Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2014-10-20

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:07 colomon joined #perl6
00:08 aborazmeh joined #perl6
00:10 TimToady well, that explains the throbbing headache, 100.5℉ == 38.1℃
00:15 geekosaur :(
00:15 colomon :(
00:24 timotimo oh damn. not as mild as we had hoped
00:37 Hor|zon joined #perl6
00:43 PZt joined #perl6
00:46 _sri_ joined #perl6
01:07 pdcawley joined #perl6
01:11 timotimo google-authenticator + FreeOTP ♥
01:12 timotimo next project: libqrencode binding :3
01:13 gfldex joined #perl6
01:13 Mouq joined #perl6
01:14 Mso150_c_r joined #perl6
01:18 Mso150_c_r joined #perl6
01:22 Mso150_c_r joined #perl6
01:40 Mso150_c_r joined #perl6
01:42 gtodd joined #perl6
02:24 gtodd1 joined #perl6
02:25 virtualsue joined #perl6
02:31 chenryn joined #perl6
02:32 erkan joined #perl6
02:32 erkan joined #perl6
02:38 Hor|zon joined #perl6
02:48 chenryn joined #perl6
02:50 noganex_ joined #perl6
02:57 pyrimidine joined #perl6
02:59 pyrimidine joined #perl6
03:05 raydiak https://gist.github.com/raydiak/6fb0016cf397bf1420f5 yay progress
03:05 pyrimidine just curious, but is there a specific way to test a IO::Handle mode(rw, a, p) after it has been instantiated?
03:06 pyrimidine I've tried checking $fh.r and $fh.w, they always return True
03:11 _sri_ joined #perl6
03:13 mauke_ joined #perl6
03:17 gfldex joined #perl6
03:45 firefish5000 joined #perl6
03:58 _sri_ joined #perl6
04:02 kaare joined #perl6
04:32 avuserow joined #perl6
04:39 Hor|zon joined #perl6
04:44 _sri_ joined #perl6
04:46 ggoebel11111115 joined #perl6
04:46 _sri_ joined #perl6
04:48 kaare_ joined #perl6
04:59 vukcrni joined #perl6
05:02 avuserow joined #perl6
05:05 [Sno] joined #perl6
05:24 avuserow joined #perl6
05:37 ecocode joined #perl6
05:41 slavik joined #perl6
05:47 anaeem1_ joined #perl6
05:50 anaeem1_ joined #perl6
06:02 chenryn joined #perl6
06:07 [Tux] joined #perl6
06:24 FROGGS joined #perl6
06:24 anaeem___ joined #perl6
06:27 immortal joined #perl6
06:27 immortal joined #perl6
06:34 moritz \o
06:35 raydiak o/
06:40 Hor|zon joined #perl6
06:41 anaeem1_ joined #perl6
06:48 rindolf joined #perl6
06:50 Mso150 joined #perl6
06:53 kaleem joined #perl6
06:59 erkan joined #perl6
07:00 zakharyas joined #perl6
07:04 firefish5000 joined #perl6
07:08 kurahaupo_ joined #perl6
07:14 jerrycheung joined #perl6
07:16 Mso150_c joined #perl6
07:17 darutoko joined #perl6
07:17 virtualsue joined #perl6
07:19 Mso150 joined #perl6
07:20 salv0 joined #perl6
07:44 sergot mornign o/
07:51 chenryn joined #perl6
08:00 Hor|zon joined #perl6
08:02 abraxxa joined #perl6
08:13 chenryn joined #perl6
08:22 rurban joined #perl6
08:28 fhelmberger joined #perl6
08:30 rindolf sergot: morning.
08:31 sergot rindolf: o/
08:31 rindolf sergot: sup?
08:32 kjs_ joined #perl6
08:32 dakkar joined #perl6
08:42 TuxCM joined #perl6
08:48 sergot rindolf: sleepy but fine
08:48 rindolf sergot: ah.
08:49 sergot rindolf: how are you?
08:50 rindolf sergot: I'm fine - working on Website Meta Language.
08:51 sergot sounds interesting  :)
09:00 dalek rakudo/nom: 7ff58f0 | (Elizabeth Mattijsen)++ | src/core/ (4 files):
09:00 dalek rakudo/nom: Streamline Bag.pick/roll/grab/pickpairs/grabpairs
09:00 dalek rakudo/nom:
09:00 dalek rakudo/nom: Using MMD: should be significantly faster on no param calls, and on Whatever/Inf
09:00 dalek rakudo/nom: calls.  Also fix #120407
09:00 synopsebot Link: https://rt.perl.org/rt3//Public/Bug/Display.html?id=120407
09:00 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/7ff58f00b0
09:16 pecastro joined #perl6
09:17 dalek roast: 01feaf8 | (Elizabeth Mattijsen)++ | S02-types/bag (2 files):
09:17 dalek roast: Some test adaptations because of Baggy streamline
09:17 dalek roast: review: https://github.com/perl6/roast/commit/01feaf8158
09:18 rmgk_ joined #perl6
09:25 lizmat seems I broke jvm  :-(
09:25 lizmat $ perl6-j -e '<a b b>.BagHash.roll(2).say'
09:25 lizmat Exception in thread "main" org.perl6.nqp.runtime.UnwindException
09:29 abraxxa when NativeCall handles Str, does it also handle CArray[Str]?
09:31 salv0 joined #perl6
09:32 denis_boyun_ joined #perl6
09:34 leont joined #perl6
09:49 timotimo https://github.com/jnthn/zavolaj/blob/master/t/05-arrays.t#L25 ← abraxxa
09:55 woshty joined #perl6
09:57 * lizmat gives up on trying to figure out what broke Bag.roll/pick/grab and goes away for the day
09:57 lizmat if anybody interested in debugging: it's probably something in ROLLPICKGRABN
09:57 lizmat - it's not the native int
09:58 lizmat - it's not having rand/seen defined outside of the gather/while
09:58 lizmat away&
09:59 kjs_ joined #perl6
09:59 abraxxa timotimo: thanks
09:59 abraxxa it seems i have to pass a CArray[int8] because it might a Str but also something else like an Int
10:00 abraxxa how do i get the length of a CArray[int8]?
10:00 timotimo usually .elems
10:03 b2gills joined #perl6
10:07 abraxxa does that look right? http://paste.scsys.co.uk/433328
10:10 timotimo doesn't look wrong to me
10:10 timotimo m: my $vbuf = "hi there".encode("UTF-16").elems
10:10 timotimo still timing out?
10:10 timotimo >_>
10:11 camelia rakudo-moar dcd32e: OUTPUT«(timeout)»
10:12 robinsmidsrod joined #perl6
10:27 chenryn joined #perl6
10:47 cognominal joined #perl6
10:48 ptc_p6 joined #perl6
10:56 abraxxa timotimo: it seems the @valuep only gets it's first element assigned
11:02 chenryn joined #perl6
11:03 timotimo what does $valuebuf.elems give you?
11:06 azawawi joined #perl6
11:06 azawawi hi
11:06 azawawi m: homedir('.')
11:06 virtualsue joined #perl6
11:06 azawawi std: homedir('.')
11:06 moritz .oO( my homedir is my castledir )
11:06 camelia rakudo-moar dcd32e: OUTPUT«(timeout)»
11:06 camelia std 14ad63b: OUTPUT«[31m===[0mSORRY![31m===[0m�Undeclared routine:� 'homedir' used at line 1�Check failed�FAILED 00:01 140m�»
11:07 azawawi im getting homedir... $*HOME is not set in my perl6 repl
11:07 azawawi Dynamic variable $*HOME not found
11:08 azawawi latest rakudo
11:08 moritz lizmat: ^^
11:09 azawawi :)
11:09 ivanshmakov joined #perl6
11:09 azawawi + Implemented tmpdir($dir) and homedir($dir)  # in docs/ChangeLog :)
11:10 azawawi moritz: doc.perl6.org needs a lot of changes in the variables section
11:10 azawawi moritz: a lot of deprecated variables like $*OS
11:10 azawawi moritz: my fault... i was reading an old copy :)
11:12 rurban Made myself finally a working 64-bit big-endian machine, a qemu mips64
11:12 azawawi moritz: a better REPL https://github.com/azawawi/farabi6/blob/master/test_repl.pl :)
11:13 FROGGS rurban++ # I've seen that on #parrot already, good work!
11:13 azawawi moritz: im thinking of generalizing it to send/recv json requests
11:14 azawawi latest rakudo is also faster a bit in json parsing
11:14 azawawi that before that is
11:14 azawawi than before that is  # correction... :)
11:15 rurban but max memory is 2GB, I only got 256M RAM working yet, so no chance for a rakudo. Need to wait for a PPC64 qemu
11:16 rurban Sparc64 qemu will probably dead
11:17 abraxxa timotimo: 1
11:17 timotimo well, uh ... there you have it?
11:17 azawawi timotimo: hi there
11:17 timotimo azawawi: last time i tried, i still got that very same error from farabi6 ;(
11:17 FROGGS rurban: there is a chance that I get my PPC64 machines working at the end of the week
11:19 abraxxa timotimo: perl6 -e 'use NativeCall;  my @valuep := CArray[int8].new; @valuep[0] = 83; @valuep[1] = 27; @valuep[2] = 34; say @valuep.elems;'
11:19 rurban Good. I found one GC assertion, but I suppose it's the memory pressure
11:19 timotimo oh
11:19 abraxxa is that a bug or can't i add elements to a CArray that way?
11:20 timotimo maybe try using $ and regular assignment instead of the @ sigil
11:20 abraxxa Variable '$valuep' is not declared. Did you mean '@valuep'?
11:20 FROGGS abraxxa: what gives you: perl6 -e 'use NativeCall;  my @valuep := CArray[int8].new; @valuep[0] = 83; @valuep[1] = 27; @valuep[2] = 34; say @valuep[0].elems;'
11:21 timotimo abraxxa: use the $ everywhere
11:21 abraxxa FROGGS: 1
11:21 abraxxa timotimo: still 1
11:21 FROGGS hmmm
11:21 abraxxa perl6 -e 'use NativeCall;  my $valuep := CArray[int8].new; $valuep[0] = 83; $valuep[1] = 27; $valuep[2] = 34; say $valuep.elems;'
11:22 FROGGS does .elems actually work here?
11:22 timotimo let me have a look
11:22 timotimo it's not implemented for CArray >_<
11:22 FROGGS :o)
11:22 timotimo but CArray derives from something
11:22 abraxxa yes, but i couldn't find out if CArray subclasses from something core
11:23 timotimo it derives from Any
11:23 FROGGS star-m: use NativeCall; say CArray.^mro
11:23 timotimo Any has an "elems" method that would return 1 in this case
11:23 camelia star-m 2014.08: OUTPUT«(timeout)»
11:23 timotimo i've been bitten by CArray not implementing methods in the past >_>
11:23 abraxxa i need something like https://github.com/jnthn/zavolaj/blob/master/t/05-arrays.t#L87
11:24 abraxxa so how do i get the number of elements then?
11:26 timotimo um, wait a minute
11:26 timotimo why are you asking the CArray for its number of elements in the first place?
11:26 timotimo just to check whether or not the assignment has worked?
11:27 abraxxa yes, to not rely on a different object and it's interface
11:28 abraxxa why does CArray not subclass from e.g. List?
11:29 timotimo i don't think it can; different REPRS
11:29 timotimo REPRs
11:29 timotimo it does implement the Positional role
11:30 abraxxa i see
11:30 timotimo as a workaround you can nqp::elems the CArray, i believe
11:32 timotimo gotta run now
11:32 abraxxa ok, thanks
11:32 abraxxa i'm using the elems method on the Buf object i'm taking the bytes from
11:33 abraxxa although someone said that Buf.elems doesn't return the number of bytes but elements which could be multibyte as well
11:33 abraxxa I need a way to loop a Buf per byte
11:33 brrt joined #perl6
11:34 abraxxa I can access each byte using $buf.[$_] but i need an index counter that goes from 0 to number of bytes
11:35 brrt \o
11:42 azawawi timotimo: do you have a $HOME/lib folder?
11:42 azawawi timotimo: or a $HOME/.perl6 ?
11:42 abraxxa when I define a NativeCall method parameter as 'Str $placeholder is encoded('utf16'),' is the bytestring null terminated?
11:43 moritz uhm
11:43 moritz null-terminated UTF-16 strings don't make much sense, since they have 0bytes quite naturally
11:43 abraxxa http://docs.oracle.com/cd/E11882_01/appdev.112/e10646/oci16rel003.htm#LNOCI17140
11:43 abraxxa placeholder
11:44 abraxxa That is, if the statement is prepared in UTF-16, so is the placeholder. As a string type parameter, the placeholder should be cast as (text *) and terminated with NULL.
11:45 timotimo azawawi, my prefix is perl6/install
11:46 ghostlines joined #perl6
11:47 ghostlines joined #perl6
11:50 abraxxa moritz: ^^
11:52 moritz abraxxa: http://en.wikipedia.org/wiki/Null-terminated_string#Character_encodings
11:53 abraxxa moritz: does that help? http://docs.oracle.com/cd/E11882_01/appdev.112/e10646/oci05bnd.htm#LNOCI16359
11:53 abraxxa strlen(ename)+1
11:53 timotimo maybe it is terminated with two null bytes?
11:56 azawawi timotimo: could you please share the rakudo build command that you are using so i can try to duplicate the issue on my machine?
11:57 timotimo i build all three components individually
11:57 timotimo in perl6/ i have the mvm, nqp and rakudo source folders
11:57 timotimo as well as ecosystem/ with panda in it
11:58 cognominal joined #perl6
11:58 timotimo I cannot laptop now and will be afk for one and a half hours soon
11:59 azawawi timotimo: could you please file an issue so i can keep track of the problem?
11:59 timotimo ok
11:59 azawawi timotimo++
12:05 bartolin joined #perl6
12:08 * [Coke] stares at #perl6 backlog over coffee.
12:12 * [Coke] wonders what "new-fap" means.
12:12 [Coke] oh.
12:15 * moritz hopes its gone now
12:21 chenryn joined #perl6
12:32 xinming_ joined #perl6
12:42 abraxxa how can i get the number of bytes in a Buf or loop over it?
12:44 azawawi abraxxa: $buf.elems ?
12:45 azawawi abraxxa: http://doc.perl6.org/type/Buf#method_elems
12:45 abraxxa that gives the same number as Str.chars where it came from
12:46 abraxxa i need a buf8
12:46 * moritz has no idea if that's implemented
12:49 guru joined #perl6
12:51 * azawawi begins working on farabi6 help search sidebar
12:53 azawawi moritz: wouldnt it be more user friendly  to run p6doc-index when the user asks for p6doc -f split for the first time?
12:54 abraxxa moritz: what is required to make it more usable? extending NativeCall or Buf?
12:58 moritz abraxxa: Buf, I think
12:58 moritz azawawi: if the user has write permissions to the appropriate directories, yes
12:59 abraxxa what is $nf? http://doc.perl6.org/type/Str#method_encode
12:59 moritz normal form
12:59 moritz not yet implemented
12:59 abraxxa can i force a utf-16 encoding returning a buf8?
13:01 moritz $ ./perl6-m -e 'say "abc".encode("UTF-16").decode("Latin-1").encode'
13:01 moritz Can not decode a utf-16 buffer as if it were Latin-1 in method decode at src/gen/m-CORE.setting:5791
13:01 moritz tough beans
13:01 moritz that would have been useful
13:01 kaleem joined #perl6
13:11 azawawi is there an internal count of gather-ed items in a gather { ... } ?
13:11 mirjam joined #perl6
13:15 [Coke] "hey, why is my laptop making that... oh, the jvm build."
13:16 moritz azawawi: I don't think so
13:19 brrt [Coke]: i have that when i run spectests :-)
13:20 [Coke] brrt: it's fine on everything up to the jvm. (moar, moar-jit, parrot...)
13:41 JimmyZ joined #perl6
13:42 gfldex joined #perl6
13:44 abraxxa moritz: can i open an issue or RFC somewhere for that feature?
13:46 moritz abraxxa: for what? Buf16 -> Buf8 conversion?
13:46 abraxxa moritz: yes, byte interface to Buf objects generally
13:46 moritz abraxxa: yes; mailto:rakudobug@perl.org
13:46 brrt iirc that's a bit of an 'open issue' :-)
13:46 abraxxa moritz: will do
13:46 abraxxa where does that mail end up?
13:46 moritz abraxxa: rt.perl.org in the perl6 queue
13:47 abraxxa then i prefer the web interface
13:54 abraxxa moritz: https://rt.perl.org/Ticket/Display.html?id=123015
13:55 kjs_ joined #perl6
14:02 kaare_ joined #perl6
14:05 dayangkun joined #perl6
14:07 dalek perl6-roast-data: 135314e | coke++ | bin/rakudo. (4 files):
14:07 dalek perl6-roast-data: Use more cores
14:07 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/135314ee4e
14:07 dalek perl6-roast-data: a42b9c3 | coke++ | bin/doit:
14:07 dalek perl6-roast-data: fix typo
14:07 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/a42b9c320e
14:07 dalek perl6-roast-data: e6db84b | coke++ | / (5 files):
14:07 dalek perl6-roast-data: today (automated commit)
14:07 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/e6db84b8b1
14:08 [Coke] ^^ that might be the first ad hoc run with all "4" backends.
14:09 [Coke] note that there are 625 parrot failures, and 627 jvm failures.
14:10 dalek perl6-roast-data: 74b00f7 | coke++ | / (5 files):
14:10 dalek perl6-roast-data: today (automated commit)
14:10 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/74b00f7211
14:16 [Coke] many of those parrot failures are probably from the lack of icu; We should probably either die if we have no icu, or not run all these failing spectests.
14:18 moritz I'd be +1 to making ICU a hard requirement for rakudo-on-parrot; jnthn, any objections?
14:22 Mouq joined #perl6
14:27 [Coke] redoing today's parrot run with icu…
14:27 jnthn moritz: Don't immediately see why not; folks who don't want to find an ICU have two other backends to choose from... :)
14:27 yoleaux 19 Oct 2014 20:30Z <lizmat> jnthn: I think there is some codegen issue with private methods, contact me for an example
14:27 yoleaux 19 Oct 2014 20:31Z <lizmat> jnthn: I would not be surprised if this would be the source of a lot of flappines, specifically in async modules
14:29 [Coke] +1 here.
14:32 timotimo lizmat: could you pin down the performance issues you encountered with multiple dispatch? was it all due to having a given number in one/more of the candidates?
14:33 brother joined #perl6
14:33 jnthn .tell lizmat It's better to do async communication with me these days than try and find a time we're both on IRC, I suspect. :) Just drop the code example that shows it or other stuff I need to reproduce it in a gist and .tell me it ;)
14:33 yoleaux jnthn: I'll pass your message to lizmat.
14:34 jnthn timotimo: Well, if you have candidates that have non-nominal stuff in them, then you end up having to run the dispatcher and maybe trial bind signatures.
14:34 jnthn timotimo: Whereas purely nominal cases hit the cache.
14:35 timotimo oh, oops :)
14:35 moritz jnthn: well, one possible objection would have been: you have trouble installing libicu, and can't fix parrot backend code anymore if we do that :-)
14:36 jnthn moritz: Given I didn't fix Parrot backend code in a LONG time... :)
14:36 jnthn FROGGS++ is mostly to thank for keeping Parrot up to date with stuff.
14:37 moritz so let's here if FROGGS++ objects
14:37 jnthn timotimo: So while multi fac(0) { 1 } is all very cute, it's (for the time being, and probably for a good while) going to be a decent amount cheaper to do such checks inside an only.
14:37 jnthn moritz: Yes, he's a better person to ask than me. :)
14:38 guru joined #perl6
14:38 FROGGS err, about what?
14:39 moritz FROGGS: any objections to making libicu mandatory for parrot?
14:39 FROGGS ohh, so detect icu in rakudo's Configure and deny to build it if absent?
14:39 timotimo jnthn: got it
14:39 moritz FROGGS: correct
14:39 FROGGS +1
14:48 ajr joined #perl6
14:50 kaleem joined #perl6
14:50 arnsholt o/
14:51 * arnsholt wrote Perl 6 (NQP, actually) code for the first time in forever last night
14:51 arnsholt That felt nice
14:51 FROGGS :o)
14:51 arnsholt Must be almost a year since last I did that
14:51 dalek snake: 2f600b0 | (Arne Skjærholt)++ | / (7 files):
14:51 dalek snake: Move to Moar from Parrot; unbitrot; much improved indent handling.
14:51 dalek snake:
14:51 dalek snake: First is self-explanatory, second simply due to the code being untended for a
14:51 dalek snake: long time and easily fixed by cargo-culting a bit of code from Rakudo.
14:51 dalek snake:
14:51 dalek snake: Much improved indent handling has been achieved by following the Pynie
14:51 dalek snake: approach much more closely. As a consequence, we now handle ifs with elses
14:51 dalek snake: properly (attaching a multiply dedented else to the correct parent rather than
14:51 dalek snake: the first available). Also adds a simple sanity test for if handling.
14:51 dalek snake: review: https://github.com/arnsholt/snake/commit/2f600b00d5
14:52 arnsholt The fact that it was for my Python compiler may or may not be a good thing ^_^
14:52 FROGGS hehe, true
14:53 arnsholt But I was happy that I figured out how to handle the indents correctly
14:53 arnsholt Turned out to be reasonably straightforward, in the end
14:53 nine_ arnsholt: you have a Python compiler in Perl 6?
14:54 arnsholt I have a Python compiler that knows how to call NQP ops and ifs with elses (but no elif)
14:54 arnsholt Embryonic, at best
14:55 arnsholt But now that I've figured out the indent stuff, it should be possible to implement a bit more interesting stuff
14:55 nine_ arnsholt: the parser may be interesting to sugar Inline::Python
15:05 araujo joined #perl6
15:09 [Coke] whoops, misread: parrot had 2807 failures.
15:09 dalek perl6-roast-data: f35ad4c | coke++ | / (5 files):
15:09 dalek perl6-roast-data: today (automated commit)
15:09 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/f35ad4c766
15:10 [Coke] I force pushed the "real" parrot run for today.
15:11 rurban joined #perl6
15:13 Ven joined #perl6
15:13 Ven o/, #perl6
15:13 Ven has anybody here used roboVM :D? (basically, jvm on ios)
15:14 timotimo no
15:17 masak greetings, #perl6lings
15:18 Ven oyez, masak
15:18 Ven I'm actually wondering if I could do my next ios app in perl6.
15:18 Ven err, not "next". an ios app.
15:18 timotimo no wai
15:19 timotimo interpreted languages are forbidden on the ios market
15:19 Ven I don't really care about the ios market though
15:20 timotimo so ... how are you going to run that app on any ios device?
15:21 Ven specific provisions
15:21 Ven if you have company-specific apps, you probably don't want them on the market
15:21 molaf joined #perl6
15:22 timotimo apple allows that?
15:23 Ven I don't have any other explanation
15:23 pecastro joined #perl6
15:23 timotimo mhm mhm
15:24 Ven timotimo: and it seems wrong.
15:24 Ven I see a few RubyMotion apps on the ios market...
15:25 anaeem1 joined #perl6
15:25 [Coke] down to 9 failures on parrot. 2 of them are long standing OS X failures, I think.
15:27 Akagi201 joined #perl6
15:27 Ven timotimo: Right. It's easy: rubymotion and robovm are ahead-of-time compilers
15:30 * brrt afk
15:30 brrt left #perl6
15:31 [Coke] is S02-types/deprecations.t failing on parrot on a non-OS X box?
15:32 timotimo hmm
15:36 chenryn joined #perl6
15:36 [Coke] p: use Test; my $latin-chars = [~] chr(0)..chr(0xFF); is $latin-chars.comb(/<blank>/)>>.ord.join(","), '9,32,160', 'blank chars';
15:36 camelia rakudo-parrot 7ff58f: OUTPUT«(timeout)»
15:38 [Coke] ^^ that test depends on a certain version of unicode, maybe? locally I am getting 9,32, not 9,32,160
15:38 [Coke] (er, 9,32)
15:40 timotimo .tell azawawi i can run farabi6 when i clone it manually and run bin/farabi6
15:40 yoleaux timotimo: I'll pass your message to azawawi.
15:40 treehug88 joined #perl6
15:40 timotimo moritz: can we get a sort of API that gives the anchor for a given line in the irclog?
15:41 timotimo something simple like "get me the anchor for the last line from $nickname" would probably be enough to make yoleaux give good links to the clogs
15:41 timotimo when .tell-ing someone something
15:41 bartolin [Coke]: I get '9,32' as well locally (Debian 7, unicode-version is 6.0)
15:43 treehug88 joined #perl6
15:44 bartolin [Coke]: btw, if I read the irclog for #parrot correctly the two OX X failures in S19-command-line/dash-e.t should be gone with the next Parrot version
15:45 rurban Really?
15:48 Mouq joined #perl6
15:51 bartolin rurban: didn't you add platform encodings for darwin?
15:53 timotimo perlpilot haz a pull request :3
15:53 rurban yes, ok
15:54 rurban I'll check latest rakudo later. just testing weird platforms
15:54 rurban now ppc64
15:54 bartolin rurban++
15:55 timotimo PerlJam: that's your github user, right?
15:59 timotimo .tell azawawi i'm not really doing anything with farabi6, but it maxes out a single cpu on my machine :\
15:59 yoleaux timotimo: I'll pass your message to azawawi.
16:01 PerlJam timotimo: aye, perlpilot is me.
16:03 kjs_ joined #perl6
16:12 odc joined #perl6
16:16 Mouq joined #perl6
16:26 kurahaupo_ joined #perl6
16:29 Mso150 joined #perl6
16:38 raydiak \o mornin #perl6
16:38 timotimo hey ray
16:39 raydiak how's it going timo?
16:40 [Sno] joined #perl6
16:41 timotimo it's all right
16:41 timotimo kind of feeling unmotivated
16:41 timotimo but i should go ahead and do the weekly
16:41 timotimo (for two weeks ... again)
16:41 kjs_ timotimo: looking forward to reading your weekly!
16:42 timotimo thank you :)
16:46 raydiak yeah the weekly is awesome, keeping up with the backlog is *hard*
16:46 raydiak sorry you're in a lull
16:46 raydiak somewhat paradoxically, one of the main things that helps me when I'm feeling low on energy is to exercise a bit
16:46 timotimo
16:46 timotimo yes, i've heard about that
16:47 timotimo i have candy now, and then a pizza :3
16:47 timotimo also i'm sitting in front of two big monitors with a beatufil photograph of a cat on them
16:47 raydiak heh a lot to be happy about, then
16:49 timotimo :)
16:49 timotimo http://www.socwall.com/desktop-wallpaper/31219/cat/
16:50 denis_boyun_ joined #perl6
16:50 raydiak wow that is some awesome photography work
16:51 timotimo and it's available for free on the internets!
16:53 raydiak mine is a pic I took of a tree in a friend's yard, looking up from under the branches...I had a few drinks one night and did everything I needed to make it my boot splash, login splash, login theme background, wallpaper, shutdown splash...
16:53 timotimo :D
16:54 timotimo so ... what could non-expert perl6ers do to bring the GLR forwards?
16:54 timotimo write a bunch of test cases?
16:54 grondilu talking about cats, there was an awesome documentary about them lately on BBC:  https://www.youtube.com/watch?v=bRgasPQnjNg
16:57 guru joined #perl6
17:03 dalek roast: c03b6f9 | usev6++ | S02-types/lazy-lists.t:
17:03 dalek roast: Fudge failing test for RT #121994 for JVM
17:03 dalek roast: review: https://github.com/perl6/roast/commit/c03b6f9284
17:03 synopsebot Link: https://rt.perl.org/rt3//Public/Bug/Display.html?id=121994
17:03 raydiak the algebra system I started can isolate a single variable in simple equations now
17:04 kaleem joined #perl6
17:04 grondilu raydiak: where is this?  I'm interested in something like that.
17:05 raydiak grondilu: it's just something I started hacking on a few days ago, I don't have it anywhere public yet
17:05 grondilu ok
17:06 grondilu ever since I wrote my geometric algebra module, I've been thinking of implementing an algebraic role.
17:06 dwarring joined #perl6
17:06 grondilu but it's kind of tough, conceptually.
17:06 raydiak grondilu: if you're interested in contributing in the near future, I can clean it up and put it on github soonlyish
17:07 vendethiel joined #perl6
17:07 grondilu before coding, I think a lot of discussion is needed on the subject.  Even the choice of vocabulary is not obvious, I think.
17:07 vendethiel I just renamed myself on github to vendethiel. FROGGS++ for the motivation
17:08 raydiak true enough...what I have currently is...a bit sloppy :)
17:08 grondilu but I really think this is something perl6 should have.  Basically imho right now Numeric kind of sucks.
17:08 grondilu (I mean the Numeric role)
17:09 grondilu though maybe it will be on perl7
17:09 raydiak how so?
17:10 [Coke] "perl7" is a loaded term.
17:10 PerlJam There will be no perl7.  Ever.  (That's my prediction)
17:11 grondilu never say never
17:11 awwaiid [digression detected]
17:11 grondilu I think there will.
17:11 PerlJam grondilu: I didn't ;)
17:11 [Coke] I just did a 'make install' on a rakudo install. then did "panda rebootstrap" which complained about:
17:11 [Coke] ===SORRY!===
17:11 grep0r joined #perl6
17:11 [Coke] Missing or wrong version of dependency 'src/gen/m-BOOTSTRAP.nqp'
17:11 grettis joined #perl6
17:12 [Coke] (any invocation of "panda" gives that error)
17:13 PerlJam grondilu: you could make a perl7 just to invalidate my prediction, but barring that, I don't think there will ever be a perl7.
17:13 timotimo [Coke]: "panda rebootstrap"?
17:14 timotimo you want to do perl6 ./rebootstrap.pl
17:14 grondilu it's pretty clear that there is some sort of schism in the perl community.  It's quite likely that at some point the part of the perl5 team which doesn't like perl 6 decide to create perl 7.  And that'd be totally OK imho.
17:15 moritz as long as they solve both perl 5's and perl 6's problems, I'm all for it :-)
17:15 timotimo right after we "release" perl6, people will get confused again and believe perl6 is dying because "now they're working on perl7!"
17:16 grondilu I don't know.  Should major version numbers indicate some sort of advancement or can't tehy just indicate chronological order or something?
17:16 PerlJam grondilu: I don't doubt that the p5p will try to find a "way around" Perl 6 at some point.   I highly doubt it will be called "perl 7"
17:16 timotimo moritz: perl5's problem is that "no strict" is the default, perl6's problem is that it was too ambitious
17:16 timotimo the obvious solution is this:
17:16 timotimo perl7 is simply perl5 with "use strict" on by default
17:16 moritz timotimo: that isn't its only problem, IMHO
17:17 [Coke] timotimo: why do I need to go back to the panda install directory?
17:17 moritz timotimo: one highly relevant is the strings vs. byte confusion
17:17 grondilu perl7 would be a bit more than that
17:17 _slade_ joined #perl6
17:17 moritz timotimo: another is the lack of an object system that's integrated with core
17:17 grondilu a bit of formal calculus for instance would be a nice addition.
17:17 timotimo [Coke]: hmm. in theory panda could include a guard around its use statements at the very top to figure out if the dependencies have moved below its feet and find and launch the rebootstrap script
17:17 timotimo moritz: oh, i thought you meant perl6
17:18 timotimo mhhh, this pizza was garlicious
17:18 moritz timotimo: no, I'm still dwelling on p5 :-)
17:18 timotimo haven't lost hope just yet? :)
17:19 moritz timotimo: I'm still here am I? :-)
17:19 timotimo hope in perl5, i meant
17:19 ptc_p6 joined #perl6
17:19 timotimo communication is hard %)
17:20 moritz timotimo: I'm not sure, really. I mean, I use it for dajob, and it works fine. But I'm not seing evolutionary approaches working to some of p5's problems
17:20 moritz timotimo: and those that state that p5 and p6 are really separate languages (and I've done that too in the past) still don't have any solutions to p5's long term future
17:21 timotimo "do you really expect me to evolve into a better language?" - "no, mister perl5. i expect you to die"
17:23 moritz well, I do believe it evolves, and for the better. I just don't know yet if it will evolve far enough
17:23 [Coke] also: panda rebootstrap should probably work without a network connection.
17:23 [Coke] ... it failed, now it thinks there's no need to -rebootstrap- then bootstrap fails. *sigh*
17:23 PerlJam moritz: anything is possible.
17:24 PerlJam moritz: Maybe with P6 out there pushing the envelope on some things, p5p will wake up and steal some good ideas from it.
17:24 grondilu you just can't expect p5 to die until p6 proves it can rival it in terms of performance.
17:24 timotimo [Coke]: ah, yeah, rebootstrap saying "regular bootstrap is enough" is steered by finding a non-empty "state file", as in: stuff that's already installed
17:25 grondilu (I mean, to me it's still quite an open question if it will ever)
17:25 timotimo oh?
17:25 timotimo well, i'll likely be working on the performance aspect for the foreseeable future
17:25 PerlJam grondilu: Is that really true though?  P5 never beat P4 in the performance dept. and mostly nobody cared.
17:25 timotimo though to be perfectly honest, i'm not really that effective at speeding up the "big picture"
17:25 [Coke] did a git pull on panda, and then a bootstrap worked. tried rebootstrap, which is now dying.
17:26 [Coke] with: No such method 'ast' for invocant of type 'Any'
17:26 * grondilu did not know P4 was faster than P5
17:26 [Coke] ... and now panda is dying with THAT error message all the time.
17:26 awwaiid I'm starting to edit my one off scripts in ~/bin to be "#!/usr/bin/env rakudo" :)
17:26 [Coke] grumpy coke is grumpy
17:27 PerlJam For a good while it was one of the things bugging me that P5 was slower than P4.  Eventually (for me), the features the P5 provided made the "slowerness" unimportant.
17:28 PerlJam And P6 has some killer features that P5 is lagging on.  As soon as P6 is "fast enough", it won't matter if it's slower than P5 as long as it's not egregiously so.
17:29 PerlJam (and IMHO, P6 has crossed the magic threshhold of being fast enough, so when most people agree with me, P6's relative speed to P5 won't matter :)
17:31 timotimo [Coke]: damn, that sucks. sorry to hear that!
17:31 timotimo wait ...
17:32 timotimo what's the third thing for 6.0.0 again? shaped arrays, NFG and ...
17:32 timotimo oh, the great list refactor, no?
17:35 PerlJam timotimo: definitely
17:41 timotimo grondilu: as always: if you have specific things that are too slow in your opinion, we'd love to have a look at your code; even better if you have a perl5 or python version that we can compare against!
17:43 dayangkun joined #perl6
17:44 spider-mario joined #perl6
17:47 [Coke] how does one nuke a panda install?
17:47 timotimo hmm
17:48 timotimo i think in the rebootstrap or bootstrap script there should be some rm-rf that'd show you how to do it
17:48 [Coke] I can bootstrap as many times as I want, but the installed panda is borked.
17:48 timotimo damn
17:48 timotimo what does it say?
17:48 [Coke] ... easier to nuke the whole thing.
17:48 timotimo dependency error?
17:48 [Coke] same Ast error.
17:48 timotimo have you tried "panda update"?
17:48 [Coke] same ast error
17:48 timotimo god fucking damnit
17:48 [Coke] ANY invocation of panda dies.
17:48 timotimo what line does it come from?
17:49 timotimo ooooh
17:49 timotimo of course
17:49 [Coke] in sub from-json at lib/JSON/Tiny.pm:51
17:49 timotimo panda opens and parses the module repository thingie first thing on startup
17:49 [Coke] because the projects file was borked at some point
17:50 timotimo you can set DESTDIR in your environment to have it look for a different path
17:50 [Coke] removing the installed projects.json file doesn't help
17:50 timotimo it doesn't?!
17:50 dalek roast: cc6dfe8 | usev6++ | S12-class/attributes.t:
17:50 dalek roast: Add tests for RT #7585 and RT #122109
17:50 dalek roast: review: https://github.com/perl6/roast/commit/cc6dfe8bd8
17:50 synopsebot Link: https://rt.perl.org/rt3//Public/Bug/Display.html?id=122109
17:50 timotimo what about the statefile?
17:50 timotimo is that also borken?
17:51 [Coke] ARGLEBARGLE. "rm -rf ~/perl6;" go back to rakudo dir, "make install". I just built this 10m ago. the installed perl6 DOES NOT WORK.
17:52 [Coke] ah. that's kind of understandable, because our "make" process doesn't know anything about our "config" process, which is kind of borken.
17:53 [Coke] ... which now brings me back to my complaint about installed moar/nqp vs. locally built moar/nqp. :P
17:53 [Coke] rebuilding moar/nqp now, even though if I had been clever it wouldn't have been necessary.
17:57 timotimo moritz: do you recall how much faster Str.trans for "\n" x 10000 became when you improved the O(n²) behavior?
17:58 dalek Perlito: 997461e | (Flavio S. Glock)++ | / (2 files):
17:58 dalek Perlito: Perlito5 - js - implement special syntax "PATTERN" for split() - https://github.com/fglock/Perlito/issues/22
17:58 dalek Perlito: review: https://github.com/fglock/Perlito/commit/997461ef04
18:00 Ven joined #perl6
18:03 [Coke] ARGLEBARGLE
18:03 [Coke] completely wiped my perl6 install. reinstalled nqp/moar/perl6 - ran perl6 panda bootstrap
18:04 [Coke] er, sorry, "perl6 bootstrap.pl" - installed... installed panda is broken, same error as before.
18:04 [Coke] No such method 'ast' for invocant of type 'Any'
18:06 [Coke] ah, right. https://github.com/tadzik/panda/issues/96
18:10 timotimo who did FROGGS work together with to improve rakudo's (well, MoarVM's) packagability?
18:10 timotimo the maintainer of our rakudo packages for SuSE i believe
18:22 FROGGS joined #perl6
18:23 Mouq joined #perl6
18:31 dalek Perlito: 491ce2e | (Flavio S. Glock)++ | / (2 files):
18:31 dalek Perlito: Perlito5 - js - split() runtime accepts regex as first argument - https://github.com/fglock/Perlito/issues/22
18:31 dalek Perlito: review: https://github.com/fglock/Perlito/commit/491ce2e446
18:35 liztormato joined #perl6
18:36 kjs_ joined #perl6
18:36 liztormato [Coke]: Sorry to hear your panda troubles. It's probably my fault.
18:37 liztormato FWIW I currently always do a perl6 rebootstrap to make sure I didn't break panda
18:38 liztormato It works on my machine ;-(
18:39 liztormato Should be home in about an hour. I will then double check right away.
18:41 amkrankruleuen joined #perl6
18:43 [Coke] liztormato: it will work if I wipe the projects.json file and then try to bootstrap when I'm not forced to use an http proxy
18:44 liztormato Ok. So you have a workaround
18:47 liztormato left #perl6
18:50 pierrot joined #perl6
18:52 NCcoachie joined #perl6
19:02 Flogator joined #perl6
19:07 liztormato joined #perl6
19:07 kjs_ joined #perl6
19:29 timotimo i weeklyposted
19:31 brrt joined #perl6
19:31 vendethiel noice
19:32 masak http://p6weekly.wordpress.com/2014/10/20/2014-41-42-apw-glr-progress-perl%E2%80%AF6-ide/
19:32 masak (for those of us who would otherwise have to search for the blog and post)
19:32 masak the first hit of my search was http://www.perl.com/pub/2002/09/p6pdigest/20020915.html
19:33 PerlJam masak++
19:33 PerlJam timotimo++
19:33 fhelmberger joined #perl6
19:37 timotimo masak: "it's surprisingly easy to stumble upon very outdated documents with a search engine" :P
19:38 PerlJam Ah ... I see I volunteered for this month's compiler release.   Suggestions for the release name are welcome.   I wanted to use Salzburg, but there doesn't appear to be a Perl Mongers group there.
19:40 masak PerlJam: quick! start one!
19:40 masak also, which one is the closest?
19:40 PerlJam btw, does "Salzburg" literally mean "salt town"?
19:41 masak PerlJam: yes.
19:41 PerlJam Do they have lots of salt or something?
19:41 masak PerlJam: how else would you preserve your Hamburg?
19:41 PerlJam A major salt mine?
19:41 mauke salt castle
19:41 PerlJam mauke: I'd like to see this castle made of salt :)
19:42 mauke I mean, that's the literal translation
19:42 vendethiel masak: this read was a very good one :-)
19:42 masak yes :)
19:44 cognominal pmichaud++ and timotimo++ for their blog and (semi-)weekly summary.
19:46 masak today's wacky DSL idea: a subroutine that takes a block and returns a dynamically constructed object with an ACCEPT method, to be passed to a .grep
19:46 masak actually used in actual code I'm writing(TM)
19:48 _sri joined #perl6
19:48 PerlJam maska++ your latest macrotic rambling has lots of ... "character"  :)
19:49 PerlJam masak_++ even
19:49 PerlJam blah!  I can't even type your nick
19:49 PerlJam masak++  (typed. real. slow. like. that time :)
19:49 masak PerlJam: sometimes the lack of rigor in the macros spec is... disappointing.
19:52 Flogator Hey all :) Im suprised seeing so many people in this IRC^^ will perl6 actuallly be published some day? sry for the question - but it's really been a while ;)
19:52 masak Flogator: first off, welcome!
19:52 masak Flogator: this is a common question here :)
19:52 PerlJam Flogator: that statement and question are oddly familiar even
19:52 PerlJam (exact phrasing)
19:52 masak Flogator: we're happy to answer it for you as best we can.
19:52 masak Flogator: first approximation at an answer.
19:53 masak Flogator: "yes."
19:53 masak someone else have a go :)
19:53 PerlJam Flogator: to my mind, Perl 6 is "published" every month.  In fact, I'm "publishing" it this month in a few days
19:53 masak Flogator: PerlJam's answer is the typical "we're releasing it, but no-one seems to notice" answer.
19:54 masak Flogator: I am now providing commentary on it for your elucidation.
19:54 Flogator :) I'm just wondering wheter it's worth beeing a "early adopter" if your not interested in developing it? So can I actually work with it already?
19:54 Flogator youre great guys!
19:54 masak Flogator: many of us here are early adopters like that.
19:55 masak Flogator: I've been an early adopter since around 2008.
19:55 masak Flogator: yes, you can definitely work with it already. I've written all kinds of stuff in Perl 6. some of it is in production now.
19:55 masak Flogator: but (as with all early stuff) expect some rough edges.
19:55 brrt and there's definitively active development
19:55 Flogator masak: I like that!
19:56 masak Flogator: I've also submitted around 1500 bug reports.
19:56 masak (though I'm an outlier, admittedly)
19:56 Flogator masak: I guess that's sure ;) (and a little scary^^ but in 6 years - ok)
19:57 masak yeah, that's much less than one per day.
19:57 mauke haha
19:57 masak it used to be around one per day, but then the bugs stopped coming as quickly.
19:57 PerlJam masak: Outliers are important and necessary!  See  http://www.ted.com/talks/shawn_achor_the_happy_secret_to_better_work   :-)
19:57 masak I blame nom and 6model -- things got so bloody stable after that. <grumble> :P
19:58 masak PerlJam: interesting. will watch.
20:00 vendethiel Flogator: I've started reading stuff on perl6 in february, started using it in prod around 2 months later
20:02 masak Flogator: if you're curious about the quickest way into the community, it's this: (a) pick a task that seems way too small -- "surely this should be easy!", (b) find and submit your first bug, (c) wake up half a year later, covered in TODOs, wondering what the heck happened.
20:02 treehug8_ joined #perl6
20:02 masak 1/2 :P
20:02 brrt we're definitely more production-ready than we were before
20:02 dalek rakudo/nom: d04b5b4 | duff++ | docs/announce/2014.10.md:
20:02 dalek rakudo/nom: a templatey first stab at the Oct release announcement
20:02 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d04b5b467d
20:02 moritz PerlJam++
20:02 Flogator ok - thanks alot- guess I will look into it^^
20:03 vendethiel brrt: "we're nearer than before" is a pretty useless comparison :P
20:03 vendethiel I certainly hope we havn't regressed!
20:03 brrt let's make it less useless then
20:04 PerlJam vendethiel: it does happen occasionally.
20:04 brrt i think we're on the upper half of the sigmoidal readiness curve
20:04 masak suggesting that from now on, convergence will be an ever-slowing process!
20:05 brrt not suggesting that your personal readiness cutoff is equal to the maximum though :-)
20:05 PerlJam nah, the target will move too, so the rate of convergence will sometimes speed up and sometimes slow down
20:05 masak wow, this is some of the weirdest code I've ever written.
20:06 vendethiel .oO( and that says a lot )
20:06 vendethiel masak: we're eagerly awaiting for the gist :P
20:07 masak vendethiel: https://gist.github.com/masak/67bbd56426b12d0e5943
20:08 masak (I'm not finished with the rest, so I can't post the whole code yet)
20:08 dalek specs: de5fbbb | usev6++ | S32-setting-library/Exception.pod:
20:08 dalek specs: [S32::Exception] Add X::Syntax::Term::MissingInitializer
20:08 dalek specs: review: https://github.com/perl6/specs/commit/de5fbbb204
20:08 PerlJam masak: you clearly have an affinity for games ;)
20:08 masak I believe line 4 is necessary, to lexically capture the dynamical.
20:08 masak PerlJam: I do.
20:09 masak PerlJam: in this case, I'm building a searchable database for a https://en.wikipedia.org/wiki/Hex_(board_game) corpus
20:10 kjs_ joined #perl6
20:10 PerlJam That line 4 business reminds me a little of the hoops you have to jump through in JS to make sure you get the right "this" sometimes.
20:11 masak PerlJam: yes, I had similar associations.
20:11 masak PerlJam: I *think* that line is necessary, though it's possible it's not.
20:11 PerlJam I tend to hope it's not :)
20:12 masak ok, this requires some empirical SCIENCE.
20:13 masak m: sub foo { my $*DYN = 42; say $*DYN; return Any.new but role { method bar { sy $*DYN } } }; my $o = foo; $o.bar
20:13 camelia rakudo-moar 7ff58f: OUTPUT«(timeout)»
20:13 masak locally, "42\nDynamic variable $*DYN not found"
20:13 masak ok, so it's necessary.
20:14 masak I guess the moral is: "dynamical variables aren't lexically bound to their original scope". and it would be kind of strange if they were.
20:14 masak and probably counterproductive.
20:14 jack_rabbit joined #perl6
20:14 pecastro joined #perl6
20:15 Ven joined #perl6
20:16 vendethiel do we have a .compact to remove any-values?
20:16 masak a what to what?
20:16 vendethiel right :
20:16 PerlJam masak: aye, but I still find it distasteful that there's this random assignment whose only purpose is so that the closure works right
20:17 masak PerlJam: I 50% agree with you, and 50% "yeah, but it's not 'only'; you need a lexical binding for the closures to work; that's how they work"
20:17 PerlJam masak: maybe we need a macro that enables us to say "close over this thing right here"  ;)
20:18 PerlJam (which makes an anonymous lexical binding for us0
20:18 PerlJam s/0/)/
20:18 masak <shaky-handwriting>DO NOT MESS WITH LEXICAL BINDING AND CLOSURES</shaky-handwriting>
20:18 Ven can I get a count of matching junctions :D?
20:19 masak Ven: received wisdom is that "you want a Set or something that you can look into"
20:19 Ven not really
20:19 Ven m: my $i = +(any(/a/, /b/) ~~ /ab/)
20:19 camelia rakudo-moar 7ff58f: OUTPUT«use of uninitialized value of type Nil in numeric context  in block <unit> at /tmp/KqT9B_kAMY:1␤␤»
20:20 PerlJam vendethiel: um ... what?
20:20 masak I am now creating a sub that wraps a dynamical sub in a sub and stores it back in the dynamical.
20:20 PerlJam er, Ven
20:20 vendethiel PerlJam: vendethiel is fine :-)
20:21 masak vendethiel: you are Not Meant To Care about the way junctions look inside. if you care about the number of elements, then you care about the inner structure of a junction.
20:21 Ven I've just used "+@blah.grep(x|y)" :D
20:21 Ven masak: I don't care about the number of elements
20:21 masak ok.
20:21 PerlJam I'm only getting more confused then
20:22 PerlJam and so ... I'll take this opportunity to go back to hacking for $work  :)
20:24 brrt left #perl6
20:26 masak omg, the code works o.O
20:26 Ven omg!
20:26 Ven must be a bug
20:27 masak no; just massive cheating(tm) and time to write more tests(tm)
20:27 Ven ™™ <- here's some for you
20:28 molaf joined #perl6
20:34 masak 😸
20:36 rurban joined #perl6
20:42 * [Coke] reads macrotic as necrotic.
20:56 dalek rakudo/nom: df4b494 | moritz++ | docs/announce/2014.10.md:
20:56 dalek rakudo/nom: [release] Announce that future Rakudos will need ICU in their parrot backend
20:56 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/df4b49441b
20:57 masak [Coke]: henceforth, my professional title will be "the Macromancer"
20:59 [Coke] moritz++ masak++ # but not for the same thing
21:00 masak things++ # for sometimes differing
21:02 lizmat [Coke]: so, should I investigate panda breakage or not?
21:02 yoleaux 14:33Z <jnthn> lizmat: It's better to do async communication with me these days than try and find a time we're both on IRC, I suspect. :) Just drop the code example that shows it or other stuff I need to reproduce it in a gist and .tell me it ;)
21:08 lizmat timotimo++ for weekly update!
21:08 lizmat fwiw, I think TimToady's statement about the 3 things we need for 6.0.0, is a good guideline for Xmas to come
21:09 lizmat GLR, NFG and NSA  :-)
21:09 kjs_ joined #perl6
21:11 [Coke] lizmat: there is already a ticket for my issue.
21:11 [Coke] it's from at least august, I think - wasn't you
21:11 lizmat ah, ok, *phew*
21:11 lizmat having broken so many things, I've become a bit sensitive to these things  :-)
21:12 [Coke] I forgot that I was having the same problem I've had in the past, and kept getting moar frustrated until I found the old ticket.
21:12 [Coke] nah, you're cool. no worries.
21:12 Tekk_ joined #perl6
21:12 Tekk_ lizmat: I'm so used to #perl being dead on here relative to irc.perl.org :P
21:12 lizmat :-)
21:12 cognominal_ joined #perl6
21:12 pmichaud good afternoon, #perl6
21:13 moritz good pm, pm
21:13 [Coke] good morning, pmichaud!
21:13 Tekk_ so, what's the best way to do globbing in 6? I got as far as IO::Path.contents but I can't figure out how to get :test to take a regex
21:14 Tekk_ or much of anything really
21:14 moritz Tekk_: you could always just .grep the results
21:14 moritz m: say '.'.path.contests
21:14 moritz m: say '.'.path.contents
21:14 camelia rakudo-moar 7ff58f: OUTPUT«No such method 'contests' for invocant of type 'IO::Path'␤  in block <unit> at /tmp/ffRTN2dlGd:1␤␤»
21:14 camelia rakudo-moar 7ff58f: OUTPUT«"/home/p6eval/jvm-rakudo".IO "/home/p6eval/star-2014.08".IO "/home/p6eval/.viminfo".IO "/home/p6eval/rakudo-inst-2".IO "/home/p6eval/.pugs_history".IO "/home/p6eval/.selected_editor".IO "/home/p6eval/star".IO "/home/p6eval/rakudo1".IO "/home/p6eval/.bash_h…»
21:15 [Coke] m: glob(/^ 'rak'/).say
21:15 camelia rakudo-moar 7ff58f: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/2DjJgCcSHiâ�¤Undeclared routine:â�¤    glob used at line 1â�¤â�¤Â»
21:15 moritz m: say '.'.path.contents.grep({ .Str ~~ /pugs/ })
21:15 camelia rakudo-moar 7ff58f: OUTPUT«"/home/p6eval/.pugs_history".IO␤»
21:15 lizmat Tekk_: $ 6 '.say for ".".IO.contents: :test(/foo/)'
21:15 lizmat "/Users/liz/Github/rakudo.moar/foo.t".IO
21:15 lizmat "/Users/liz/Github/rakudo.moar/fooz".IO
21:15 [Coke] (glob is referenced in S32/IO ...)
21:15 moritz m: say '.'.path.contents(test => /pugs/)
21:15 camelia rakudo-moar 7ff58f: OUTPUT«"/home/p6eval/.pugs_history".IO␤»
21:15 pmichaud m: say '.'.path.contents.grep(/pugs/)  # curious
21:15 camelia rakudo-moar 7ff58f: OUTPUT«"/home/p6eval/.pugs_history".IO␤»
21:16 moritz Tekk_: just do it right, and it'll work :-)
21:16 lizmat [Coke]: that's really a fossil, atm
21:17 lizmat or at least part of the "Conjectural Stuff"
21:17 Tekk_ ah, I just had the syntax wrong then :)
21:18 pmichaud m: say '.'.path.contents.grep(/rak/)  # curious #2
21:18 lizmat it's because of this functionality, that I don't think we need glob() at all, really
21:18 camelia rakudo-moar 7ff58f: OUTPUT«"/home/p6eval/jvm-rakudo".IO "/home/p6eval/rakudo-inst-2".IO "/home/p6eval/rakudo1".IO "/home/p6eval/rakudo-inst-1".IO "/home/p6eval/rakudo-inst".IO "/home/p6eval/rakudo".IO "/home/p6eval/rakudo2".IO␤»
21:18 pmichaud \o/
21:18 Tekk_ lizmat: yeah
21:19 Tekk_ I just had to figure out how you're supposed to do it :)
21:19 * TimToady wonders when convenience suddenly became passe...
21:19 lizmat BTW, .IO.contents is deprecated as of 2014.10, please use .IO.dir instead  :-)
21:20 Tekk_ TimToady: when the inconvenient alternative is sufficiently in style. Java being a fantastic example
21:20 lizmat TimToady: perhaps since I rewrote large parts of S16/S32:IO ?
21:22 lizmat glob() was still NYI, fwiw
21:23 lizmat if we have a good spec, we could implement it  :-)
21:23 lizmat is "perldoc -f glob" is the spec, we could do that too
21:23 TimToady you don't want me speccing when I have a fever :)
21:23 lizmat *if
21:24 Tekk_ lizmat: do you mean IO.directory?
21:24 Tekk_ rakudo fails saying that .dir is deprecated :)
21:24 Tekk_ or maybe I have an ancient rakudo..
21:24 lizmat yes, that must be ancient
21:24 Tekk_ 2013.05. Yep, that qualifies as ancient...
21:24 * Tekk_ goes to fix the slackbuild
21:24 lizmat you also might want to try moarVM instead
21:24 lizmat you will be surprised by its speed
21:25 Tekk_ not having 2 second start times on the interpreter sounds pretty nice
21:25 Tekk_ that was annoying me quite a bit
21:25 lizmat $ 6 1
21:25 lizmat real0m0.209s
21:27 cognominal_ lizmat, it seems that your recent changes have borked local-fetch in Panda::Fetcher as per the error messages when rebootstrap.pl-ing panda.
21:27 * Tekk_ has rakudobrew doing its thing, instead of using slackbuilds :P
21:27 cognominal_ No such method 'cleanup' for invocant of type 'Str'
21:27 cognominal_ in sub local-fetch at /Users/cog/git/panda/lib/Panda/Fetcher.pm:46
21:28 * lizmat looks
21:28 cognominal_ I am so out ouf tune with the perl6 recent development I may have missed somthing important.
21:29 TimToady btw, mkdir(Date.today) stopped working
21:29 lizmat cognominal_: how do you invoke that ?
21:29 [Coke] ^^ someone add a spectest.
21:30 cognominal_ cd ~/git/panda  # go to the panda repository (with my setting)
21:30 cognominal_ perl6 rebootstrap.pl
21:31 lizmat TimToady: this is what you see?
21:31 lizmat $ 6 'mkdir(Date.today)'
21:31 lizmat No such method 'IO' for invocant of type 'Date'
21:31 cognominal_ my $cleanup       = $from.IO.path.cleanup;   # The line where it blows up.  Being an IO.path stuff, I dare to infer this is due to your recent changes
21:32 lizmat aha... but maybe you need to pull panda then
21:32 lizmat my version reads: my $cleanup       = $from.IO.cleanup
21:35 lizmat TimToady: perhaps we need to move method Cool.IO to Any
21:36 * lizmat tests that
21:37 kjs_ joined #perl6
21:38 cognominal_ lizmat++, sorry I was on a fork. Silly me.
21:42 denis_boyun_ joined #perl6
21:42 lizmat hmmm... interesting failure mode again  :-(
21:42 lizmat Stage parse      : moar(54226,0x7fff7c1a8310) malloc: *** error for object 0x7f8dc4e95c70: pointer being freed was not allocated
21:42 masak 'night, #perl6
21:42 lizmat gnight, masak
21:47 timotimo good day TimToady
21:47 lizmat grrrr... it seems that just a "take" in jvm bombs ?
21:48 timotimo because it's outside a gather?
21:48 lizmat no, it's very much inside
21:50 lizmat timotimo: see src/core/Baggy.pm, sub ROLLPICKGRABN
21:51 timotimo :(
21:53 dalek rakudo/nom: 3975e16 | (Elizabeth Mattijsen)++ | src/core/Temporal.pm:
21:53 dalek rakudo/nom: Fix mkdir/rmdir(Date) breakage TimToady++
21:53 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/3975e16b10
21:53 lizmat timotimo: do you see anything wrong there ?
21:55 lizmat .tell jnthn suggestions about take failing on jvm? https://gist.github.com/lizmat/ab00347d5cc44a23c65b
21:55 yoleaux lizmat: I'll pass your message to jnthn.
21:56 lizmat .tell jnthn code is from src/core/Baggy.pm
21:56 yoleaux lizmat: I'll pass your message to jnthn.
21:58 thou joined #perl6
22:03 kjs_ joined #perl6
22:03 lizmat .tell jnthn the gist with the sub/private method codegen issue: https://gist.github.com/lizmat/16382a07084c5bb33d19
22:03 yoleaux lizmat: I'll pass your message to jnthn.
22:05 lizmat ooh wow
22:05 lizmat seems if I take away the extra { } of the gather, it works
22:06 * lizmat is going to test that on moar as well
22:06 timotimo lizmat: wat :(
22:06 timotimo glad it works, but ...
22:06 colomon joined #perl6
22:11 dalek rakudo/nom: f0efc88 | (Elizabeth Mattijsen)++ | src/core/Baggy.pm:
22:11 dalek rakudo/nom: Apparently fix "take" failure on JVM
22:11 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/f0efc88310
22:32 erkan joined #perl6
22:32 erkan joined #perl6
22:36 camelia joined #perl6
22:48 colomon joined #perl6
22:52 lizmat alas, that didn't fix it after all...
22:52 lizmat giving up for today.... gnight #perl6!
22:58 colomon o/
23:05 timotimo hi u
23:13 BenGoldberg joined #perl6
23:21 japhb .botsnack
23:21 yoleaux :D
23:25 davido__ joined #perl6
23:33 kurahaupo joined #perl6
23:34 fhelmberger joined #perl6
23:55 virtualsue joined #perl6

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

Perl 6 | Reference Documentation | Rakudo