_ ah, good point. i can never figure out the obvious questions
21:00 * diakopter learned about POSIX::pause today
21:01 diakopter in the absence of Time::HiRes, I can use it to imitate setitimer
21:03 lumi joined #perl6
21:12 payload joined #perl6
21:13 orafu joined #perl6
21:18 orafu joined #perl6
21:40 PerlJam whoa.   I just ran "make spectest" for the first time in a while and it was *fast*.    I'm guessing that's because of all of the currently-non-functional tests that have been commented out.
21:40 Tene hehe
21:41 colomon PerlJam: ng branch?
21:41 PerlJam aye
21:42 Tene PerlJam: I enabled a test file last night that didn't pass on master. :)
21:42 japhb joined #perl6
21:43 Tene I don't know what exactly to work on next, as so much is pending on the lists refactor plan from pm.
21:43 colomon PerlJam: yeah, something like 90% of the spectests are turned off there.
21:46 mberends pm's probably busy working on it offline. that's how he created nqp-rx. it needs intense concentration.
21:47 Tene It always makes me just a little bit nervous how spread thin the p6 development effort is.  I'm not quite sure what I can do to fix it, though.
21:48 PerlJam Tene: become a clone of either Pm or jnthn so when one of them disappears, you're there to take up the slack  :)
21:48 colomon Hmmm... it seems to me we're doing okay on the overall number of people working on it at the moment.
21:48 Tene That was one of my motivations for working on ng.
21:48 mberends Tene: you are helping out a great deal. I understand your thought, though.
21:48 colomon But we're horribly blocked on ng and jnthn's stuff.
21:49 colomon s/ng/pmichaud/
21:49 patspam joined #perl6
21:49 colomon and wow, that's an interesting freudian slip....
21:49 mberends heh
21:49 PerlJam colomon: though there are still things that can be done to make ng do things that can currently be done in master but not ng.
21:49 PerlJam (without regard to jnthn or pmichaud)
21:50 colomon PerlJam: Sure, but it's a pain in the neck picking through everything trying to find things that can be made to work at the moment.
21:52 zamolxes joined #perl6
21:54 PerlJam colomon: well, mberends is probably correct.  When I talked to Pm the other day he said he was going to focus on get ng in shape to become master.  So, it's likely that we'll see barriers removed over the next couple of weeks.
21:54 colomon Tene: I keep on thinking a few of us should just go ahead and figure out how to make lists work.
21:55 PerlJam colomon: that would be good even if it duplicated effort because of the expertise gained.
21:55 PerlJam (IMHO)
21:55 Tene colomon: Sure.  I just have limited time, and if it's getting done, I'd rather not duplicate effort.
21:57 colomon Tene: understood, I'm in the same boat.  (Actually, my time may be effectively none at the moment.)
21:58 Tene HAH I HAVE MORE TIME THAN YOU!
21:59 orafu joined #perl6
22:00 colomon Tene: yeah, that's what happens when you have to get ready to try to sell your house AND figure out how to make about 30-40% more money at $work this year.
22:01 Tene I can empathize, unfortunatly.
22:02 PerlJam colomon: That second thing is easy ... just get another job and forego sleep or something  :)
22:02 colomon I already got myself a small child to make sure I never get enough sleep.  :)
22:04 PerlJam I've got 4 year old twins that still crawl into bed with us in the middle of the night. Also, one falls asleep quickly, the other tries to stay up all night.
22:05 ash_ I'd be willing to try to contribute to the ng branch more if I knew what needed to be done
22:05 PerlJam ash_: everything!  :)
22:05 ash_ i found some bugs or rather things not yet implemented in ng that i am trying to tackle and just submit a patch
22:06 ash_ PerlJam: see, when you say everything it makes it almost hard to help though, because then i worry someones already doing it
22:07 PerlJam ash_: Find a reasonable sized area of interest and ask here if anyone is working on it.  Either the answer is "no" and you're relatively safe or the answer is "yes" and at worst you can collaborate.
22:08 dalek rakudo/ng: f7ad5dc | duff++ | src/ (2 files):
22:08 dalek rakudo/ng: rename break/continue to succeed/proceed
22:08 dalek rakudo/ng: review: http://github.com/rakudo/rakudo/commit/f7ad5dcf196a0d399a5ef1c5160131da5ed27a73
22:09 ash_ PerlJam: hence my working on meta object stuff, but still, jnthn has said he was going to work on some of the stuff i am doing, i just don't know which parts, also, MOP isn't really speced much in S12~14
22:10 PerlJam ash_: Then you could spec it (either by documenting existing design or by stealing from Moose or whatever)
22:12 ash_ there is a document in ng/docs/metamodel.pod that jnthn started
22:15 colomon Hmmm... perhaps the most productive thing we could do at the moment would be to try to make a partial list of areas in ng that need work and aren't blocked at the moment....
22:15 colomon Also, there's a LOT of room for more/better spectests.
22:16 ash_ metamodel stuff isn't really blocked by anything, its just needs to get implemented as far as i can tell
22:17 ash_ i did notice most of the HOW methods expect a HOW object, but are passing tests so I am assuming there is room for more tests in the metamodel stuff, thats what I was going to do if i finished adding ^methods to roles
22:21 pmurias joined #perl6
22:21 pmurias mberends: the default backend is via-C
22:21 pmurias mberends: -Cm0ld is used to view the m0ld for a given program
22:22 mberends pmurias: oh. see, how much I still need to learn!
22:23 mberends pmurias: would you do a few README edits and commit before I continue with it? I'm re-installing Linux a few times.
22:24 pmurias ok
22:25 colomon afk
22:25 * mberends is very happy with Ubuntu 10.4-alpha1. it even includes libv8-dev :)
22:25 pmurias wow does it work with JavaScript::V8?
22:26 mberends yes! :)
22:27 sjohnson hi
22:27 ash_ mberends: " I'm re-installing Linux a few times." just curious why do you need to do it more than once?
22:27 ash_ multiple computers?
22:29 mberends ash_: one computer, but there are usually a few matters like partitioning that I try out various ways. Also, software installed without using packages is not uninstalled properly.
22:32 ash_ mberends: ah, makes sense. I always setup my computers with a ~/Projects directory and have ~/Projects/local for software not installed with a package system.
22:33 mberends this time around, the root partition will be ext4 and /home gets either ext2 or ext3. There is also hibernate in the swap partition to verify.
22:34 ash_ either that or make my own portfile/ebuild files that i want to install outside that folder, so i can use a package system to uninstall/update automattically if i update the files properly
22:35 mberends there is also multi boot (Win 7, 2008 and XP) to factor in for $work
22:36 pugs_svn r29460 | pmurias++ | fix a couple of things in the README
22:39 ash_ i think the key to all of this is having a reliable backup system, lol the second i start tampering with my setup i almost always mess something up, ususally i can recover just fine, but there are losses sometimes
22:41 mberends so true. for backups I prefer spare drives and computers, and just replicate content often
22:42 ash_ is rakudo bug #71816 really a bug?
22:43 ash_ i don't see it...
22:45 ash_ oh, nevermind i see what the complaint is, the error message doesn't refer to a line# or filename, thats to do with $?FILE and $?LINE but those don't work in rakudo
22:46 cognominal joined #perl6
22:58 lumi joined #perl6
22:59 dalek rakudo/ng: a867b39 | duff++ | src/Perl6/Grammar.pm:
22:59 dalek rakudo/ng: Add unspace parsing
22:59 dalek rakudo/ng: review: http://github.com/rakudo/rakudo/commit/a867b396c6287d001d0cc247629a3dbcb948fdc5
22:59 dalek rakudo/ng: 30136a1 | duff++ | src/Perl6/Grammar.pm:
22:59 dalek rakudo/ng: Parse variable shape in declarations
22:59 dalek rakudo/ng: review: http://github.com/rakudo/rakudo/commit/30136a118fb712ccaeb39c879b1ada7578122b80
23:30 colomon What do those patches do, then?
23:37 ruoso joined #perl6
23:43 lumi joined #perl6

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

Perl 6 | Reference Documentation | Rakudo