Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2014-04-02

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:26 raiph joined #perl6
00:33 jnap joined #perl6
00:41 colomon joined #perl6
00:56 japhb r-m freezes trying to 'panda install p6doc' at:  Compiling lib/Perl6/TypeGraph/Viz.pm to mbc
01:12 iolympian joined #perl6
01:12 iolympian left #perl6
01:18 jnap joined #perl6
01:34 lustlife joined #perl6
01:35 woosley joined #perl6
01:40 xenoterracide joined #perl6
01:40 lustlife joined #perl6
01:44 FROGGS joined #perl6
01:47 klapperl joined #perl6
02:03 * timotimo is experimenting with docker
02:03 benabik joined #perl6
02:04 japhb r: say "$_" for < a b >; say "{$_}" for < c d >;
02:04 camelia rakudo-parrot 358582: OUTPUT«a␤b␤c␤d␤»
02:04 camelia ..rakudo-jvm 358582, rakudo-moar 358582: OUTPUT«a␤b␤␤␤»
02:05 japhb ^^ Why the difference here?
02:05 timotimo oh yikes
02:06 japhb Yeah, that led to a rather strange bug, let me tell you.
02:06 timotimo i can imagine
02:06 timotimo but i don't have time to hunt that right now :\
02:07 japhb .tell jnthn Two r-m bugs: http://irclog.perlgeek.de/perl6/2014-04-02#i_8527156 and http://irclog.perlgeek.de/perl6/2014-04-02#i_8527351
02:07 yoleaux japhb: I'll pass your message to jnthn.
02:07 colomon joined #perl6
02:07 japhb timotimo: Understood.
02:13 xenoterracide joined #perl6
02:14 japhb .tell moritz Since I wasn't around when you announced it, congrats re: http://irclog.perlgeek.de/perl6/2014-03-31#i_8519218  :-)
02:14 yoleaux japhb: I'll pass your message to moritz.
02:17 BenGoldberg joined #perl6
02:21 colomon joined #perl6
02:25 yeltzooo joined #perl6
02:26 japhb .tell jnthn On a positive note, your r-j hang-on-exit fix did indeed work for me.  :-)
02:26 yoleaux japhb: I'll pass your message to jnthn.
02:30 xragnar joined #perl6
02:54 xenoterracide joined #perl6
03:15 colomon joined #perl6
03:23 ssutch joined #perl6
03:25 ssutch joined #perl6
03:27 jnap joined #perl6
03:38 rurban1 joined #perl6
04:03 Ben_Goldberg joined #perl6
04:04 kaare_ joined #perl6
04:16 BenGoldberg_ joined #perl6
04:28 jnap joined #perl6
04:30 Psyche^ joined #perl6
04:48 thou joined #perl6
04:52 woolfy joined #perl6
05:28 jnap joined #perl6
05:30 aborazmeh joined #perl6
05:34 kurahaupo joined #perl6
05:37 zby_home_ joined #perl6
05:38 [Sno] joined #perl6
05:55 moritz japhb: thanks
05:55 yoleaux 02:14Z <japhb> moritz: Since I wasn't around when you announced it, congrats re: http://irclog.perlgeek.de/perl6/2014-03-31#i_8519218  :-)
05:58 woolfy left #perl6
05:59 lizmat oohh, wow, congrats moritz++ and mom!
06:02 moritz lizmat: thanks
06:03 rurban1 joined #perl6
06:05 rurban2 joined #perl6
06:06 arnsholt joined #perl6
06:13 lizmat decommute&
06:29 jnap joined #perl6
06:30 kaleem joined #perl6
06:32 [Sno]_ joined #perl6
06:32 denisboyun joined #perl6
06:55 zakharyas joined #perl6
07:01 arnsholt joined #perl6
07:01 arnsholt o/
07:09 WJB joined #perl6
07:10 rurban1 joined #perl6
07:21 dmol joined #perl6
07:28 kurahaupo joined #perl6
07:30 jnap joined #perl6
07:44 FROGGS joined #perl6
07:45 FROGGS morning
07:45 arnsholt Moin
07:46 nwc10 \o/
07:56 hummeleB1 joined #perl6
08:05 dmol1 joined #perl6
08:13 fhelmberger joined #perl6
08:27 virtualsue joined #perl6
08:30 ashleydev joined #perl6
08:31 jnap joined #perl6
08:31 dakkar joined #perl6
08:32 darutoko joined #perl6
08:40 jnthn morning o/
08:40 yoleaux 02:07Z <japhb> jnthn: Two r-m bugs: http://irclog.perlgeek.de/perl6/2014-04-02#i_8527156 and http://irclog.perlgeek.de/perl6/2014-04-02#i_8527351
08:40 yoleaux 02:26Z <japhb> jnthn: On a positive note, your r-j hang-on-exit fix did indeed work for me.  :-)
08:41 jnthn .tell japhb yes, the panda p6doc thing is known to fail install on both JVM and MoarVM. It's not entirely clear why yet, but the failures are likely related.
08:41 yoleaux jnthn: I'll pass your message to japhb.
08:42 jnthn I'm surprised the other one didn't trigger a spectest...
08:54 pecastro joined #perl6
09:16 rurban1 joined #perl6
09:23 aborazmeh joined #perl6
09:26 SHODAN joined #perl6
09:32 jnap joined #perl6
09:36 dalek star: d381a68 | (Tobias Leich)++ | modules/ (2 files):
09:36 dalek star: update submodules to newer revision
09:36 dalek star: review: https://github.com/rakudo/star/commit/d381a68147
09:39 dalek star: f9a0c4c | (Tobias Leich)++ | tools/star/release-guide.pod:
09:39 dalek star: no special handling needed, PR got applied
09:39 dalek star: review: https://github.com/rakudo/star/commit/f9a0c4c607
09:48 Guest6533 joined #perl6
09:51 rindolf joined #perl6
09:57 rindolf joined #perl6
10:05 iolympian1 joined #perl6
10:07 denis_boyun joined #perl6
10:08 bjz joined #perl6
10:17 woosley left #perl6
10:17 rurban1 joined #perl6
10:17 woosley joined #perl6
10:25 pecastro joined #perl6
10:32 jnap joined #perl6
10:43 colomon joined #perl6
11:05 daniel-s_ joined #perl6
11:09 iolympian1 left #perl6
11:33 jnap joined #perl6
11:58 telex joined #perl6
12:09 daniel-s_ joined #perl6
12:26 FROGGS TheDamian++ # http://www.infoq.com/interviews/conway-perl
12:27 FROGGS seems to be more than a year old, but is still very very good
12:34 jnap joined #perl6
12:39 Ven r: FROGGS "“Isn’t Pearl dead?” - What do you answer to that?" erm :(
12:39 camelia rakudo-parrot 358582, rakudo-jvm 358582, rakudo-moar 358582: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/tmpfileâ�¤Two terms in a rowâ�¤at /tmp/tmpfile:1â�¤------> [32ml dead?â€� - What do you answer to that?" [33mâ��[31merm :([0mâ�¤    expecting any of:â�¤ …»
12:39 moritz Ven: Mu
12:40 Ven Seems like my brain is thinking "this is #perl6, I need to prefix my message with r:" sorry moritz
12:40 FROGGS *g*
12:40 FROGGS but yeah, that typo is slightly annoying
12:40 espadrine joined #perl6
12:41 moritz Ven: I actually meant that 'Mu' is a good answer to such a question
12:41 Ven moritz: Then I'll just keep on wondering what that means
12:41 moritz compare http://en.wikipedia.org/wiki/Mu_%28negative%29
12:42 moritz in particular http://en.wikipedia.org/wiki/Mu_%28negative%29#.22Unasking.22_the_question
12:43 Ven moritz: I'm only talking about the typo, I don't care about the question haha ... Didn't realize that's what you were talking about
12:43 d0nk joined #perl6
12:45 xenoterracide joined #perl6
12:47 mzgcz joined #perl6
12:48 [Coke] FROGGS: I think the typo is intentional. "this user knows so little about perl they misspell it"
12:48 [Coke] sorry, I meant PERL, of course.
12:48 [Coke] That's a shibboleth for me when looking at resumes. we don't perl here, but I don't want people who claim to have PERL'ed.
12:49 FROGGS yeah, that might be true
12:50 daniel-s_ joined #perl6
12:51 Ven "no system is as good as Haskell’s type system, but also unlike Haskell’s type system, ordinary human beings can understand it" I get sad when I read that
12:51 [Coke] booking.com++ # perl5 core maint fund
12:51 Ven [Coke]: might be, yeah
12:51 [Coke] Ven: I loved that quote!
12:51 [Coke] also, if you speak haskell, do you want to try to get pugs back up on crutches?
12:52 Ven [Coke]: I'm not sure it's even kinda up-to-date enough for it not to blow up with recent ghc updates
12:52 Ven also, I'm not smart enough
12:54 * Ven notes - that looks like a really neat project to have, though ...
12:55 [Coke] ven - if you speak haskell, you're already a step ahead of me.
12:55 [Coke] let me grab a snapshot of what the issue looks like here.
12:55 dmol joined #perl6
12:59 guru joined #perl6
13:01 FROGGS Ven: I tried to fix a bug in pugs like two months ago, and I was able to build it but was not able to fix the bug
13:05 Ven FROGGS: well, I'm just waiting for the day to end anyway, do you remember what the bug was ?
13:05 FROGGS ohh, now I remember...
13:05 FROGGS it wasn't even a bug, I wanted to change sub "eval" to "EVAL"
13:06 FROGGS but [Coke] has the problem that his pugs installation does not run the spectests...
13:06 FROGGS so I dunno what is the problem there
13:06 iolympian joined #perl6
13:06 iolympian left #perl6
13:07 Ven FROGGS: you mean it refuses to *run* the spectests themselves?
13:08 [Coke] I'm getting the build output, one sec.
13:08 [Coke] no, it runs them, they just nearly all fail.
13:09 FROGGS Impl date percentage roast sha1 pass fail todo skip plan spec notes
13:09 FROGGS pugs 2014-04-01 00.06% 2f754e1 19 14733 0 1 14753 28848
13:09 FROGGS pugs 2014-02-03 21.86% a4574df 6331 4691 2884 856 14762 28451
13:09 sqirrel joined #perl6
13:09 FROGGS so, it passes >6k at one point
13:09 FROGGS passed*
13:09 Ven 20% for an implementation that's been dead since 2009 seems good ;o)
13:09 Ven (not really sure about the date, sry)
13:11 FROGGS I have heard about 2007, but that was way before my time here :o)
13:11 [Coke] lots of deprecation warnings...
13:11 FROGGS it passed 9.1k tests in may 2012
13:11 Ven (I actually work on an altJS where au is a contrib, though)
13:12 Ven That's still cool
13:13 Ven "it will probably be based on ideas from languages like Erlang" mmh ?
13:14 [Coke] "this version of PCRE compiled without utf8 support" is oft-repeated.
13:14 [Coke] s/utf8/UTF/
13:17 * [Coke] assumes he hasn't gone back and tried to fix this lately.
13:18 Util r: my @x = [<a b>], [<c d>]; say .perl for [@x.list], [@x.flat], [@x>>.flat], [@x>>.list];
13:18 camelia rakudo-parrot 358582, rakudo-jvm 358582, rakudo-moar 358582: OUTPUT«[["a", "b"], ["c", "d"]]␤[["a", "b"], ["c", "d"]]␤["a", "b", "c", "d"]␤["a", "b", "c", "d"]␤»
13:18 Util How can I flatten @x into <a b c d> ?
13:18 Util I have tried .list and .flat as methods and hyper-methods.
13:21 TimToady it appears the latter two worked, but you put an extra [] around them after
13:21 Ven also, @x.list returns an array ?
13:21 Ven r: say [].^HOW.name;
13:21 camelia rakudo-parrot 358582, rakudo-jvm 358582, rakudo-moar 358582: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/tmpfileâ�¤Cannot use .^ on a non-identifier method callâ�¤at /tmp/tmpfile:1â�¤------> [32msay [].^HOW[33mâ��[31m.name;[0mâ�¤    expecting any of:â�¤     …»
13:22 TimToady HOW is already ^-ish
13:22 TimToady doesn't need ^
13:22 Ven r: say ([1]).HOW.name;
13:22 camelia rakudo-moar 358582: OUTPUT«Not enough positional parameters passed; got 1 but expected 2␤  in any name at src/gen/m-Metamodel.nqp:69␤  in block  at /tmp/tmpfile:1␤␤»
13:22 camelia ..rakudo-jvm 358582: OUTPUT«Wrong number of arguments passed; expected 2..2, but got 1␤  in block  at /tmp/tmpfile:1␤␤»
13:22 camelia ..rakudo-parrot 358582: OUTPUT«too few positional arguments: 1 passed, 2 (or more) expected␤  in any name at gen/parrot/Metamodel.nqp:66␤  in block  at /tmp/tmpfile:1␤␤»
13:23 Util TimToady: Thanks! (/me needs to `git pull` and re-make)
13:23 TimToady say [1].^name
13:23 TimToady r: say [1].^name
13:23 camelia rakudo-parrot 358582, rakudo-jvm 358582, rakudo-moar 358582: OUTPUT«Array␤»
13:23 Ven ;( thanks
13:23 Ven r: [1].list.^name
13:24 camelia rakudo-parrot 358582, rakudo-jvm 358582, rakudo-moar 358582: ( no output )
13:24 Ven r: say [1].list.^name
13:24 camelia rakudo-parrot 358582, rakudo-jvm 358582, rakudo-moar 358582: OUTPUT«Array␤»
13:24 TimToady r: say [1].list.WHAT
13:24 Ven Looks weird to me,though
13:24 camelia rakudo-parrot 358582, rakudo-jvm 358582, rakudo-moar 358582: OUTPUT«(Array)␤»
13:24 Util I was working from local:    Jan 20 12:54 perl6-p
13:25 TimToady that's like...ancient...
13:26 [Coke] Ven: http://feather.perl6.nl/~coke/PUGS_BUILD.txt
13:27 [Coke] the PCRE bug is the most obvious issue, googling..
13:27 arnsholtAtHome joined #perl6
13:27 arnsholtAtHome o/
13:29 [Coke] I don't see how to have cabal tell me what versions I have installed locally. :P
13:33 [Coke] ven: any any tips on how to diagnose that issue?
13:35 jnap joined #perl6
13:37 bluescreen10 joined #perl6
13:45 kbaker_ joined #perl6
13:45 FROGGS hi arnsholtAtHome
13:45 [Coke] (or anyone)
13:45 [Coke] .seen au?
13:45 yoleaux I haven't seen au? around.
13:45 [Coke] .seen au
13:45 yoleaux I haven't seen au around.
13:45 [Coke] yoleaux: help
13:46 FROGGS TimToady: I've heard you are/were travelling around and give talks... are any of them recorded?
13:46 [Coke] .mangle is perl six the best perl?
13:46 yoleaux River, Pearl used?
13:47 mmachenry joined #perl6
13:47 TimToady FROGGS: they had a camera there, dunno anything else
13:47 FROGGS hmmm... do they have a website? :o)
13:48 FROGGS nvm, I've seen a website... I was just not able to read it
13:49 [Coke] .gcs "perl" "perl 5" "perl 6"
13:49 yoleaux [site] "perl" (32,000,000), "perl 6" (872,000), "perl 5" (366,000)
13:49 [Coke] .gcs "rakudo"
13:49 yoleaux [site] "rakudo" (186,000)
13:50 FROGGS .gcs "peril"
13:50 yoleaux [site] "peril" (6,930,000)
13:50 FROGGS /o\
13:50 [Coke] we can't search that, it's too dangerous.
13:51 dalek rakudo-star-daily: 7718ffb | coke++ | log/ (5 files):
13:51 dalek rakudo-star-daily: today (automated commit)
13:51 dalek rakudo-star-daily: review: https://github.com/coke/rakudo-star-daily/commit/7718ffbdb0
13:51 dalek rakudo-star-daily: 9937ef2 | coke++ | log/ (5 files):
13:51 dalek rakudo-star-daily: today (automated commit)
13:51 dalek rakudo-star-daily: review: https://github.com/coke/rakudo-star-daily/commit/9937ef27fc
13:52 FROGGS +Makefile:45: *** missing separator.  Stop.
13:52 FROGGS :/
13:53 Woodi hi + :)
13:54 FROGGS hi Woodi
13:55 Woodi new R* ! I was missing it for some time... but it's just emotions :)
13:55 btyler joined #perl6
13:57 Woodi I hope OpenLDAP bindings would work via NativeCalls
13:58 Woodi and can we have { ... } are export sugar ?
13:58 FROGGS ++Woodi # I'd like to be your tester!
13:59 FROGGS hmmm, we have nothing like that atm... so I'd say not in the near future
13:59 FROGGS ahh, btw:
13:59 Woodi brave FROGGS :) Just yesterday I realy understand programming is not realy for me :)
13:59 FROGGS TimToady: when we autogenerate a proto because we just declare a multi, should the proto inherit some of the traits?
14:00 FROGGS TimToady: I am think of "is hidden_from_backtrace" for example
14:00 FROGGS Woodi: bah... don't you give up! :o)
14:00 FROGGS thinking*
14:04 Util TimToady: `perl6-p` updated from git HEAD; problem resolved. Thanks!
14:04 rurban1 joined #perl6
14:06 Woodi FROGGS: and few days before my brain decided: "C++ ! this is the language ! why so furiously stick to C ? with C++ all GUI frameworks can be used ! even Perl6 should be done in C++ !"
14:06 FROGGS ohh noes!
14:07 Woodi but I try and now trying WxPerl with some success :)
14:07 Woodi Perl - it's so nice and easy language :)
14:07 FROGGS btw, do we meet in Poznan in may?
14:08 jnap joined #perl6
14:09 Woodi hmm, probably not if you ask Woodi :)
14:09 FROGGS Woodi: why not?
14:09 ascent_ what's in Poznan in May?
14:10 retupmoca on rakudo-moar, $?FILE is a relative path. Is there a way to get the absolute path to the current file?
14:10 FROGGS ascent_: the polish perl workshop
14:10 retupmoca IO::Path.absolute just gets me $*CWD ~ $?FILE
14:10 Woodi I didn't know that
14:10 jnap joined #perl6
14:10 Woodi now I will try to mess a bit there :)
14:11 * jnthn hasn't been to Poznan before. :)
14:11 FROGGS me neither :o)
14:11 FROGGS I was in Poland once but just because I was driving wrongish
14:12 Woodi heh :)
14:12 jnthn Driving wrongish isn't allowed in Germany, so you went to Poland to do it instead? :P
14:12 FROGGS m: say $*PROGRAM
14:12 camelia rakudo-moar 358582: OUTPUT«===SORRY!===␤Dynamic variable $*PROGRAM not found␤»
14:13 Woodi we are allowed to drive limit+10km/h unofficially :)
14:13 retupmoca My use case: I want to load a library.so file that's in the same directory as the installed library.pm6
14:13 FROGGS m: say $*PROGRAM_NAME
14:13 camelia rakudo-moar 358582: OUTPUT«/tmp/Rr4xo9uSSC␤»
14:14 FROGGS m: say $*PROGRAM_NAME.path.directory # retupmoca>
14:14 camelia rakudo-moar 358582: OUTPUT«/tmp␤»
14:14 FROGGS dunno what that does in a .pm though
14:15 jnthn Same; it's about the original program
14:15 FROGGS it perhaps gives you still the script that invoked it
14:15 jnthn Is there a %*INC for "hwere did modules we used come from"?
14:15 FROGGS if we had CompUnitRepos in nom you could query that :o)
14:15 treehug88 joined #perl6
14:16 retupmoca if we had $?RESOURCES in nom/panda master I could use that instead :)
14:18 jnthn FROGGS: How far off are having CompUnitRepos in nom? :)
14:18 retupmoca yeah, it looks like $*PROGRAM_NAME just gives the .pl file
14:19 retupmoca even when used in a .pm
14:19 jnthn $* means dynamically coped
14:19 jnthn *scoped
14:19 retupmoca yeah, that's what I expected
14:19 jnthn So not so surprising.
14:21 FROGGS jnthn: we would need to discuss its "design"
14:22 FROGGS and then do some testing, but there is not that much left to do for it
14:22 salv0 joined #perl6
14:22 Woodi btw. R* is annoingly long... IMO "Pleas note..." and all below should be refactored to separate "todo" page...
14:23 aborazmeh joined #perl6
14:23 Woodi /R\*/R* release text/
14:24 Ven [Coke]: I'm back -- sorry ! Looking at it right now
14:25 hoverboard joined #perl6
14:28 Ven Lots and lots of deprecation warning, uh
14:28 hoverboard joined #perl6
14:29 pecastro joined #perl6
14:29 benabik joined #perl6
14:30 retupmoca so, I can scan @*INC to find the .so file
14:30 retupmoca next problem: is native($path-to-library) gets compiled in staticly
14:31 retupmoca so I get "Cannot locate native library '.../.work/1396448921_1/blib/lib/...'
14:31 FROGGS mhpf
14:31 Ven Never seen that "this version of PCRE is compiled without UTF support". Google results blame apache but that's with php so meh
14:31 retupmoca even when $path-to-library is completely different
14:32 Woodi R* README: "To build R* you need at least a C compiler, a 'make' utility, and Perl 6.1.0 or newer."  Hmm ? 5.16 ?
14:32 FROGGS ohh
14:32 FROGGS Woodi: that was me
14:33 FROGGS there was a 5.9.0 before and that was the Parrot revision we had in the past...
14:33 FROGGS damnit
14:34 dalek star: 0757d26 | (Tobias Leich)++ | README:
14:34 dalek star: fixed thinko in version bump, Woodi++
14:34 dalek star: review: https://github.com/rakudo/star/commit/0757d2671e
14:34 Woodi also no MoarVM mentioning in README
14:35 Woodi can I run R* on Moar now ?
14:39 FROGGS if you wait a minute, I fix a vcs conflict in the makefile
14:39 FROGGS Woodi: a few modules will fail tests though
14:40 FROGGS and one module (something::graphviz::something) will hang when you build it, just ctrl+c it
14:41 dalek star: f79eaed | (Tobias Leich)++ | tools/build/Makefile.in:
14:41 dalek star: fix VCS conflict
14:41 dalek star: review: https://github.com/rakudo/star/commit/f79eaed04e
14:41 jnap1 joined #perl6
14:41 FROGGS Woodi: let me give you a release tarball where you can build moar...
14:42 Woodi there is INSTALL.txt in rakudo/
14:42 woolfy joined #perl6
14:42 Woodi FROGGS: I wgetted one from rakudo.org...
14:42 FROGGS Woodi: this one does not build for moarvm
14:44 FROGGS I've just made a release tarball that contains MoarVM/nqp/rakudo HEAD as of now
14:44 FROGGS there are quite some fixes since 2014.03
14:44 Woodi I see rakudo/src/vm/ but it will not build ?
14:45 FROGGS Woodi: you are able to build the compiler, but not the modules I think
14:46 Woodi means: not automatically/via panda or at all ?
14:46 FROGGS Woodi: it does not use panda, no
14:46 FROGGS damn
14:46 FROGGS that tarball I created will not build rakudo, because of an outdated MANIFEST :/
14:47 Woodi upload the .patch :)
14:49 Hugh joined #perl6
14:50 jnthn train &
14:52 Ven [Coke]: Seems like its' missing that HsBridge.pm, though. (even if you have some custom code to deal with it)
14:53 [Coke] s/you/pugs/ :)
14:53 spider-mario joined #perl6
14:54 Ven [Coke]: haha yeah, that's what I mean : `unshift @INC, '/home/coke/.cabal/share/Pugs-6.2.13.20120717/blib6/pugs/perl5/lib';`
14:58 [Coke] *blank stare* Ok.
14:59 Woodi actually I would be happy to just have the list of 3 or 4 urls to parot/moar/nqp/rakudo .tgz files and build them as usual on *nix systems... Packaging R* could have such nice side effect, declaration: "R vX recuires: Parrot vY, Moar vZ and NQP vAA"
15:00 kaleem joined #perl6
15:01 FROGGS http://moarvm.org/releases/MoarVM-2014.03.tar.gz
15:01 FROGGS http://rakudo.org/downloads/rakudo/
15:01 FROGGS http://rakudo.org/downloads/rakudo/rakudo-2014.03.01.tar.gz
15:02 FROGGS http://rakudo.org/downloads/nqp/nqp-2014.03.tar.gz
15:03 FROGGS http://ftp.parrot.org/releases/all/6.1.0/parrot-6.1.0.tar.gz
15:03 xinming_ joined #perl6
15:04 guru joined #perl6
15:05 FROGGS Woodi: and finally: http://froggs.de/perl6/rakudo-star-2014-04-02-TRIAL.tar.gz
15:06 FROGGS bbiab
15:06 colomon joined #perl6
15:07 Ven [Coke]: I'll try to understand why that error pops up, though I can't say anything atm :(
15:11 FROGGS[mobile] joined #perl6
15:11 Util moritz++ (quite literally!) # http://irclog.perlgeek.de/perl6/2014-03-31#i_8519218
15:16 Ven thanks FROGGS
15:18 moritz thanks Util :-)
15:21 colomon joined #perl6
15:24 kivutar joined #perl6
15:25 rindolf joined #perl6
15:25 pmurias joined #perl6
15:29 lizmat joined #perl6
15:48 SamuraiJack__ joined #perl6
16:04 btyler one of the things TheDamien mentions in the interview linked earlier is something like "my Str $foo where length > 20"; does that (or something like that) actually exist in rakudo?
16:09 FROGGS[mobile] it would be chars instead of length, but I've not seen it
16:10 FROGGS[mobile] m: my Str $foo where chars > 20
16:10 camelia rakudo-moar 358582: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/UMpCNXHQmn�Preceding context expects a term, but found infix > instead�at /tmp/UMpCNXHQmn:1�------> [32mmy Str $foo where chars >[33m�[31m 20[0m�»
16:11 FROGGS[mobile] m: my Str $foo where { *.chars > 20}
16:11 camelia rakudo-moar 358582: ( no output )
16:11 Ven m: my Str $foo where .chars > 20 # should work ?
16:11 camelia rakudo-moar 358582: ( no output )
16:11 FROGGS[mobile] can you assign something?
16:12 FROGGS[mobile] it is hard to type on a phone :o)
16:12 Ven m: my Str $foo where .chars > 20; my $foo = "abc; say "alive"; my $foo = "x" x 50; say "alive"; # should work ?
16:12 btyler m: my Str $foo where .chars < 5; $foo = 'abc';
16:12 camelia rakudo-moar 358582: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Two terms in a rowâ�¤at /tmp/v3HNxstdST:1â�¤------> [32mwhere .chars > 20; my $foo = "abc; say "[33mâ��[31malive"; my $foo = "x" x 50; say "alive";[0mâ�¤    expecting any of:â�¤        postfixâ�¤        infix stopperâ�¤â€¦Â»
16:12 camelia rakudo-moar 358582: ( no output )
16:12 Ven m: my Str $foo where .chars > 20; my $foo = "abc"; say "alive"; my $foo = "x" x 50; say "alive"; # should work ?
16:12 camelia rakudo-moar 358582: OUTPUT«Potential difficulties:â�¤    Redeclaration of symbol $fooâ�¤    at /tmp/jmly5TGVk_:1â�¤    ------> [32mmy Str $foo where .chars > 20; my $foo [33mâ��[31m= "abc"; say "alive"; my $foo = "x" x 50[0mâ�¤    Redeclaration of symbol $fooâ�¤    at /tmp/jmly5…»
16:13 Ven m: my Str $foo where .chars > 20; $foo = "abc"; say "alive"; $foo = "x" x 50; say "alive"; # ;_;
16:13 camelia rakudo-moar 358582: OUTPUT«alive␤alive␤»
16:13 Ven m: my Str $foo where .chars > 20; $foo = "abc"; say "alive - $foo.chars()"; $foo = "x" x 50; say "alive - $foo.chars()"; # let's see
16:13 camelia rakudo-moar 358582: OUTPUT«alive - 3␤alive - 50␤»
16:13 Ven FROGGS[mobile]: apparently doesn't check
16:14 FROGGS[mobile] perhaps that only works in sub or method calls
16:14 Ven m: sub a(Str where .chars > 20); a('a'); say "alive"; a('a' x 50); say 'alive';
16:14 camelia rakudo-moar 358582: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/WOAv1Nj6kAâ�¤Cannot do non-typename cases of type_constraint yetâ�¤at /tmp/WOAv1Nj6kA:1â�¤------> [32msub a(Str where .chars > 20[33mâ��[31m); a('a'); say "alive"; a('a' x 50); say[0mâ�¤    expe…»
16:14 Ven m: sub a(Str $foo where { *.chars > 20}) {}; a('a'); say "alive"; a('a' x 50); say 'alive';
16:14 camelia rakudo-moar 358582: OUTPUT«alive␤alive␤»
16:15 Ven FROGGS[mobile]: apparently no
16:17 FROGGS[mobile] someone needs to implement that then :o)
16:19 btyler I saw that in the interview and thought "wait whaaaaaat. that is too cool!"
16:21 FROGGS[mobile] m: sub a($foo where {$_.chars > 20} ) {}; a "bar"
16:21 camelia rakudo-moar 358582: OUTPUT«Constraint type check failed for parameter '$foo'␤  in sub a at /tmp/n7UVxcehma:1␤  in block  at /tmp/n7UVxcehma:1␤␤»
16:21 FROGGS[mobile] \o/
16:21 FROGGS[mobile] note: no whatever star in a block
16:22 FROGGS[mobile] sub a($foo where *.chars > 20) {}; a "foo"
16:23 FROGGS[mobile] damn -.-
16:23 FROGGS[mobile] can somebody paste that to a m: ?
16:23 lizmat m: sub a($foo where *.chars > 20) {}; a "foo"
16:23 camelia rakudo-moar 358582: OUTPUT«Constraint type check failed for parameter '$foo'␤  in sub a at /tmp/vB_KHslriv:1␤  in block  at /tmp/vB_KHslriv:1␤␤»
16:23 lizmat cycling&
16:23 FROGGS[mobile] thanks!
16:24 FROGGS[mobile] so it works as it should in subs
16:24 btyler hm, so it works in sub parameter checks, but not for regular declarations
16:24 btyler cool
16:29 FROGGS joined #perl6
16:30 FROGGS m: :($foo where *.chars > 20); $foo = "abc"
16:30 camelia rakudo-moar 358582: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/s43Y15YPnDâ�¤Variable '$foo' is not declaredâ�¤at /tmp/s43Y15YPnD:1â�¤------> [32m:($foo where *.chars > 20); $foo[33mâ��[31m = "abc"[0mâ�¤    expecting any of:â�¤        postfixâ�¤Â»
16:31 FROGGS :/
16:33 colomon joined #perl6
16:37 denis_boyun joined #perl6
16:39 [Sno] joined #perl6
16:41 cognominal joined #perl6
16:50 mmachenry joined #perl6
16:55 Ven FROGGS: uuh, matching signatures with blocks gotta be nasty
17:04 colomon joined #perl6
17:15 virtualsue joined #perl6
17:18 mmachenry joined #perl6
17:19 zakharyas joined #perl6
17:20 guru joined #perl6
17:33 sivoais_ joined #perl6
17:38 vendethiel joined #perl6
17:40 dalek perl6-roast-data: 7752803 | coke++ | / (6 files):
17:40 dalek perl6-roast-data: today (automated commit)
17:40 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/775280379e
17:41 [Coke] yay, no more hangs. If someone is bored, they could add the ulimit stuff to the harnesses for the remaining implementations.
17:42 [Coke] r.j - 3 fails, r.m - 5 fails, r.p - 179 fails.
17:42 [Coke] All the r.p fails are aborts in the concurrency tests.
17:42 [Coke] can probably fudge those to skip all if r.p
17:43 [Coke] pugs refuses to die at 00.06%
17:45 colomon \o/
17:46 treehug8_ joined #perl6
17:48 FROGGS [Coke]: btw, your star tests should run tomorrow again
17:48 Yappo__________8 joined #perl6
17:48 [Coke] FROGGS: they run every day?
17:49 [Coke] I can do another run today if that bug is fixed.
17:50 FROGGS [Coke]: "rakudo-star-daily" implies every day, no? :o)
17:50 FROGGS but yeah, the bug is fixed
17:50 [Coke] oh, I see. you meant "should run and work this time"
17:51 [Coke] I gets it now.
17:51 FROGGS ahh
17:51 FROGGS k
17:52 [Coke] I kicked off a run right now, will post results when it finishes.
17:52 FROGGS cool!
17:54 kurahaupo joined #perl6
18:00 xinming__ joined #perl6
18:02 retupmoca jnthn: I posted a PR for NativeCall when you get a chance
18:03 * vendethiel opens [Coke]'s log to try and look what's going on for pugs now that he's home
18:19 dalek ecosystem: e8dd3c5 | (Andrew Egeler)++ | META.list:
18:19 dalek ecosystem: Add LibraryMake
18:19 dalek ecosystem:
18:19 dalek ecosystem: An attempt to simplify native compilation
18:19 dalek ecosystem: review: https://github.com/perl6/ecosystem/commit/e8dd3c5416
18:28 isBEKaml joined #perl6
18:31 dalek rakudo-star-daily: b55e7d5 | coke++ | log/ (5 files):
18:31 dalek rakudo-star-daily: today (automated commit)
18:31 dalek rakudo-star-daily:
18:31 dalek rakudo-star-daily: second run to grab bugfix.
18:31 dalek rakudo-star-daily: review: https://github.com/coke/rakudo-star-daily/commit/b55e7d5dcb
18:32 FROGGS nice, all module pass
18:32 hoverboard joined #perl6
18:34 hoelzro FROGGS: on which backend?
18:35 FROGGS parrot
18:35 mmachenry joined #perl6
18:36 hoelzro I figured =/
18:36 hoelzro MIME::Base64 just stopped working for me on moar
18:36 FROGGS hmmm, but it should work... I works on my boc
18:36 FROGGS box
18:37 FROGGS it*
18:37 hoelzro curious
18:37 hoelzro I'll try it at home
18:37 hoelzro when I get there =/
18:45 timotimo i like the sound of "LibraryMake"
18:48 retupmoca timotimo: https://github.com/retupmoca/P6-Auth-PAM-Simple for usage example
18:49 retupmoca (needs a NativeCall patch to actually install that library, though)
18:50 timotimo oh cool, pam authentication
18:51 timotimo hm, would we need one module for each PAM method, like fingerprint, smartcard, ldap, ... ?
18:51 retupmoca well, Auth::PAM::Simple has an arg for the pam service to auth against
18:52 retupmoca but honestly I have only a vaguely understand PAM. I just copy-pasted some C code so I could auth against the 'login' service
18:52 retupmoca s/have only a/only/
18:54 timotimo oh, i didn't know you could re-route the library name of a sub at run-time
18:54 timotimo that's cool
18:54 retupmoca timotimo: that's a NativeCall PR I have up
18:54 retupmoca doesn't work with HEAD yet
18:55 timotimo ah
18:55 timotimo yeah, i like that
18:55 retupmoca and after the PR is merged, doing &sub.libname = '...'; after you call it won't do anything
18:55 retupmoca because the library has already been loaded
18:56 retupmoca but this is my plan for fixing Imlib2
19:00 timotimo fair enough
19:00 timotimo how exactly does it have to be fixed?
19:01 retupmoca need to remove the existing Makefile and Build.pm and replace it with stuff that uses LibraryMake
19:02 retupmoca then just need to add code to find the native library like Auth::PAM::Simple uses
19:02 retupmoca there could be issues with the module itself, but this will get the native library compiled and loaded anyway
19:02 retupmoca which is where it's breaking right now
19:03 timotimo well, i'm very glad you've created this
19:04 retupmoca LibraryMake.pm6 needs a couple of patches for parrot and JVM if you want to help (I'm not sure how to get the config for '-l' for them)
19:04 sqirrel joined #perl6
19:04 timotimo i haven't looked at that part of the whole thing yet :(
19:04 retupmoca for moar it's $*VM<config><ldusr>, and then strip out the %s
19:04 timotimo other than "inspect $*VM<config>", no clue
19:04 retupmoca gotcha
19:04 retupmoca I may have to actually install rakudo-j and -p then
19:05 timotimo fortunately it's easy :)
19:05 timotimo however
19:05 timotimo you can get the output of that right here on the channel
19:05 retupmoca oh, right
19:06 retupmoca p: for $*VM<config>.kv -> $k, $v { if $v ~~ /\-l/ { say "$k: $v" } }
19:06 camelia rakudo-parrot 358582: OUTPUT«archname: x86_64-linux-gnu-thread-multi␤ccwarn: -falign-functions=16 -funit-at-a-time -fexcess-precision=standard -maccumulate-outgoing-args -Wall -Wextra -Waggregate-return -Wcast-qual -Wdisabled-optimization -Wdiv-by-zero -Wendif-labels -Wformat-extr…»
19:06 FROGGS that gives good hints: https://github.com/jnthn/zavolaj/blob/master/t/CompileTestLib.pm
19:06 retupmoca p: for $*VM<config>.kv -> $k, $v { if $v ~~ /^\-l/ { say "$k: $v" } }
19:06 camelia rakudo-parrot 358582: OUTPUT«libs: -ldl -lm -lpthread -lcrypt -lrt -lgmp -lreadline  -lffi  ␤»
19:06 retupmoca FROGGS: that's what I used for everything else
19:06 FROGGS k
19:07 retupmoca FROGGS: but that doesn't load any libraries of it's own, so it doesn't give me -l
19:07 FROGGS it uses perllibs for example
19:07 retupmoca j: for $*VM<config>.kv -> $k, $v { if $v ~~ /^\-l/ { say "$k: $v" } }
19:07 camelia rakudo-jvm 358582: OUTPUT«nativecall.libs: -lgdbm -lgdbm_compat -ldb -ldl -lm -lpthread -lc -lcrypt␤nativecall.perllibs: -ldl -lm -lpthread -lc -lcrypt␤»
19:07 retupmoca m: for $*VM<config>.kv -> $k, $v { if $v ~~ /^\-l/ { say "$k: $v" } }
19:07 camelia rakudo-moar 358582: OUTPUT«ldlibs: -lm -lpthread -lrt -ldl␤ldusr: -l%s␤ldsys: -l%s␤»
19:07 FROGGS j: say $*VM<config><nativecall.perllibs>
19:07 camelia rakudo-jvm 358582: OUTPUT«-ldl -lm -lpthread -lc -lcrypt␤»
19:08 retupmoca I need the '-l' by itself
19:08 FROGGS do not use libs
19:08 hoverboard joined #perl6
19:08 retupmoca looks like parrot and jvm don't have that :/
19:08 FROGGS retupmoca: what do you mean?
19:08 FROGGS a pattern or something?
19:08 retupmoca m: say $*VM<config><ldusr>
19:08 camelia rakudo-moar 358582: OUTPUT«-l%s␤»
19:08 retupmoca something like that
19:09 retupmoca so I can say "Link against pam", which on my machine is 'gcc ... -lpam'
19:09 dwarring joined #perl6
19:09 FROGGS isn't it always the same?
19:10 FROGGS I mean, we are talking about dynamic linking only
19:11 jiraffe joined #perl6
19:11 jiraffe left #perl6
19:11 retupmoca I thought there were some windows compilers that used some other argument for "Link this library in"
19:11 retupmoca all I know is GNU stuff
19:13 timotimo i think it's weirder on mac os x
19:14 FROGGS retupmoca: make it work for your system, and others care for their... (i.e. I have windows and linux)
19:14 FROGGS ... and an osx vm somewhere
19:14 geekosaur you're not going to be using PAM on windows anyway
19:14 FROGGS why not?
19:14 retupmoca geekosaur: right, but you will hopefully be using LibraryMake on windows
19:15 geekosaur PAM is only useful if the system uses it, since it's plugins for system authentication/uthorization
19:15 geekosaur Windows does not use PAM for that
19:15 FROGGS ahh, now I understand
19:15 geekosaur (Windows has GINA which is a bit different)
19:15 geekosaur (among other things(
19:16 FROGGS m: nqp::how
19:16 camelia rakudo-moar 358582: OUTPUT«===SORRY!===␤Error while compiling op how: Arg count 0 doesn't equal required operand count 2 for op 'gethow'␤»
19:26 retupmoca Anyway, LibraryMake patches are welcome, and I'll probably try and make Imlib2 use it in the next couple of days if noone beats me to it
19:28 timotimo it's funny; i was about to want to use imlib2 for something fun recently
19:29 retupmoca timotimo: I think that's what started me down this whole LibraryMake path
19:29 timotimo oh!
19:29 retupmoca you mentioned it in here and I started poking at it
19:30 timotimo \o/
19:30 retupmoca \o/ indeed
19:31 sivoais joined #perl6
19:42 timotimo oh hey
19:42 timotimo my optimization isn't the one that breaks trigonometry!
19:44 * jnthn waves from his rather nice Goteborg hotel
19:45 jnthn So design. Such Scandinavian. Wow.
19:45 lue Here's my potential replacement for S32::Str, could use some feedback :)  https://gist.github.com/lue/9941658
19:48 * FROGGS waves from his couch
19:49 * lue wonders what would happen if things like :21<1⑳> were allowed...
19:49 * lizmat waves from really
19:50 jnthn lizmat: OH ECHT?
19:51 lizmat yes, really!
19:51 FROGGS *groan*
19:52 jnthn retupmoca: I'm not too comfortable with the PR. I agree the problem needs a solution, but I think there's probably a better one. Give me a bit of thinking time. :)
19:53 FROGGS we could invent a special var like $- or $;
19:53 FROGGS we could have plenty of those :P
19:53 retupmoca jnthn: not a problem
19:54 retupmoca FROGGS: I can find the .so file fine, but the problem is that it's different at run time and compile time
19:54 lue FROGGS: it should be $# :)
19:54 FROGGS we could treat the given path specially if it starts with './'
19:54 jnthn $💩 :P
19:55 FROGGS hehe
19:55 retupmoca that was my other idea - if the path starts with './', look through @*INC to find it
19:55 jnthn m: my \� = 'crap'; say �
19:55 camelia rakudo-moar 358582: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/HO2404Pagcâ�¤Malformed myâ�¤at /tmp/HO2404Pagc:1â�¤------> [32mmy [33mâ��[31m\� = 'crap'; say �[0mâ�¤    expecting any of:â�¤        scoped declaratorâ�¤Â»
19:55 jnthn Aww
19:55 jnthn m: my \term:<�> = 'crap'; say �
19:55 camelia rakudo-moar 358582: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/jrQJZNeDUkâ�¤Term definition requires an initializerâ�¤at /tmp/jrQJZNeDUk:1â�¤------> [32mmy \term[33mâ��[31m:<�> = 'crap'; say �[0mâ�¤    expecting any of:â�¤        scoped declaratorâ�¤Â»
19:55 FROGGS or use the file location as a base
19:55 FROGGS I dunno
19:55 jnthn aww
19:55 lue m: my \💩 = 'crap'; say 💩
19:55 camelia rakudo-moar 358582: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/K41XmsyJncâ�¤Malformed myâ�¤at /tmp/K41XmsyJnc:1â�¤------> [32mmy [33mâ��[31m\💩 = 'crap'; say 💩[0mâ�¤    expecting any of:â�¤        scoped declaratorâ�¤Â»
19:56 FROGGS there should be a sane solution... somewhere
19:56 dmol joined #perl6
19:56 jnthn m: say '�' ~~ /<ident>/
19:56 camelia rakudo-moar 358582: OUTPUT«Nil␤»
19:56 lizmat FROGGS: the %?RESOURCE hash ?
19:56 jnthn Ah, that explains it a bit
19:56 FROGGS lizmat: perhaps, yes
19:57 retupmoca lizmat: will that be usable during development? Before I've done a panda install?
19:57 FROGGS retupmoca: I think so
19:57 retupmoca well then, somebody needs to hurry up and implement that :P
19:57 lizmat I would assume a "resource" directory in the -I directory should work in that case
19:58 jnthn Well, I think in general we want to have a way to provide a mechanism to resolve a library
19:58 FROGGS because we then do our best to gather the information that would be collection during installation
19:58 FROGGS so, if you had a META6.json we might consider it
19:58 lizmat I guess that makes sense, no?
19:59 FROGGS I think so, but my brane is somewhere else atm
19:59 lizmat you would only have a %?RESOURCE from an -I directory if there is a META6.json file in there that specifies it
19:59 lizmat I think that would make a lot of sense  :-)
19:59 retupmoca that's what I would expect
19:59 FROGGS yes
20:04 * lizmat displaces her brain to bed along with the rest of her
20:05 FROGGS gnight lizmat
20:05 jnthn o/ lizmat
20:05 retupmoca hmm, %?RESOURCE will probably have the same issue as currently though - at compile time the path will be different than run time
20:06 FROGGS retupmoca: %?RESOURCE will be made up when we load a module I think
20:06 FROGGS and at that time we know its "path"
20:06 FROGGS if it actually has a path, it can live in a database in theory
20:06 retupmoca FROGGS: right, but that will happen after the is native('...') path is compiled and set
20:07 FROGGS hmmm
20:07 FROGGS damn
20:07 FROGGS yeah
20:07 FROGGS then you want to install compiled libs?
20:07 retupmoca yes - I'm using Build.pm to compile a library and put it into the lib directory
20:07 FROGGS and compile at the target location or so?
20:08 retupmoca when you do a panda install
20:08 retupmoca but so far I haven't found a good way to find and load that library. I can find it at runtime, but NativeCall needs it at compile time right now
20:09 retupmoca so I think it needs some kind of solution in NativeCall.pm6 - and I will defer to jnthn on the best way to do that
20:11 arnsholtAtHome retupmoca: Maybe a bit of a hack, but have you tried just doing "trait_mod:<is>(&my-sub, :native($libname))" at runtime? I think that should work
20:12 jnthn I'd like a solution that lets us get rid of the need to put the library name into "is native" every time also
20:12 arnsholtAtHome There is that too, though
20:12 retupmoca arnsholtAtHome: I haven't tried that. It would probably work, but I would rather figure out a "proper" way to do it
20:13 arnsholtAtHome I've been pondering maybe something like "module Library::Internals is native<libname>", but I'm not sure how feasible that is
20:13 denis_boyun joined #perl6
20:13 jnthn Well, I was wondering about something like
20:13 arnsholtAtHome Yeah, a proper way would be great too. But if that works you can keep on hacking and plop in the right way once we know what the right way is =)
20:13 jnthn sub res() is native_lib_resolver { return 'foo'; }
20:14 jnthn And then sub blah() is native { * }
20:14 jnthn And it'll use the resolver.
20:14 jnthn But needs some care over scoping.
20:14 arnsholtAtHome Yeah, that sounds neat
20:15 arnsholtAtHome And maybe allow for sub res(&the-sub) { ... } as well?
20:15 jnthn And then you can write whatever code you want in there
20:15 arnsholtAtHome Yeah, I like that idea
20:16 retupmoca much cleaner than the current solution
20:16 retupmoca +1 as long as it's runtime :P
20:16 FROGGS p: use NativeCall; sub foo is native(-> { 'path/to/lib.so' }) { * }
20:16 camelia rakudo-parrot 358582: OUTPUT«===SORRY!===␤Could not find NativeCall in any of: /home/p6eval/rakudo-inst-2/lib/parrot/6.1.0-devel/languages/perl6/site/lib, /home/p6eval/rakudo-inst-2/lib/parrot/6.1.0-devel/languages/perl6/vendor/lib, /home/p6eval/rakudo-inst-2/lib/parrot/6.1.0-deve…»
20:17 arnsholtAtHome FROGGS: Won't work with current Zavolaj
20:17 FROGGS arnsholt: I know :o)
20:17 FROGGS I just wanted to see how it explodes
20:18 FROGGS if we wanna allow callables in that trait, it should perhaps get either the subname or the sub object as an argument
20:19 jnthn That would be another way
20:19 arnsholtAtHome That's not as useful though, since traits are applied at compile-time, not run-time
20:19 jnthn then you could
20:19 jnthn sub &imlib() { ... }
20:19 jnthn sub foo is native(&imlib) { * }
20:19 jnthn And it'd just stash a reference to the sub
20:20 jnthn and call it on first invocation
20:20 arnsholtAtHome Right. That'd work, of course
20:20 jnthn And means we don't have scope trickery to do
20:20 timotimo i'd still love to have something that means we don't have to write "is native(&thing)" over and over and over again
20:20 jnthn Yeah
20:21 FROGGS but how?
20:21 jnthn "is native" should be enough.
20:21 jnthn FROGGS: Well, maybe with something like I suggested.
20:21 FROGGS apply a trait over a collection?
20:21 FROGGS hmmm
20:41 effbiai joined #perl6
20:53 FOAD joined #perl6
20:59 mmachenry joined #perl6
21:02 zby_home_ joined #perl6
21:02 eiro joined #perl6
21:02 eiro hello everyone
21:03 timotimo hi there
21:03 dalek rakudo/nom: e872810 | (Tobias Leich)++ | src/Perl6/World.nqp:
21:03 dalek rakudo/nom: noop in case method resolved did not run yet
21:03 dalek rakudo/nom:
21:03 dalek rakudo/nom: This gets us a little further though this really hides a deeper problem.
21:03 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/e8728107ec
21:08 hoverboard joined #perl6
21:20 treehug88 joined #perl6
21:42 rurban1 joined #perl6
21:46 btyler joined #perl6
21:49 jdv79 joined #perl6
21:49 jdv79 is there a "pretty print" dump?
21:49 jdv79 i naively thought .gist would do that
21:50 FROGGS jnthn: jvm is unaffected by the patch... I'll fix that tomorrow: https://gist.github.com/FROGGS/0ea5537eb675588baaa2
21:50 jnthn k
21:51 FROGGS jdv79: .gist is just for human beings
21:51 FROGGS .perl is intended for reevaluation
21:51 jdv79 a hash strung out on one line is not "human friendly" at to this human
21:51 FROGGS true, this can be improved
21:52 jdv79 so to dump something large a module is needed at this point?
21:52 FROGGS I've not seen a pretty dumper... (except when you take JSON::Pretty into acount)
21:53 FROGGS ... which would be okay for hashes
21:53 jnthn FROGGS: Wow, just 2 modules with test fials
21:53 jnthn *fails
21:53 jdv79 ok
21:53 FROGGS jnthn: yeah, which is very surprising
21:53 * [Coke] would hope we'd end up with something like .perl(:pretty) or something.
21:53 [Coke] which would do the nested levels and indents.
21:54 jnthn [Coke]: That'd be kinda nice, yes.
21:54 jnthn FROGGS: Yeah. April having a Moar star is starting to look pretty good.
21:54 jdv79 that would be cool
21:55 [Coke] m: say <a b c >.perl()
21:55 camelia rakudo-moar e87281: OUTPUT«("a", "b", "c")␤»
21:55 [Coke] m: say <a b c >.perl :thing
21:55 camelia rakudo-moar e87281: OUTPUT«("a", "b", "c")␤»
21:55 [Coke] m: say <a b c >.perl :pretty
21:55 camelia rakudo-moar e87281: OUTPUT«("a", "b", "c")␤»
21:55 [Coke] m: say <a b c >.perl :iamtheverymodelofamodernmajorgeneral #huh
21:55 camelia rakudo-moar e87281: OUTPUT«("a", "b", "c")␤»
21:56 jnthn [Coke]: Methods always slurp up any extra names
22:01 [Coke] a $*perl-indent var would be safe to use for tyhis, yes?
22:02 jnthn 'night, #perl6
22:03 colomon joined #perl6
22:14 timotimo for what exactly?
22:15 timotimo oh
22:15 timotimo i see
22:29 rurban did you see fperrad's secret nqp already? https://github.com/fperrad/tvmjit/tree/master/t/nqp
22:33 [Coke] i don't see any info about what kind of nqp that is.
22:33 rurban it written in tvmjit
22:33 rurban it's written in tvmjit
22:34 rurban and used by https://github.com/richardhundt/shine (which is nice)
22:36 timotimo secret nap? mhhh
22:40 rurban This is the src for his nqp: https://github.com/fperrad/tvmjit/blob/master/src/nqp/compiler.lua
22:40 timotimo suddenly a bunch of lisp
22:40 [Coke] I wonder if it's like parrot/nqp, perl6/nqp, one of the other nqps, or a new nqp. (in terms of what he's trying to support)
22:40 rurban yes, tvmjit is featuring a proper s-expr language :)
22:41 rurban I guess it's his parrot project
22:41 [Coke] if perl6/nqp, I wonder how much effort it would be to add it as a backend.
23:07 rurban1 joined #perl6
23:08 rurban2 joined #perl6
23:28 raiph joined #perl6
23:33 raiph lue: comments added to your s32::Str gist
23:34 raiph almost all nitpicky; main thing is confusion about your use of the word "codepoints"
23:35 lue raiph: codepoint is specifically a unicode codepoint; there should be no confusion with that term.
23:35 raiph that's what i thought, but i don't think your spec achieves that
23:36 lue well, it's not in isolation. S15 is the actual Unicode spec :) .
23:37 lue (besides, since NFG-only graphemes don't have codepoints, they don't count as one. I honestly don't see where confusion can arise.)
23:39 raiph Maybe you'll see what I mean when you've read the gist comment. If not, ping me again.
23:40 lue I've read through the whole thing; I still don't see how one could confuse "codepoints" to refer to things that don't have codepoint numbers (that is, NFG graphemes).
23:40 raiph k, let me try explain what's bugging me
23:40 raiph "C< Unicodey> operates on the codepoint level as its view of what a character is. It doesn't enforce any particular Normalization Form on its contents."
23:41 raiph first, are you counting NFG as a Normalization Form?
23:41 lue on the enforcement statement, yes.
23:41 lue (NFG isn't enforced any more than NFC or NFKD)
23:42 raiph so one could have something that does Unicodey that's in NFG?
23:43 raiph and then do .chars on it
23:43 raiph which will give a count of the graphemes
23:43 raiph but your statement says Unicodey operates on the codepoint level as its view of what a character is
23:43 lue Ah, I think I see now. NFG is meant to be a specialization of Unicodey, so only Str would deal with NFG stuff.
23:44 lue s/Unicodey/Unicode/
23:45 raiph or, tc: "Returns a string with each codepoint lowercased."
23:45 lue Unicodey and Uni are basically the same thing, fwiw.
23:45 lue tc shouldn't say that.
23:45 raiph but what if one is working at the NFG level?
23:45 raiph I think it ought to say chars where you've got codepoints
23:45 raiph (for tc and cousins)
23:46 stevan_ joined #perl6
23:46 lue Well no, for Unicodey "codepoints" is the correct term. I didn't think I'd have to restate every method on C<Str> to specify that it operates on graphemes in that class :(
23:47 [Coke] if it's universal, hoist it out and put it at the top.
23:47 [Coke] said the guy who hasn't read the doc or the gist.
23:47 stevan_ joined #perl6
23:48 timotimo gnite #perl6
23:48 raiph \o timotimo
23:49 btyler joined #perl6
23:49 raiph lue: aiui tc and cousins work on "characters", not codepoints; if a string is NFG that'll be graphemes, if NFC it'll be codepoints, and so on.
23:50 lue raiph: yes, but the Unicodey role operates on codepoints. Only Str handles NFG graphemes. Those method definitions are attached the Unicodey role. Thus they are described as operating on "codepoints".
23:51 raiph gotchya, so Unicodey explicitly does not include NFG, right?
23:51 lue (note how Stringy's definition of .tc doesn't mention "title case", but rather "upper case", because ASCII has no notion of "title case". Stringy's method definitions don't try to generalize for all things that may do it; neither does Unicodey.)
23:52 lue raiph: yes, Unicodey is meant to be a Unicode-based role, so adding a non-Unicode extension to it is less-than-ideal.
23:54 raiph lue: gotchya
23:54 lue IIRC S15 explains all this already, and I tried to not re-explain things here (esp. stuff that doesn't really fit S32's purpose of laying out the API).
23:55 lue printf, pack, and unpack need expansion, I'll get to those later :) .
23:58 lue btw, I don't index and rindex have any RTL concerns; RTL text *looks* like it's going in the reverse direction, but the start of text is still at the left side in storage.
23:59 raiph lue: k

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

Perl 6 | Reference Documentation | Rakudo