Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2013-01-02

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:12 sftp joined #perl6
00:36 Targen_ joined #perl6
00:40 screencast joined #perl6
00:43 timotimo er, is the rendering of S26 correct? it says "D<POD> is an easy-to-use markup language" and "B<=begin pod>" and "B<=end pod>" :(
00:45 JJ_Brain joined #perl6
00:48 timotimo oh, right, it says so right next to it that it's known to be incomplete
00:48 timotimo so that's fine i guess
01:01 adu joined #perl6
01:02 cognominal joined #perl6
01:13 sftp joined #perl6
01:30 geekosaur joined #perl6
01:30 geekosaur joined #perl6
01:56 timotimo i'll get back to the screencast project after a bit of sleep.
02:00 ggoebel joined #perl6
02:14 cognominal joined #perl6
02:34 FROGGS_ joined #perl6
02:42 asdf12 joined #perl6
02:54 ggoebel wondering if initial jvm backend intends to make use of the "da vinci machine" (JSR 292). Looks like the folks at Oracle are continuing to work on improving support for dynamic languages. whereas .Net's DLR seems to have stagnated
02:58 ggoebel Nashorn (Oracle's new open source javascript runtime for jvm jdk8 uses it) http://mail.openjdk.java.net/pipermai​l/mlvm-dev/2012-December/005158.html
02:59 au InvokeDynamic++
03:11 au JSR 335 (Java 8, project Lambda) may also be relevant, with JVM-level support of "Default methods" dovetailing with method-carrying roles. # http://cr.openjdk.java.net/~brian​goetz/lambda/lambda-state-4.html
03:17 ggoebel Current thread on LF performance http://mail.openjdk.java.net/pipermai​l/mlvm-dev/2012-December/005162.html
03:21 au looks pretty good.
03:25 xinming joined #perl6
03:32 [Coke] au: sorry to bug you, but pugs is broken on feather. Not sure if you have any tuits to poke at it.
03:35 au /usr/bin/pugs looks OK'ish. how exactly is it broken?
03:36 au (please typeahead; I'll be back in a bit and take a look after lunch.)
03:36 [Coke] sorry, building a fresh pugs is broken.
03:37 [Coke] (so the daily test runs have been showing pugs at 0 for some time.)
03:37 au ah, ok. rebuilding it in the background screen now; will see how it goes.
03:42 [Coke] au++
03:42 au [Coke]: A fresh "cabal update ; cabal install Pugs" worked for me producing /home/audreyt/.cabal/bin/pugs .
03:42 au trying a Github checkout "make" in Pugs.hs/Pugs/ now.
03:43 [Coke] let me reproduce my failure for you.
03:43 au ok. "rm -rf ~/.cabal" might help.
03:44 au compiling from Git checkout also worked for me, producing /home/audreyt/Pugs.hs/Pugs/pugs .
03:44 au for the time being please feel free to copy/use my checkout.
03:44 au (bbiab, lunch)
03:45 * au is happy that a fresh recompile-the-world takes less than 7mins and Pugs itself takes just 3mins.
03:46 cognominal joined #perl6
03:46 [Coke] * Missing (or bad) header file: perl5/p5embed.h
03:46 [Coke] removing ~/.cabal...
03:46 au also "git clean -dfx ; git pull" if you're working from a git checkout
03:47 [Coke] then I get:      fatal error: libutil.h: No such file or directory
03:47 au "git reset --hard" may help too
03:47 [Coke] then I get the p5embed.h error again.
03:48 [Coke] git clean -xdf ; git reset --hard; git pull; rm -rf ~/.cabal; make # gets the p5embed.h error.
03:49 au apparently it's because you're using a perlbrew
03:49 au perlbrew off ; make
03:51 au actually, before make, run "cabal install Pugs" first
03:51 au but in any case that should work... the perlbrew was perhaps installed with an out-of-sync toolchain or something.
03:53 [Coke] Had removed ~/.ghc as part of "remove everything" test.. reinstalling some stuff, will try sans perlbrew shortly.
03:53 [Coke] getting a lot of "Error while Error while updating world-file. : /home/autrijus/.cabal/: openNewBinaryFile: perm"
03:54 [Coke] (though it doesn't seem to be impacting my installs)
03:55 * shachaf recommends mkdir old; mv .cabal .ghc old/ rather than rm -rf .cabal .ghc7
03:55 shachaf Often it turns out there was something bbin there that you wanted.
03:56 [Coke] au++ #muuch better.
03:57 [Coke] shachaf: I use haskell only for pugs, so rebuilding the dozen deps isn't a pain.
03:58 [Coke] of course, now every test is failing because "this version of PCRE was compiled without UTF support".
03:58 [Coke] still much better, though.
03:58 orafu joined #perl6
04:02 dalek perl6-roast-data: 3be2124 | coke++ | / (2 files):
04:02 dalek perl6-roast-data: today (automated commit)
04:02 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/3be2124f68
04:02 dalek perl6-roast-data: 09ae7ee | coke++ | bin/pugs.sh:
04:02 dalek perl6-roast-data: disable perlbrew for pugs build. (au++)
04:02 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/09ae7ee1aa
04:16 au [Coke]: fixed the PCRE support
04:19 [Coke] how do I get the udpated version?
04:20 [Coke] do I need to rebuild something with cabal?
04:20 au run the command "sleep 300" :)
04:21 au (I'm running "cabal update ; cabal install --reinstall regex-pcre-builtin pugs-compat ; make" su'd as you atm)
04:21 skids joined #perl6
04:21 au alright, /home/coke/sandbox/Pugs.hs/Pugs/pugs now has utf8 pcre support.
04:26 au ...and so does /usr/bin/pugs. thank you for bringing it back to the roast
04:27 adu joined #perl6
04:28 sorear au \o/
04:28 [Coke] au++
04:28 au ohai sorear \o/
04:29 adu hi
04:35 dduncan joined #perl6
04:35 dduncan is there an admin for https://github.com/perl6 around?
04:36 dduncan I used to be a member of that GitHub organization but had removed myself in order to reorganize my accounts; now I'd like to be re-added to perl6
04:36 dduncan specifically, "muldis" was a perl6 member, but I converted that to an organization and created a new personal account duncand
04:36 dduncan can duncand be added to perl6?
04:37 au neat, haven't noticed the "turn into an organization" feature before
04:38 au <- not an admin for github.com/perl6/
04:39 dduncan I was setting things up to be more social, so its easier to have a group working on my Muldis projects with direct commit access, as is common in the Perl community
04:39 dduncan rather than everyone else having to do pull requests
04:39 au *nod*
04:40 adu is there a way to add metadata to a sub?
04:40 dduncan I actually removed myself from perl6 several months ago in preparation, but just did the actual account conversion today
04:40 dduncan that is all
04:48 sorear dduncan: you want duncand added to perl6?  I can do this
04:49 au adu: yes (with trait_mod, see https://perl6advent.wordpress.com/2011/​12/04/traits-meta-data-with-character/ ), but what would you like to use it for?
04:49 dalek Pugs.hs: ddfcad8 | au++ | pugs-compat/pugs-compat.cabal:
04:49 dalek Pugs.hs: * Bump min regex-pcre-builtin to 0.94.4.4.8.31 to support UTF8 and GHC7.6
04:49 dalek Pugs.hs: review: https://github.com/perl6/Pugs.hs/commit/ddfcad8d84
04:50 dduncan sorear, yes, thank you
04:50 dduncan https://github.com/duncand
04:50 dduncan is my new personal profile
04:51 dduncan I can't promise to be very active, but I was a member before (under "muldis") and occasionally updated the Perl 6 spec or such
04:54 tadzik joined #perl6
04:54 * [Coke] wonders why feather is so slow. top not showing anything.
04:55 dduncan "duncand" matches my CPAN id, fyi
04:55 [Coke] ... and now it seems responsive again (after about 10m of slowdown)
04:55 sorear dduncan: done
04:55 dduncan thank you
05:29 dalek ecosystem: 8815b57 | (David Warring)++ | META.list:
05:29 dalek ecosystem: Update META.list
05:29 dalek ecosystem:
05:29 dalek ecosystem: Added PDF-Grammar
05:29 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/8815b57c2b
05:29 dalek ecosystem: 23bc919 | util++ | META.list:
05:29 dalek ecosystem: Merge pull request #17 from dwarring/patch-1
05:29 dalek ecosystem:
05:29 dalek ecosystem: Update META.list
05:29 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/23bc91951a
05:39 att__ joined #perl6
06:12 dduncan left #perl6
06:15 kaleem joined #perl6
06:20 GlitchMr joined #perl6
06:31 ismail joined #perl6
06:58 JJ_Brain joined #perl6
06:58 daniel-s_ joined #perl6
07:15 Chillance joined #perl6
07:20 baest joined #perl6
07:42 geekosaur joined #perl6
07:42 JJ_Brain joined #perl6
07:45 geekosaur joined #perl6
07:58 asdf12 joined #perl6
07:58 SamuraiJack joined #perl6
08:44 JJ_Brain joined #perl6
08:45 quester left #perl6
08:45 JJ_Brain joined #perl6
08:45 rindolf joined #perl6
09:14 Rix joined #perl6
09:27 SmokeMachine joined #perl6
09:28 SmokeMac_ joined #perl6
09:29 rindolf joined #perl6
09:35 rindolf joined #perl6
09:42 Psyche^ joined #perl6
09:49 Pupubear joined #perl6
10:31 kivutar joined #perl6
10:36 JJ_Brain joined #perl6
10:44 hoelzro in token statement_control:sym<...>, :sym is an adverb, and <...> are arguments to that adverb, yes?
10:47 arnsholt Good question. In previous versions of the Rakudo I think it's been treated separately
10:48 arnsholt But last time I looked into the grammar was before we actually got proper adverb handling
10:48 arnsholt *the Rakudo grammar
10:50 Kharec joined #perl6
10:57 gaussblurinc_ joined #perl6
11:03 masak I'm not sure that particular colonpair gets to be called "adverb".
11:04 masak adverbs are more like named options you pass to operators.
11:04 masak like `$a + $b :carefully`
11:04 masak (g'day #perl6)
11:07 hoelzro masak: so is :sym<...> special?
11:13 ankitsingh joined #perl6
11:13 masak std: sub foo:hihihihi<14>() {}
11:13 p6eval std a8bc48f: OUTPUT«ok 00:00 43m␤»
11:13 masak hoelzro: no.
11:14 hoelzro I see.
11:14 hoelzro masak: so :sym is a trait?
11:15 masak "trait" is used for something else in the spec. in 'is rw', 'rw' is the trait.
11:15 hoelzro I see
11:15 hoelzro so what *is* :sym, then?
11:15 hoelzro is it actually part of the name?
11:15 masak as far as I know, it's just a somewhat unusual part of the... right.
11:15 FROGGS_ it is like a named parameter for a multi sub
11:15 FROGGS_ right?
11:16 masak dunno. haven't thought much about this.
11:16 masak in some sense, it shouldn't be part of the name, because this is the process that underlies protoregexes.
11:17 FROGGS it's sad that you dont get the sym<> part when dumping a match though
11:17 masak std: sub foo:several<names>:at<once>() {}
11:17 p6eval std a8bc48f: OUTPUT«ok 00:00 43m␤»
11:18 masak FROGGS: yeah, that sounds lossy.
11:19 hoelzro I bring this up because token statement_control:sym<{ }> is one of the few things that my pygments lexer gets wrong
11:19 hoelzro and I'm trying to think of the best way to codify the rules
11:20 jnthn morning, #perl6
11:20 phenny jnthn: 01 Jan 18:33Z <FROGGS> tell jnthn that I have a better approach: https://github.com/FROGGS/rakudo/pull/new/nom
11:20 phenny jnthn: 01 Jan 18:34Z <FROGGS> tell jnthn I would like to make a new branch in the rakudo/rakudo repo, and would like to have commit bit for this, do I get your ok for this? I will also ask pmichaud
11:20 phenny jnthn: 01 Jan 21:05Z <[Coke]> tell jnthn I split up the qast tests, but I am not really confident about my organization.
11:20 phenny jnthn: 01 Jan 21:25Z <[Coke]> tell jnthn I have 'chars' added locally - it maps directly to String's length, but since that's an instance method, we can't map it directly like we did with the math methods, so I'm wrapping it in the ops class.
11:20 phenny Further messages sent privately
11:21 masak jnthn! \o/
11:22 jnthn phenny: tell FROGGS you linked me to the new pull request form ;) I'm OK with you having a commit bit; please drop pmichaud an email about it, since he has the info on received CLAs.
11:22 phenny jnthn: I'll pass that on when FROGGS is around.
11:22 masak hoelzro: if I were you I'd look at how STD.pm6 parses :sym
11:22 hoelzro ok, thanks
11:23 jnthn the foo:sym<bar> is just parsed by longname.
11:23 jnthn And sym is treated specially in the implementation of <sym>
11:24 masak so things can have "the same name" (from a proto/multi perspective), but different longnames? makes sense.
11:24 jnthn Yeah
11:24 jnthn We really do take the proto name and scan the method list for things that start with that_name:sym<
11:25 jnthn All colonpair names are canonicalized.
11:25 jnthn So if you write :sym['lol'] it is actually installed as :sym<lol>
11:27 masak hm... so using several colonpairs wouldn't work so well? :)
11:27 jnthn phenny: tell [Coke] the approach you took for chars is the right one (or right enoubh for now...) - we shouldn't bind ourselves too tightly to Java String to the abstraction there is welcome.
11:27 phenny jnthn: I'll pass that on when [Coke] is around.
11:28 jnthn phenny: tell [Coke] arrays are [Type, e.g. [Integer
11:28 phenny jnthn: I'll pass that on when [Coke] is around.
11:28 * masak hands jnthn a keyboard where the keys aren't right next to each other :)
11:29 masak I'm running *what* Niecza version!?
11:29 * masak updates
11:30 jnthn heh, I can't English
11:31 spider-mario joined #perl6
11:31 FROGGS jnthn: I gave you this link because you can see the diff there, but I dont want to make a pull request with these repo, wanna make a fresh and clean commit
11:31 phenny FROGGS: 11:22Z <jnthn> tell FROGGS you linked me to the new pull request form ;) I'm OK with you having a commit bit; please drop pmichaud an email about it, since he has the info on received CLAs.
11:32 FROGGS hmmm, I need his mail address then...
11:32 FROGGS ohh, faund it
11:32 FROGGS found
11:44 moritz \o
11:44 masak apparently, Quentin Tarantino has now made a movie about a Python MVC framework. I never thought I would live to see it.
11:44 masak moritz! \o/
11:44 moritz masak! \o/
11:46 jnthn o/ moritz
11:46 masak moritz: many happy ⏎s :)
11:47 moritz .u ⏎
11:47 phenny U+23CE RETURN SYMBOL (⏎)
11:47 moritz :-)
11:47 masak or maybe ䷗s
11:47 masak .u ䷗
11:47 phenny U+4DD7 HEXAGRAM FOR RETURN (䷗)
11:48 masak au++ # App::Uni
11:50 xinming joined #perl6
11:53 birdwindupbird joined #perl6
11:59 MayDaniel joined #perl6
12:05 dalek nqp-jvm-prep: e72ad7d | jonathan++ | Makefile:
12:05 dalek nqp-jvm-prep: Add a missing dependency.
12:05 dalek nqp-jvm-prep: review: https://github.com/jnthn/nqp​-jvm-prep/commit/e72ad7de59
12:05 dalek nqp-jvm-prep: 46250d9 | jonathan++ | t/qast_ (2 files):
12:05 dalek nqp-jvm-prep: Pull 6model tests out to a separate file.
12:05 dalek nqp-jvm-prep: review: https://github.com/jnthn/nqp​-jvm-prep/commit/46250d970c
12:05 dalek nqp-jvm-prep: ee7efde | jonathan++ | t/qast_ (2 files):
12:05 dalek nqp-jvm-prep: Split out aggregate tests also.
12:05 dalek nqp-jvm-prep: review: https://github.com/jnthn/nqp​-jvm-prep/commit/ee7efde0bc
12:06 jnthn coke++ # picking off lots of the LHF
12:06 au TimToady++ # original author; App::Uni is just an implementation of it
12:06 jnthn phenny: tell [Coke] tests breakup was good, I just did a few extra bits
12:06 phenny jnthn: I'll pass that on when [Coke] is around.
12:06 au . o O { "Implementation is the sincerest form of flattery." - L. Peter Deutsch }
12:07 dalek nqp-jvm-prep: b26a32d | jonathan++ | docs/ROADMAP:
12:07 dalek nqp-jvm-prep: Remove completed item.
12:07 dalek nqp-jvm-prep: review: https://github.com/jnthn/nqp​-jvm-prep/commit/b26a32d857
12:09 nwc10 which fruit is now the lowest hanging? (or am I asking a stupid question because you're updating that file as I type?)
12:10 jnthn nwc10: coke++ has been updating the LHF file. There's...not a lot left in it now. :)
12:11 jnthn ROADMAP gives the longer view of stuff to do
12:11 nwc10 aha
12:14 masak au: https://twitter.com/carlmasa​k/status/286445378038018050 :)
12:15 masak there's something slightly loopy about all this...
12:16 au you can say that again...
12:21 masak :P
12:22 masak nice, insightful comment on when to use right-associativity: http://arclanguage.org/item?id=16820
12:22 masak I like the rule "if it takes a Foo and a Bar and gives you a Bar in return". I'll keep on the lookout for that.
12:27 denisboyun_ joined #perl6
12:34 skids joined #perl6
12:39 hoelzro my pygments lexer successfully hilights all of Grammar.pm! \o/
12:39 hoelzro (with two minor bugs)
12:40 moritz \o/
12:40 moritz minor bugs are OK for a language as difficult to parse as Perl 6
12:42 hoelzro well, one is easy to fix
12:42 hoelzro it's the one I mentioned earlier in the backlog
12:42 hoelzro the other is role q { ... }
12:43 hoelzro which the lexer is parsing as the role keyword + a quote word string
12:44 hoelzro so I think what I'll do is fix the token sym: thing and announce an "alpha" of the lexer, so people can try it on their code, and see what breaks
12:44 arnsholt If it makes you feel better, it took quite a while before Rakudo realized that it's supposed to be a role decl, not a quote string too =)
12:44 arnsholt I think that particular bug was fixed with one of the improved LTM patches
12:45 hoelzro I just have to tell the role rule in the lexer to look for an identifier and a look-ahead brace
12:45 hoelzro it shouldn't be too bad!
12:46 hoelzro I was going to test it on Actions.pm too, but I think that Grammar.pm is probably a good enough test
12:47 moritz note again that rakudo's Grammar.pm is NQP, not Perl 6
12:47 hoelzro ah, true
12:48 arnsholt NQP's syntax is a subset of Perl 6 though, isn't it?
12:48 hoelzro well, the reason I picked Grammar.pm is because a) it's a large body of Perl 6 code and b) it has a lot of token/rule/regex decls
12:52 FROGGS ... and a lot of quoted and escaped things
12:54 hoelzro FROGGS: which made it a pretty good test =)
12:54 hoelzro I caught a *lot* of issues because of Grammar.pm
12:54 masak std: q { ... }
12:54 p6eval std a8bc48f: OUTPUT«ok 00:00 41m␤»
12:54 masak I can see how that one is tricky, yes... :)
12:56 hoelzro if anyone wants to try it out now, it's at https://bitbucket.org/hoelzro/pygments-main
13:03 moritz hoelzro: https://github.com/perl6/std/blob/master/STD.pm6 might be an even better test :-)
13:03 hoelzro heh
13:04 hoelzro yikes, 6300 lines?!
13:04 hoelzro 3800 was enough
13:04 masak arnsholt: it is if you squint.
13:04 masak arnsholt: about on the same level as JSON is a subset of JavaScript syntax.
13:05 arnsholt Right, right
13:19 _sri masak: JSON isn't really a subset of JavaScript http://timelessrepo.com/jso​n-isnt-a-javascript-subset
13:21 rindolf joined #perl6
13:21 masak _sri: ...which, if you read the context above, was really my point...
13:21 _sri ok :)
13:21 rindolf joined #perl6
13:22 _sri is there a list of incompatibilities between nqp and perl6?
13:22 masak none that I'm aware of.
13:23 masak nqp is fairly underdocumented, pending core dev tuits for such luxuries :)
13:24 _sri yea, guess it's rather unimportant
13:24 * _sri can't wait to see rakudo run on the jvm
13:25 arnsholt Me too!
13:25 wtw joined #perl6
13:25 arnsholt And when it does, I'm gonna try to submit a talk to the local Java developer's conference =D
13:26 arnsholt From the looks of it, I should probably aim for the 2014 edition
13:26 arnsholt Since this year's conference is early february
13:27 masak ++arnsholt
13:27 bbkr__ is there any built-in method to get hostname in p6?
13:29 masak not to my knowledge.
13:30 masak three hits for 'hostname' in S29.
13:30 masak seems 'gethost' might be what you're looking for.
13:30 masak (NYI anywhere)
13:31 arnsholt There's always Zavolaj =)
13:31 arnsholt Or maybe a Parrot builtin of some kind
13:34 FROGGS there is nothing in parrots config AFAICT
13:35 dalek doc: 5913946 | abcxyzp++ | lib/operators.pod:
13:35 dalek doc: Update lib/operators.pod
13:35 dalek doc:
13:35 dalek doc: One character typo fix.
13:35 dalek doc: review: https://github.com/perl6/doc/commit/5913946b15
13:35 dalek doc: cdb1cbb | (Konrad Borowski)++ | lib/operators.pod:
13:35 dalek doc: Merge pull request #12 from ghost/patch-1
13:35 dalek doc:
13:35 dalek doc: Update lib/operators.pod
13:35 dalek doc: review: https://github.com/perl6/doc/commit/cdb1cbb916
13:35 bbkr__ star: use NativeCall; sub gethostname(Str is rw, Int) returns Int is native { ... }; my $x = ""; gethostname($x, 32);
13:35 p6eval star 2012.12: OUTPUT«[31m===[0mSORRY![31m===[0m�Null PMC access in get_bool()�at /tmp/D_Ccua3CBR:1�------> �»
13:35 [Coke] jnthn: Danke.
13:35 phenny [Coke]: 11:27Z <jnthn> tell [Coke] the approach you took for chars is the right one (or right enoubh for now...) - we shouldn't bind ourselves too tightly to Java String to the abstraction there is welcome.
13:35 phenny [Coke]: 11:28Z <jnthn> tell [Coke] arrays are [Type, e.g. [Integer
13:35 phenny [Coke]: 12:06Z <jnthn> tell [Coke] tests breakup was good, I just did a few extra bits
13:37 FROGGS star: use NativeCall; sub gethostname( Str $a is rw, Int $n) returns Int is native { ... }; my $x = ""; say gethostname($x, 32); say $x
13:37 p6eval star 2012.12: OUTPUT«0␤␤»
13:39 arnsholt Oh, it wants to write to a buffer
13:39 arnsholt That doesn't work yet
13:39 FROGGS Inline::C would be your friend ;o)
13:40 arnsholt And I'm not sure if we can make it work 100% transparently
13:40 FROGGS (or an external c file)
13:40 arnsholt Strings in Parrot are immutable, so we can't really diddle with the Parrot strings themselves
13:40 arnsholt It might be possible to do some trickery with the containers though
13:42 jnthn arnsholt: Probably we need to do something with Buf
13:42 arnsholt Yeah, that's probably the best option
13:43 arnsholt Other option would be an operation similar to the write-barrier we do on complex objects coming out of a foreign function
13:43 arnsholt But that'd mean string compares on all function calls, which isn't going to be good for performance
13:47 bbkr__ ok, so if NativeCall way does not work yet can system command be executed? like $x = `hostname` in p5?
13:48 bbkr__ r: say `hostname`
13:48 p6eval rakudo 13e805: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Two terms in a rowâ�¤at /tmp/Z7HWQOLMZb:1â�¤------> [32msay [33mâ��[31m`hostname`[0mâ�¤    expecting any of:â�¤        argument listâ�¤        prefix or termâ�¤        prefix or meta-prefixâ�¤        postfixâ�¤        infix or meta-infixâ�¤        in…
13:48 jnthn try qx/hostname//
13:49 jnthn er, with a single / at the end
13:50 bbkr__ jnthn++ works, thanks!
13:57 rindolf joined #perl6
13:58 rindolf joined #perl6
14:00 kresike joined #perl6
14:00 kresike hello all you happy perl6 people
14:00 FROGGS hi kresike
14:01 kresike FROGGS, o/
14:03 rindolf kresike: hi.
14:03 [particle] joined #perl6
14:04 cognominal joined #perl6
14:07 atrodo joined #perl6
14:10 PacoAir joined #perl6
14:12 grondilu joined #perl6
14:17 kresike rindolf, o/
14:17 rindolf kresike: what's up?
14:18 kresike rindolf, well if I look up, I see the ceiling, so I'd say that's up :)
14:19 wtw joined #perl6
14:19 am0c joined #perl6
14:20 * rindolf grumbles.
14:21 ggoebel jnthn: with regard to jvm backend are you aware of JSR 292 (InvokeDynamic) and JSR 335 (project Lambda)?
14:21 ggoebel jnthn: http://irclog.perlgeek.de/​perl6/2013-01-02#i_6290048
14:29 masak rindolf: if you're not happy with such answers, would you please mind telling us what *is* up? :P
14:29 [Coke] ggoebel: neither of those are in a shipping java, neh?
14:29 au 292 is. 335 is in the next one
14:29 au JRuby uses 292 pretty heavily, as does Nashorn (Node.jar).
14:31 jnthn ggoebel: I'm aware of inovkedynamic; the JAST thing I have should already be able to emit it or can with relatively little work. I'm not using it just yet, but have some ideas on how to.
14:31 JJ_Brain joined #perl6
14:31 rindolf masak: well, I've tweaked the CSS on http://www.shlomifish.org/ so it won't overflow when the font size is enlarged too much.
14:31 jnthn ggoebel: 335 seems to be primarily about Java language changes, and I didn't yet see anything on changes being made to the JVM itself, which is what would be of more interest.
14:31 rindolf At least not without enlarging the images simultaneously.
14:32 rindolf And I tried E17 ZERO but it kept eating my Ctrl, Alt, etc. keys (at least in KDE apps) so I'm now back in KDE 4.
14:33 jnthn oh, seems that default methods cause some changes
14:34 au yeah, the semantic of invokeinterface and invokevirtual changed
14:38 kaleem joined #perl6
14:39 cognominal joined #perl6
14:46 JJ_Brain Hi. I may have solved P=NP.  Here is the link http://mathbin.net/133965
14:47 moritz JJ_Brain: please stop bothering us with that
14:47 JJ_Brain ok. sorry.
15:03 JJ_Brain joined #perl6
15:06 [particle]1 joined #perl6
15:31 * masak wishes he had thought of making http://www.regexper.com/
15:32 masak ooh, it uses Raphael.
15:35 [Coke] masak++ #nifty.
15:36 jnthn cute :)
15:37 masak clearly I need to investigate this RaphaelJS thing.
15:37 masak I mean, who's to say a presentation framework can't be built with it? :P
15:39 dalek roast: c039739 | moritz++ | S10-packages/basic.t:
15:39 dalek roast: RT #74592, return value from package declarations
15:39 dalek roast: review: https://github.com/perl6/roast/commit/c0397397da
15:39 kaare_ joined #perl6
15:40 moritz any idea on how best to fix  https://rt.perl.org/rt3/Ti​cket/Display.html?id=71544 ?
15:40 * jnthn has a look
15:40 moritz it would be nice if one could copy the signature from one block to another
15:41 moritz and use a \c param on that block, so that one only copies the constraints, not the binding
15:41 moritz or something like that
15:42 jnthn hm
15:42 moritz rakudo: say 1 if 2 if 3 { say 3 }
15:42 p6eval rakudo 13e805:  ( no output )
15:42 moritz std: say 1 if 2 if 3 { say 3 }
15:42 p6eval std a8bc48f: OUTPUT«[31m===[0mSORRY![31m===[0m�Missing semicolon at /tmp/yvUU8Qf3Ba line 1:�------> [32msay 1 if 2[33m�[31m if 3 { say 3 }[0m�Parse failed�FAILED 00:00 42m�»
15:43 jnthn moritz: If we want to cheat a little...
15:44 jnthn Signature caches arity/count rather than calcuating them each time.
15:44 moritz woah
15:44 moritz I see where the cheat is coming up :-)
15:44 jnthn Which menas you can...yeah, you guessed. :D
15:44 * jnthn feels a little dirty for suggesting it
15:45 jnthn But it should fix the bug since MapIter only cares for .count, iirc.
15:45 jnthn And the closure in grep is private to grep and passed straight on to MapIter.
15:46 moritz aye
15:46 jnthn We probably wanna shallow-clone the $!signature so two threads doing a grep at once don't hit issues
15:46 jnthn Or just keep a cache of those by arity.
15:47 jnthn The other way is to have a GrepIter, but it'd be laregly a copy paste of MapIter
15:48 masak it's only cheating if you get caught :)
15:48 [Coke] says the man who is going to catch them.
15:48 jnthn :P
15:50 isBEKaml joined #perl6
15:50 isBEKaml hello, #perl6!
15:52 masak isBEKaml! \o/
15:53 xinming_ joined #perl6
15:53 isBEKaml hola, masak! I see there was more activity on #perl6 than I thought. :)
15:54 pmichaud Happy New Year, #perl6
15:54 phenny pmichaud: 01 Jan 18:37Z <FROGGS> ask pmichaud if it would be possible to get commit bit for rakudo, my CLA is in walnut since november
15:54 masak pmichaud! \o/
15:54 [Coke] pmichaud: happy new year.
15:54 isBEKaml oh, yeah - Happy new year!
15:54 masak isBEKaml: yeah. there is activity. the bastards haven't ground us down yet. :)
15:54 pmichaud I'm fine with FROGGS getting a commit bit if nobody objects; I don't have his github id to do it myself.
15:54 pmichaud Yes, TPF received his CLA.
15:55 * masak was gonna ask :)
15:55 isBEKaml masak: if you're referring to doomsday naysayers, yeah - they got their calculations wrong. :)
15:55 FROGGS pmichaud: happy new year!
15:55 FROGGS I'm FROGGS there
15:55 moritz it's FROGGS, same as the nick
15:55 pmichaud FROGGS: adding.
15:55 FROGGS https://github.com/FROGGS/
15:55 FROGGS yay
15:55 jnthn happy new year, pmichaud :)
15:56 masak isBEKaml: no, it's just an expression. https://en.wikipedia.org/wik​i/Illegitimi_non_carborundum
15:56 isBEKaml jnthn: good to see some work going on nqp-jvm. jnthn++
15:56 dalek nqp-jvm-prep: 4557e1c | jonathan++ | docs/LHF.md:
15:56 dalek nqp-jvm-prep: Dangle some more LHF. :-)
15:56 dalek nqp-jvm-prep: review: https://github.com/jnthn/nqp​-jvm-prep/commit/4557e1c5fa
15:57 moritz rakudo: say :foo<<string\ here>>.perl
15:57 p6eval rakudo 13e805:  ( no output )
15:57 jnthn isBEKaml: coke++ also :)
15:57 pmichaud FROGGS: Added.
15:57 moritz r: say <<a\ b>>.perl
15:57 p6eval rakudo 13e805:  ( no output )
15:57 isBEKaml masak: Oh, I had never heard of that expression before.
15:57 FROGGS pmichaud: thanks
15:58 pmichaud all:  note that we don't require CLAs for commitbits to the star repo
15:58 jnthn [Coke]: 4557e1c ;-)
15:58 isBEKaml jnthn:  I remember adding a build.xml to the older nqp-jvm repo. Can I add it here too? Of course, all we need to do is add bcel jar to dependencies. :-)
15:58 Kharec_ joined #perl6
15:58 isBEKaml coke++
15:58 timotimo what's a CLA?
15:58 moritz pmichaud: so are star commit bits to be handed out roughly as freely as to the perl6/ repos?
15:59 PerlJam Happy new year all!
15:59 moritz timotimo: contributor's license agreement
15:59 pmichaud moritz: yes
15:59 moritz pmichaud: ok, noted
15:59 pmichaud it's just the rakudo repo that requires CLAs
15:59 timotimo oh, ok
15:59 jnthn isBEKaml: I know just about nothing about the Right Way to handle the building stuff with regard to Java projects. If you know about that, you're most welcome to a commit bit and to work on it :)
15:59 pmichaud I'd be fine with us changing the membership team names to reflect that, fwiw.
15:59 isBEKaml jnthn: https://github.com/svatsan/6model/b​lob/master/java/compiler/build.xml
15:59 brrt joined #perl6
15:59 [Coke] isBEKaml: Hurm?
15:59 [Coke] jnthn: Hey, I'm at $DAYJOB, stop that!
15:59 brrt wait, i understand why for loops are lazy unless when in sink context
16:00 isBEKaml jnthn: the problem with build stuff in java-land is that there are more ways to build it than I care to count. :-)
16:00 isBEKaml jnthn: for instance, ant, ivy, maven - the big three.
16:00 brrt suppose you say my @foo = for (1..200) { $_ * 2 }; then return()-ing doesn't make sense
16:00 brrt because, if you return from the routine, whatvalue will @foo have?
16:00 brrt it'd be nonsense
16:01 isBEKaml [Coke]: Keep shooting down those LHFs! :)
16:01 jnthn [Coke]: I'm gonna be traveling a lot for the next several days, so I figured it couldn't hurt to stock it up with ideas of things to work on before I go :)
16:01 [Coke] isBEKaml: I'd probably go with ant, given we're not pulling in anything from outside the repo.
16:01 isBEKaml [Coke]: but that would mean that we keep dependencies in source repo. I hope jnthn's okay with that.
16:02 brrt anyway, am i anywhere near the truth?
16:02 isBEKaml [Coke]: all binary blobs that are jars.
16:03 [Coke] I will happily use something else if it's setup. Note that we already have a 3rdparty dir with bundled jars.
16:03 isBEKaml Okay - 6model was 2 years ago. Really? (Time ran by fast. :( )
16:03 isBEKaml (I'm looking at my own repo)
16:04 [Coke] r: say(nqp::p6box_s(nqp::sha1("hey there, rakudo"));
16:04 p6eval rakudo 13e805:  ( no output )
16:05 isBEKaml r: say(nqp::p6box__s(nqp::sha1("Hey there")));
16:05 p6eval rakudo 13e805:  ( no output )
16:05 moritz r: say 42
16:05 p6eval rakudo 13e805:  ( no output )
16:05 * [Coke] thinks about half of those new ones look like LHF, and the others will require more thinking.
16:05 [Coke] moritz: ... nice golf.
16:06 jnthn isBEKaml: Lemme know your github ID if you want a commit bit for nqp-jvm-prep
16:06 [Coke] maybe I should give you a .... golf clap.
16:06 isBEKaml jnthn: not yet - I'm just afraid it might fall around like 6model. :(
16:07 [Coke] jnthn: ok, a good chunk of that really does look like LHF. :)
16:07 [Coke] thanks.
16:07 isBEKaml jnthn: the github link I gave you above was from my own repo - 2 years ago. :(
16:08 fgomez joined #perl6
16:09 jnthn isBEKaml: The 6model repo was largely a place to do resarch. That fed into building the pieces that Rakudo works on today, and considered how they could look on other VMs. What's being done in nqp-jvm-prep is an implementation of the things we need to run today's NQP and Rakudo on the JVM.
16:11 cognominal joined #perl6
16:12 isBEKaml jnthn: I understand that. I'm more concerned that I may not be able to contribute as much as I want to, if the past is anything to go by.
16:13 jnthn isBEKaml: ah, I see
16:13 jnthn isBEKaml: Well, just let me know if you fancy contributing something, and I'm happy to let you have a commit bit. :)
16:14 isBEKaml jnthn: sure, if something comes out of my eclipse IDE. :)
16:15 jnthn FROGGS: $foreign_proto := self.find_symbol([$_.key]); - is the foreign dispatcher not just be $_.value?
16:16 jnthn I mean, the thing that's in scope when you do the find_symbol call is the same thing as the target proto, I think.
16:16 isBEKaml joined #perl6
16:18 isBEKaml jnthn: in terms of the ROADMAP, where are we currently now?
16:18 * masak .oO( not in Kansas anymore )
16:19 isBEKaml lol
16:20 masak don't let the wicked witch grind you down.
16:22 isBEKaml masak: $day-job's playing that role now.
16:22 moritz r: my $x := $(); say $x.perl
16:22 p6eval rakudo 13e805:  ( no output )
16:23 isBEKaml r: my $x = $(); $x.WHAT.say;
16:23 p6eval rakudo 13e805:  ( no output )
16:23 moritz nr: $_ = 'abc'; m/(.)/; say $)(
16:23 p6eval rakudo 13e805:  ( no output )
16:23 p6eval ..niecza v24-12-g8e50362: OUTPUT«[31m===[0mSORRY![31m​===[0mâ�¤â�¤Non-declarative sigil is missing its name at /tmp/OKdZ4FdoPw line 1:â�¤------> [32m$_ = 'abc'; m/(.)/; say [33mâ��[31m$)([0mâ�¤â�¤Unsupported use of bare 'say'; in Perl 6 please use .say if you meant $_, or use an explicit invoc…
16:23 moritz nr: $_ = 'abc'; m/(.)/; say $()
16:23 p6eval rakudo 13e805:  ( no output )
16:23 p6eval ..niecza v24-12-g8e50362: OUTPUT«a␤»
16:23 isBEKaml it's really broken now?
16:25 jnthn isBEKaml: https://github.com/jnthn/nqp-jv​m-prep/blob/master/docs/ROADMAP is most of the stuff that needs building to get most of what's needed for cross-compiling NQP in place
16:25 moritz feather1's disc has run full
16:26 isBEKaml jnthn: so, we are still laying the bricks, but no cement?
16:26 moritz everybody with a feather account, please check your home dirs
16:27 pmichaud moritz: checking
16:27 bluescreen10 joined #perl6
16:28 jnthn isBEKaml: Depends what you consider as cement :)
16:28 moritz pmichaud: yours looks not too big, with 84M
16:28 jnthn .oO( ain't the cement meant to go between the bricks... )
16:28 pmichaud moritz: it was 335M  :-)
16:28 pmichaud I just removed 251M
16:28 moritz pmichaud: thanks
16:28 isBEKaml jnthn: yeah, we have nothing to stand on. :-)
16:28 moritz ... and something is using up all that free space immediately
16:29 pmichaud the 84M is the synopsis subversion dump from when we converted  svn->git
16:29 pmichaud I'll keep it around just-in-case for a bit longer
16:29 moritz pmichaud: +1
16:29 [Coke] 11584 azawawi   20   0  5556 1484 1252 R 96.3  0.1 158:45.48 perl                ?
16:29 moritz usually when the disc is full, it's one or two home dirs that are really huge
16:29 isBEKaml jnthn: about your makefile, we depend on nqp to build and test nqp-jvm?
16:30 moritz [Coke]: your home dir is 5.1G, do you need that all?
16:30 arnsholt If space is immediately eaten up, maybe you have a runaway process?
16:30 jnthn isBEKaml: Yes
16:30 [Coke] moritz: I'm poking now.
16:30 jnthn isBEKaml: nqp by now is written almost entirely in nqp
16:30 isBEKaml moritz: I'm not sure, how much swap space is there now?
16:30 arnsholt If you have lsof, that might  give you a pointer
16:30 moritz isBEKaml: is that relevant?
16:31 isBEKaml moritz: If we are talking about runaway processes. Like I said, I don't have any picture on this.
16:31 jnthn isBEKaml: The idea being that we'll get a cross-compiler going then getting it to spit out an nqp on JVM, that has no dependency on NQP on Parrot. :)
16:31 moritz can I ask lsof to only list FDs opened for writing?
16:32 isBEKaml jnthn: well, I was more of the impression that we have to have an nqp2jvm translator.
16:32 jnthn isBEKaml: Not really. At the moment it is a QAST -> JVM translator. :)
16:32 isBEKaml jnthn: that is, translate nqp code directly to run on jvm :-) (naive, maybe)
16:33 cognominal joined #perl6
16:33 arnsholt moritz: From a skim of the man page, it seems that the FD output item describes the mode (read, write, ...) of an fd as well as its number
16:33 pmurias joined #perl6
16:35 thou joined #perl6
16:35 isBEKaml jnthn: I have your project open in eclipse and am source-diving now. :)
16:36 moritz oh dammit
16:36 pmichaud -s to lsof shows file size, if that helps.
16:36 moritz if a still-open file handle uses up all that space, du won't show it, right?
16:36 [Coke] is there any point to more pruning of my homedir at this point?
16:37 pmichaud moritz: correct, du doesn't find files not linked into the tree
16:37 moritz [Coke]: just on general principle
16:37 pmichaud s/tree/filesystem
16:43 [Coke] is azawawi here?
16:43 moritz http://moritz.faui2k3.org/tmp/lsof.out # outptu of lsof -s|grep home
16:44 pmichaud /tmp/UIR5AJNtCP
16:44 pmichaud rootfs is also at 100%
16:44 moritz but that's not mounted on /home
16:44 * moritz deletes that temp file
16:45 pmichaud I'm pretty sure it's the azawawi process that must be eating up space
16:45 pmichaud there's also a long-running "gitit" process running.
16:45 moritz gitit is our wiki, no?
16:45 [Coke] pmichaud: he's running a temp file that shouldn't really be taking all that long.
16:45 moritz killed
16:46 pmichaud moritz:  I don't know anything about gitit
16:46 pmichaud other than it's showing up at the top of 'top'
16:46 [Coke] and now we're at 97%
16:46 pmichaud and nothing appears to be eating up the free space
16:47 pmichaud and  root, /var, and /home all appear to be the same filesystem or device somehow
16:47 moritz phenny: tell azawawi that I killed a process of his on feather that filled up the file systems. Please be more careful in future (ulimit is your friend)
16:47 phenny moritz: I'll pass that on when azawawi is around.
16:47 pmichaud (see output of 'df')
16:48 [Coke] moritz: I'll clean up more homedir space after lunch.
16:48 kresike bye folks
16:48 moritz [Coke]: thanks
16:48 * moritz runs apt-get autoclean
16:49 pmichaud fwiw, the command I used to find the big temp file was   'lsof -s | grep -v "/lib/" | less'   and then search for azawawi
16:49 FROGGS jnthn: I got errors when using $_.value instead of self.find_symbol([$_.key]), but yes, I thought too that these should be the same
16:51 Pleiades` joined #perl6
16:53 jnthn FROGGS: What errors, ooc?
16:53 * jnthn tracks down RT#116268 to an optimizer issue
16:55 rindolf FROGGS: hi.
17:00 isBEKaml what's .WHO meant to do?
17:01 isBEKaml r: Int.WHO.say
17:01 p6eval rakudo 13e805:  ( no output )
17:03 sftp joined #perl6
17:04 masak r: say Int.WHO
17:04 p6eval rakudo 13e805:  ( no output )
17:04 masak hrm.
17:05 masak r: module M {}; say M.WHO
17:05 p6eval rakudo 13e805:  ( no output )
17:05 isBEKaml in my local run, it always returns ().hash
17:05 nyuszika7h joined #perl6
17:05 wtw joined #perl6
17:06 jnthn .WHO returns the underlying Stash (which is a kind of Hash)
17:06 jnthn The place where our things inside the package live
17:07 masak r: module M { our $foo }; say M.WHO
17:07 p6eval rakudo 13e805:  ( no output )
17:07 masak the stash is very... silent, though.
17:07 jnthn masak: p6eval's Rakudo build currently busted due to disk space issue
17:07 masak oh!
17:07 isBEKaml masak: I thought you already knew that.
17:08 isBEKaml (p6eval build brokenness)
17:08 REPLeffect joined #perl6
17:09 isBEKaml jnthn: S12 says that it's used to disambiguate methods between classes and traits. That's as far as my english skills go. :)
17:09 moritz http://tonyarcieri.com/dci-i​n-ruby-is-completely-broken # seems that in most (all) ruby VMs, mixing a role into an instance invalidates all method cashes
17:10 masak isBEKaml: I saw people talk about feather in the backlog, but I didn't connect the wires, it seems.
17:11 moritz on feather, it seems that /var/spool/qmail is huge
17:11 * isBEKaml thinks slow masak-ness is not something we see often. :P
17:11 moritz I've been running a 'du' process on it for several minutes now, and now response yet
17:12 masak isBEKaml: if that is so, it's only because I hide it well most of the time :P
17:12 moritz and load average is >7
17:21 wtw joined #perl6
17:23 dalek rakudo/nom: 0f0bb50 | jonathan++ | src/binder/bind.c:
17:23 dalek rakudo/nom: Teach compile time bind analysis about (|) sig.
17:23 dalek rakudo/nom:
17:23 dalek rakudo/nom: It now recognizes such a signature as accepting anything, and thus
17:23 dalek rakudo/nom: certain to be a successful match.
17:23 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0f0bb50054
17:23 dalek rakudo/nom: f63c42d | jonathan++ | src/Perl6/Optimizer.pm:
17:23 dalek rakudo/nom: Check proto as well as candidates in optimizer.
17:23 dalek rakudo/nom:
17:23 dalek rakudo/nom: This fixes issues where we could end up inlining calls to candidates
17:23 dalek rakudo/nom: that would never be reachable due to restrictions in the proto. Also,
17:23 dalek rakudo/nom: we catch some such issues at compile time now.
17:23 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/f63c42d9a9
17:23 dalek rakudo/nom: 799d759 | jonathan++ | src/core/Numeric.pm:
17:23 dalek rakudo/nom: Fix incorrect exp proto.
17:23 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/799d759c27
17:24 REPLeffect joined #perl6
17:30 moritz r: say 42
17:31 moritz ok, /var/spool/qmail on feather is 16G
17:31 FROGGS jnthn: give me a few minutes, I'll paste then...
17:31 p6eval joined #perl6
17:31 FROGGS rindolf: hi
17:31 rindolf FROGGS: what's up?
17:31 moritz any idea what to do about that?
17:31 moritz I guess rm -rf isn't such a good idea, is it?
17:32 p6eval rakudo : OUTPUT«Can't exec "./nom-inst/bin/perl6": No such file or directory at lib/EvalbotExecuter.pm line 200.␤exec (./nom-inst/bin/perl6 --setting=RESTRICTED /tmp/a6Y4pYPKQM) failed: No such file or directory␤Server error occurred! Closing Link: feather3.perl6.nl (Quit: p6eval)␤Lost …
17:33 FROGGS rindolf: the kids (but only just), and perl 6 of course
17:33 rindolf FROGGS: OK.
17:34 FROGGS grammars a bit to parse perl 6 code, and bughunting
17:38 moritz rn: say 42
17:39 moritz sorry, I guess I'll have to postpone fixing p6eval until feather has normal-ish performance characteristics again
17:39 moritz and after my table tennis training :-)
17:39 p6eval joined #perl6
17:39 moritz though I won't complain if anybody beat me to it (fixing feather, that is :-)
17:40 moritz bbi3h
17:40 moritz &
17:40 cognominal joined #perl6
17:40 djanatyn joined #perl6
17:40 p6eval ..niecza v24-12-g8e50362: OUTPUT«42␤»
17:44 raiph joined #perl6
17:45 p6eval joined #perl6
17:50 dalek rakudo/nom: bd41d2f | jonathan++ | src/Perl6/Optimizer.pm:
17:50 dalek rakudo/nom: Improve error reporting when proto is to blame.
17:50 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/bd41d2f6a0
17:51 regreg joined #perl6
17:54 tadzik moritz: re feather diskspace, do you know of ncdu?
17:54 timotimo ncdu is a nice tool, i use it often
17:55 tadzik apparently it's even installed, my $HOME is ~750 meg
17:55 tadzik was, actually :)
17:55 jnthn std: say "You can even call &uc('functions')!"
17:55 p6eval std a8bc48f: OUTPUT«ok 00:00 44m␤»
18:00 isBEKaml jnthn: I almost got SHA1 working. (yes, "almost") :)
18:04 jnthn :)
18:05 Chillance joined #perl6
18:12 cognominal joined #perl6
18:15 jnthn afk for a bit
18:15 FROGGS jnthn: here it is: https://gist.github.com/4436609
18:21 Kharec joined #perl6
18:26 Targen joined #perl6
18:26 MikeFair__ joined #perl6
18:33 ismail joined #perl6
18:43 JJ_Brain Thanks Raptor
18:47 rindolf joined #perl6
18:48 PacoAir joined #perl6
18:48 [Coke] argh, other people stealing my LHF! ;)
18:49 [Coke] It really shouldn't take minutes for 'ccache -C' to clear my cache, should it? ;(
18:51 isBEKaml [Coke]: were you looking to implement sha1?
18:51 marloshouse joined #perl6
18:51 isBEKaml [Coke]: You can take that if you want to. I'm just too tired today or I don't see anything at all. :)
18:51 flussence [Coke]: depends on the filesystem. It's basically doing an rm -rf
18:52 [Coke] isBEKaml: nono, that's fine, take it!
18:52 [Coke] I'm stuck at $DAYJOB anyway.
18:52 isBEKaml [Coke]: Ah, day-jobs eat up all our time. :(
18:53 [Coke] spent last 2 days of break happily poking at nqp-jvm-prep as time permitted.
18:55 thou joined #perl6
18:55 isBEKaml hmm, I only have minutes to count before I drop off. (Another day wasted)
19:05 rindolf joined #perl6
19:05 ifim joined #perl6
19:07 rindolf joined #perl6
19:08 SamuraiJack joined #perl6
19:11 isBEKaml now, where do we have tests for sha1 in nqp?
19:27 [Coke] anyone know if it's safe to rm perlbrew's build/ dirs?
19:28 [Coke] (uninstall didn't remove them.)
19:38 MikeFair_ joined #perl6
19:41 masak nice Rob Pike talk about lexing. http://www.youtube.com/watch?v=HxaD_trXwRE summary: don't write state machines as switch statements, that's boring. instead, just have lexing functions HOPly return each other.
19:43 flussence H?
19:43 geekosaur higher order programming
19:43 [Coke] Higher Order Programmingly.
19:44 flussence oh, duh. my brain decided "OP" was "-Oriented Programming" and got stuck :)
19:44 isBEKaml jnthn: Here's my stab at sha1. Admittedly, all indentation is screwed, sleepiness and everything. https://github.com/svatsan/nqp-jvm-prep/comm​it/fd9587caf7b8ac3e348246fdb0481fabacf9c3b5 :)
19:45 isBEKaml masak: I read that as "Happily" return each other. :|
19:47 [Coke] isBEKaml: are you using eclipse?
19:47 isBEKaml [Coke]: no,vim.
19:47 isBEKaml [Coke]: I originally did that in eclipse and I found that I needed to set up git. :/
19:48 flussence an odd idea: is it possible to take a list of subs and attempt to use them as multis?
19:48 isBEKaml (on windows), though I had it in cygwin. (so I switched to vim)
19:48 [Coke] haivng all the whitespace changes makes it hard to digest that.
19:49 flussence (by "list", I mean an actual runtime-assigned array of &code things)
19:49 isBEKaml [Coke]: yeah, I know - here are the basic changes: https://gist.github.com/4437043
19:49 [Coke] other than that, +1, though.
19:49 isBEKaml and follow down to the last one: in qast_string.t
19:59 [Coke] someone bored? take over http://search.cpan.org/~tobix/OO-DateTime-0.01/ and kill it.
20:07 masak [Coke]: what do you mean "and kill it"? there must be some assumed procedure I'm missing here.
20:08 FROGGS well, once you are the owner you could write to ANDK and let him remove the registered namespace I guess
20:09 isBEKaml masak: "kill boredom", mayhaps. :P
20:09 zby_home joined #perl6
20:09 masak isBEKaml: :)
20:16 MikeFair_ joined #perl6
20:27 isBEKaml :q
20:27 * hugme hugs isBEKaml, good vi(m) user!
20:28 isBEKaml :q
20:28 * hugme hugs isBEKaml, good vi(m) user!
20:28 cognominal joined #perl6
20:33 dalek rakudo/froggs_mergemulti: 763a621 | (Tobias Leich)++ | src/Perl6/World.pm:
20:33 dalek rakudo/froggs_mergemulti: possible fix for multi-method import bug
20:33 dalek rakudo/froggs_mergemulti:
20:33 dalek rakudo/froggs_mergemulti: This patch allows you to import multi methods from several modules.
20:33 dalek rakudo/froggs_mergemulti: It merges the dispachtee lists. Two identical dispatchee signatures
20:33 dalek rakudo/froggs_mergemulti: dont clash yet, since I am unable to compare signatures.
20:33 dalek rakudo/froggs_mergemulti: review: https://github.com/rakudo/rakudo/commit/763a62173c
20:36 [Coke] masak: remove it from CPAN is what I meant (so people don't confuse it with DateTime, as I just did)
20:36 [Coke] Date::Time is just a pod file, that's it.
20:36 masak ah.
20:37 Santosh joined #perl6
20:38 PacoAir joined #perl6
20:39 Guest10671 left #perl6
20:41 SantoshR joined #perl6
20:49 VIVIsectI joined #perl6
20:58 cognominal joined #perl6
21:00 Pleiades` joined #perl6
21:05 raiph joined #perl6
21:14 * jnthn back
21:16 [Coke] jnthn: there's a pull request. I was going to try to tease out the whitespace formatting, but don't really care. Not sure if you do. (if we do, we should add project settings to enforce something.)
21:18 jnthn I'd rather it was consistently formatted
21:18 jnthn Also, it shouldn't just go printing stuff to stderr, if there's an exception it should be thrown.
21:19 [Coke] wasn't sure if we needed to rethrow there or not.
21:20 [Coke] just let the original exception out?
21:20 jnthn Can do that also
21:20 jnthn s/also//
21:22 tadzik obtw, jnthn, do you have your AOP module around somewhere?
21:22 tadzik I thought today, now that there are threads and semaphores, this one is lines away from being a Monitor implementation possibly :)
21:23 jnthn tadzik: I don't think I ever made it into a real module, but the code is probably easy to steal from the PDF
21:25 snearch joined #perl6
21:30 masak ++tadzik # publishing jnthn's AOP module :)
21:31 jnthn ;0
21:31 tadzik oh noes, I'm doomed
21:35 moritz rn: say 42
21:35 p6eval rakudo bd41d2, niecza v24-12-g8e50362: OUTPUT«42␤»
21:35 dalek rakudo/nom: f5037ef | moritz++ | / (3 files):
21:35 dalek rakudo/nom: remove "SAFE" setting
21:35 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/f5037efe39
21:35 moritz \o/ p6eval fixed
21:37 jnthn \o/
21:37 jnthn Righty, let's see if I can get anywhere with attributes in nqp-jvm...
21:37 moritz I didn't do anything for that now, seems the cronjobs did it for me :-)
21:38 masak the cronjobs made a commit?
21:38 masak wow.
21:38 moritz no, fixing p6eval :-)
21:38 masak oh, right.
21:39 masak by the way, we should totally have cronjobs making commits for us :)
21:39 moritz I think [Coke]++ has
21:39 masak anyway, clearly I'm too tired to IRC.
21:39 masak 'night, #perl6
21:39 jnthn 'night, masak
21:39 moritz good night. Dream of not being tired.
21:39 alex_Perl_new joined #perl6
21:41 alex_Perl_new Is Python Perl6?
21:42 tadzik hugme: hug alex_Perl_new
21:42 * hugme hugs alex_Perl_new
21:42 alex_Perl_new Can anyone tell me if Python is perl6?
21:42 sorear No
21:42 tadzik welcome
21:43 sorear Python is not Perl 6
21:43 alex_Perl_new Thanks tadzik
21:43 alex_Perl_new thanks sorear
21:44 moritz nr: say 'Python' eq 'Perl6'
21:44 p6eval rakudo bd41d2, niecza v24-12-g8e50362: OUTPUT«False␤»
21:46 alex_Perl_new I Installed Rakudo but I could not find out how to instal an editor that could help with my Syntax. After 4 days I am going Crazy.
21:46 moritz use vim
21:46 moritz it has Perl 6 syntax hilighting
21:46 tadzik alex_Perl_new: do you just need syntax hilighting, or something more sophisticated?
21:47 tadzik that's a tricky questions, as I don't think anything more sophisticated is available :)
21:47 moritz padre has some more sophisticated stuff, I think
21:48 alex_Perl_new I think just hilithing would work, I have found many ways but can get them installed for windows. I think I just do not know much and Brut force is not working.
21:48 MikeFair_ joined #perl6
21:49 alex_Perl_new I even had problem to get on the IRC channel. I was using Browser. Finaly Instaled IRC App.
21:50 * moritz hugs alex_Perl_new
21:50 moritz root@feather:~# qmail-qstat
21:50 moritz messages in queue: 1179019
21:50 moritz messages in queue but not yet preprocessed: 1179019
21:51 tadzik whoa
21:51 jnthn yowser!
21:52 jnthn Guess that's where the 16 gigs are :)
21:52 moritz indeed
21:53 moritz root@feather:~# /etc/init.d/qmail stop
21:53 moritz Stopping mail-transfer agent: qmail not running.
21:53 alex_Perl_new To All: I think here you all are just at a higher level than me. At this time I have been reading and reading and did not get any place yet. Finaly some one said get Python as it coms Complete. Use it Learn it then go to perl6.
21:54 alex_Perl_new Is that correct? I cant even do simple thing like prodo::xxxx
21:55 moritz wtf is prodo::xxxx ?
21:55 alex_Perl_new do not even know were to type it.
21:57 alex_Perl_new I whish very much to undrestand "wtf is prodo::xxxx ?" Where do I Start to learn perl <period> ? lol
21:57 jnthn alex_Perl_new: Perl 6 isn't yet at the point where things are packaged up in the way you're seeking. If you're looking for getting into Perl then Perl 5 may be easier in some senses, since there's for example the Padre IDE, which probably gives you the experience you are after.
21:58 jnthn alex_Perl_new: And I'd expect Perl 5 => Perl 6 will be an easier jump than Python => Perl 6 if that's your end goal.
21:59 alex_Perl_new <alex_Perl_new> What do you recomend I am a good programer I just lost to get started.
21:59 [Coke] moritz: yes, I have cronjobs that commit for me. I do the push, though.
22:00 * moritz just deletes all the large files in the qmail queue
22:00 jnthn alex_Perl_new: See http://padre.perlide.org/download.html
22:00 jnthn The windows download on the left.
22:01 alex_Perl_new Thanks jntn I will do that.
22:01 flussence alex_Perl_new: what is this "prodo::xxxx" you speak of? http://www.google.com/search?q=%22prodo::xxxx%22
22:03 am0c joined #perl6
22:04 alex_Perl_new jnthn, Should I remove Rakudo and Activeperl and python before installing  DWIM Perl 5.14.2.1 (v7).
22:04 moritz alex_Perl_new: no need for that
22:06 alex_Perl_new Maritz, Would they not conflect with each othe. and in the Command line of Windows when i enter perl -v what would it say and How does it know wiich perl?
22:06 moritz they do not conflict.
22:07 moritz alex_Perl_new: most or all of the Rakudo developers use Perl 5 too, so we had incentive to make them not conflict
22:08 jnthn alex_Perl_new: Note that Rakudo's executable is called perl6.exe, and python's is probably python.exe, so no risk there :)
22:09 alex_Perl_new Thanks to all I hope as time passes I can learn to be helpful to others here. I am going to try now.
22:10 moritz something is really strange on feather
22:10 moritz I've been running rm -rf on the qmail queue (which is 16G in size) for several minutes
22:10 moritz and the load goes up to 7
22:10 flussence is it on ext3?
22:10 moritz but 'df' doesn't report more free space
22:10 moritz flussence: yes
22:11 flussence it's probably sorting through a gigantic maildir and everything else is getting blocked on IO
22:11 moritz you mean it sorts before it starts to delete?
22:12 [Coke] moritz: is it the same thing were somethig else is grabbing space as you open it?
22:12 flussence well, it has to read the directory before deleting... and if it's an old machine it probably doesn't have dir_index enabled so that's a linear scan each time
22:21 [Coke] might explain why it was so slow to clear out my ~/.ccache
22:23 [Coke] Glitchmr is holding at 17% Mem util.
22:24 [Coke] p6eval should probably nice itself when building.
22:24 moritz aye, it should
22:25 * moritz -> sleep
22:30 [Coke] ugh. feather is behaving oddly. "ls .cpanm" hangs, ^C doesn't kill it.
22:30 flussence try ^\
22:31 alex_Perl_new left #perl6
22:32 [Coke] p6eval is no longer building, and now things are much better.
22:33 [Coke] and ^\ doesn't help when things are hanging in this case, unfortunately.
22:33 [Coke] (not immediately, anyway)
22:37 bitonic joined #perl6
22:38 bitonic is there a go-to document to learn about perl6?  I know next to 0 about perl5.
22:39 colomon joined #perl6
22:40 FROGGS bitonic: get the book: https://github.com/perl6/book/downloads
22:40 bitonic FROGGS: thank you.  I suppose that’s going to be good with the rakudo shipped with wheezy?
22:41 bitonic rakudo 0.1~2012.01-1 specifically
22:41 FROGGS well, these are a bit outdated so some things might not work
22:41 FROGGS wow, there are huge improvements since then
22:42 bitonic OK, so I should get a newer one...
22:42 FROGGS right
22:42 FROGGS bitonic: maybe install this: https://github.com/rakudo/star
22:42 FROGGS it comes with some useful modules
22:43 bitonic FROGGS: thanks
22:43 FROGGS bitonic: if you arn't used to git you might download a zipball from here: http://rakudo.org/
22:43 FROGGS bitonic: you're welcome
22:44 bitonic FROGGS: I am used to scm, but I might download something else to get something more stable then an HEAD
23:15 skids joined #perl6

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

Perl 6 | Reference Documentation | Rakudo