»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
00:11 araujo left, araujo joined
dalek ast: ffb801f | (David Warring)++ | S29-conversions/ord_and_chr.t:
chr() on invalid code-points
 00:14
TimToady .tell jnthn also, if implicit method new is a no-op wrapper around .bless, we can probably just poke the .bless routine in under the name of .new to call it directly 00:18
yoleaux TimToady: I'll pass your message to jnthn.
00:23 zwut00 left 00:33 BenGoldberg joined 00:47 Bzek joined 00:50 Bzek left 01:23 bluescreen10 joined 01:26 FROGGS_ joined 01:31 FROGGS left 02:06 xinming left 02:10 xinming joined 02:24 jaldhar left, jaldhar joined 02:35 Targen left 03:05 xenoterracide_ is now known as xenoterracide 03:11 jeff_s2 joined 03:12 jeff_s1 left 03:21 preflex_ joined, ChanServ sets mode: +v preflex_ 03:22 preflex left, preflex_ is now known as preflex 03:33 abnorman joined 03:51 BenGoldberg left 03:53 thou joined 04:16 aindilis joined 04:24 thou left, grondilu joined 04:33 thou joined 04:37 yoleaux left 04:42 cbk joined 05:02 hummeleBop joined 05:04 SamuraiJack joined
dalek kudo/nom: 7cb364b | (Geoffrey Broadwell)++ | tools/build/NQP_REVISION:
Bump NQP_REVISION so that Rakudo will get the fix for ConfigureJVM.pl in latest NQP
 05:05
05:28 ksh joined
dalek ar: dddc8a5 | moritz++ | modules/rakudo-debugger:
updated rakudo-debugger submodule
 05:33
05:51 ggoebel2 joined 05:52 ggoebel left 06:02 denis_boyun joined 06:18 denis_boyun left, denisboyun joined 06:33 yoleaux joined, ChanServ sets mode: +v yoleaux 06:40 xenoterracide left, xenoterracide_ joined 06:42 Mouq left 06:51 kaleem joined 06:54 zakharyas joined
[Coke] (OS X failures) - just updated this list: gist.github.com/coke/5428159 - it's been failing that way for months and months. 06:59
(rakudo.parrot)
07:01 thou left 07:04 ksh left 07:07 ksh joined 07:17 sidus joined
lizmat good *, #perl6! 07:18
bbkr speaking of OS X - Rakudo Star tools/star/Makefile doesn't work due to lack of "tar --xform". should it be reported as an issue or we assume tarball can be only created under Linux?
good morning :)
lizmat [Coke]: all but the last indicate an improperly installed icu library
oddly enough, after a rebuild of parrot yesterday, all of those disappeared for me 07:19
(must have installed icu properly in the meantime)
dalek kudo-star-daily: 7fbccfd | coke++ | log/ (5 files):
today (automated commit)
lizmat the last one is from missing libreadline-dev library
still seeing that as well
breakfast& 07:20
[Coke] (things to help rakudo star release - fix github.com/rakudo/star/issues/27 07:23
lizmat: macports, icu-49.1.2 07:25
readline-dev is optional, shouldn't be running the tests if we don't have it.
(well, optional for parrot)
bbkr: I think everyone who does the build so far has access to linux, so not a showstopper. 07:26
bbkr [Coke]: Not a showstopper but I've reported it anyway as low priority issue to be taken care of during future refactoring of build process - github.com/rakudo/star/issues/28 07:33
[Coke] hokay 07:34
dalek rl6-roast-data: 3e42306 | coke++ | / (5 files):
today (automated commit)
 07:35
[Coke] another regression on rakudo.jvm 07:36
+S32-io/io-path-win.rakudo.jvm aborted 19 test(s) 07:37
+S32-io/io-spec-win.t aborted 33 test(s)
+integration/advent2009-day02.rakudo.jvm aborted 3 test(s)
+S02-names-vars/fmt.t aborted 14 test(s)
note that S02-magicals/env.rakudo.jvm 14 - ENV members persist to child processes is still failing - i thought we merged a fix for that. 07:38
nwc10 bbkr: I see that I have a /usr/bin/gnutar on OS X. So in theory OS X is OK. But likely most of the BSDs are not, let alone other sane Unix platforms.
mmm, but my gnutar on OS X seems to be too old to have --xform 07:39
bbkr mine is "bsdtar 2.8.3 - libarchive 2.8.3"
nwc10 oh, quirky. The *man page* doesn't have it
tar is bsdtar. gnutar is a GNU tar
moritz tarpit tar! 07:40
nwc10 nope. info gnutar is also out of date 07:41
anyway, not sure what the right trade off is. --xform is useful
[Coke] eh. we have a very small audience for cutting * releases. 07:48
(and there's always feather)
[Coke] tries updating ICU. 07:49
tadzik lizmat: just to make double-sure: 14-15 september, frankfurt am mein? 07:54
well, tickets bought :) 07:58
07:58 crab2313 joined 07:59 pernatiy left 08:01 atroxaper joined 08:02 pernatiy joined
timotimo tadzik: that's correct indeed 08:12
dalek rlito: 48b3691 | (Flavio S. Glock)++ | / (3 files):
Perlito5 - js - tweak the javascript pretty-printer
 08:17
08:22 fhelmberger joined, Rotwang joined
jnthn morning, #perl6 08:23
yoleaux 00:18Z <TimToady> jnthn: also, if implicit method new is a no-op wrapper around .bless, we can probably just poke the .bless routine in under the name of .new to call it directly
jnthn Epic cheating :P
tadzik so, my AFTERPARTY method may not be that useless after all? ;) 08:28
08:37 kaleem left
lizmat tadzik: am Main, ja! 08:40
08:41 jeff_s2 left
timotimo no star release in a long time. i feel like i should punch some modules in the bugs! 08:41
08:44 dmol joined, atroxaper left
jnthn
.oO( Which Frankfurt? The main one! )
 08:44
08:46 bluescreen10 left 08:48 Rotwang left
lizmat
.oO( Mainhattan )
 08:48
nwc10 I can't hugely make this work: Hahn == Hahn 'aint hanywhere near 08:51
08:52 stevan_ joined
jnthn no...no, you can't :P 08:53
nwc10 and I'm utterly failing on Frankfurt am Elbe puns 08:54
give up and write code?
08:55 pecastro joined 08:58 berekuk joined
jnthn It code be a good idea 08:59
09:02 dmol left 09:04 daxim joined
lizmat jnthn: is gist.github.com/lizmat/6177720 still an accurate description of the RaNIW ? 09:05
jnthn lizmat: There's the occasional bit of re-ordering within the modules that I've discovered while working on it, but the topic list still matches up, yeah. 09:07
So, "close enough" :)
09:07 donaldh joined
lizmat ok, so I'll put this in perl6-compiler, see if we can attract more participants (currently 6 confirmed + you) 09:08
which then should hit planetsix 09:09
hmm.. maybe need to blog about it as well :-) 09:10
09:15 stevan_ left
nwc10 bloggage is awesome. 09:17
mention (in passing, eg 1 sentance or less) the JVM-ness and the concurrency stuff?
09:17 dayangkun joined 09:18 dayangkun left
donaldh morning o/ 09:24
09:24 abnorman left, abnorman joined 09:34 jeff_s1 joined 09:35 rindolf joined
dalek ecs: 93dd4af | (Donald Hunter)++ | S29-functions.pod:
Insert line to fix 'run =item shell' on perlcabal.org/syn/S29.html
 09:35
donaldh hmm, cwd doesn't seem to be specced anywhere 09:38
09:38 jeff_s1 left 09:46 rindolf left, rindolf joined 09:50 rindolf left, rindolf joined 09:52 rindolf left 09:53 rindolf joined
donaldh How do I create a PROCESS::chdir ? 09:53
I thought it would just be: module PROCESS { multi sub chdir($path) { … } } 09:54
jnthn PROCESS::<&chdir> = anon sub chdir($dir) { ... }
donaldh ah
And I can do that in IO.pm, i.e. right in the middle of CORE.setting ? 09:55
jnthn Or if you need multis then put it in an anonymous block and then poke it in
{
proto sub chdir($) {*}
...multi candidates...
PROCESS::<&chdir> := &chdir;
}
Or so
09:55 jeff_s1 joined
donaldh thx 09:55
jnthn The block stopping it leaking into user-visible space.
09:59 hypolin joined 10:03 abnorman left, dmol joined, abnorman joined
lizmat blogs.perl.org/users/liz/2013/08/ra...5-sep.html 10:06
FROGGS_ awesome! 10:08
10:08 FROGGS_ is now known as FROGGS
jnthn lizmat++ 10:09
daxim will the materials be published?
lizmat the materials are supposed to be open source 10:10
daxim excellent
lizmat but, at the discretion of jnthn and Edument
they're free to give the workshop again
daxim are edument aware this is niche knowledge?
lizmat I'm pretty sure they are 10:11
jnthn yes :)
It is very nice to be doing Perl 6 things as part of my $dayjob :) 10:12
lizmat OTOH, that means it's the ultimate growth market :-)
10:13 rindolf left 10:14 rindolf joined
FROGGS when looking at this it seems like nqp::sprintf doesn't break anything! github.com/coke/rakudo-star-daily/...0a2#diff-2 10:14
what a relief
lizmat wonders how to get my blog post on planetsix ? 10:21
tadzik contact planetsix admins, I don't remember who that is :) 10:25
[Coke], maybe?
10:40 fhelmberger_ joined
hoelzro 's interest is piqued by this workshop 10:41
tadzik komm, komm 10:42
10:44 fhelmberger left, sqirrel joined
hoelzro my girlfriend will kill me if I leave her alone for weekends so close together =) 10:45
will there be screencasts?
donaldh would love to go, but cannot unfortunately. 10:46
jnthn hoelzro: No screencast, I'm afraid. 10:47
hoelzro curses
jnthn If it works out well, maybe another one will happen in the future... 10:48
lizmat when it works out well :-)
wrt to screencast: not very interesting seeing up to 12 core developers to be sweating doing jnthn's exercises :-) 10:49
tadzik it'll be like a realistic hacker movie 10:50
"they're just sitting there... for 12 hours straight!"
lizmat I'm not sure it will really be up to snuf 10:51
hoelzro hahhaha
I'm looking into hotels now
just in case ;)
moritz I'm sure there'll be lunch in between :-)
lizmat hoelzro: rooms have been booked already
don't worry about hotels, just worry about getting there
hoelzro how much for the rooms, though?
lizmat rooms are on the house, as is the workshop 10:52
10:52 sqirrel left
hoelzro seriously? 10:52
so I just need to get there?
lizmat yup 10:53
hoelzro hmm
this might be an easier sell than I think...
lizmat I didn't want to put that in the announcement
hoelzro ;)
tadzik don't save on your knowledge, especially when it's free :)
hoelzro I have to sell it to my girlfriend is all 10:54
10:54 pernatiy left
moritz hoelzro: where do you live? 10:56
hoelzro Amsterdam
moritz there round-trip flights for ~100EUR 10:57
hoelzro I know
it's cheap
soooo cheap
but I'm going to Bonn tonight
and my girlfriend is not happy about being alone for the weekend =/ 10:58
moritz froscon?
hoelzro mhmm
lizmat moritz: FrosCon is this weekend
donaldh hah, my naive attempt to use IO::Spec.rel2abs in the IO file ops was bound to fail
hoelzro so I have to convince her that leaving her alone again is a good idea =P 10:59
donaldh "Dynamic variable $*CWD not found" while loading the setting
lizmat well, maybe a partner programme should be set up
hoelzro that's an idea 11:00
"A reasonable knowledge of the Perl 6 language" # what qualifies as "reasonable"? 11:01
lizmat woolfy will be in Frankfurt as well, but probably not at the workshop all the time
jnthn hoelzro: As in, have written some Perl 6 before, can read the common stuff without difficulty, etc. 11:02
lizmat looking at the program of the workshop, is it *all* klingon to you, or does at least 40% of it mean something to you
hoelzro ok
I wrote an XMPP bot last year 11:03
jnthn hoelzro: We'll be seeing a lot of NQP code, which is a Perl 6 subset.
hoelzro and a few small programs
I've dealt with NQP a bit
not much, though
lizmat I think you qualify :-)
11:03 crab2313 left
jnthn I'd say so. 11:03
hoelzro my limits are at introspection and the crazy Z <<+>> stuff
ok, cool =)
lizmat those limits will be stretched :-) 11:04
hoelzro yay
FROGGS lizmat: why does klingon implies that one can't understand it?
jnthn If you think you're reading klingon when it's actually English, you don't understand it :P 11:05
FROGGS lizmat: you can have on of my copies of klingon<->deutsch :o) 11:06
lizmat I was going for "chinese" originally, but figured the chance of chinese reading people looking, was too large 11:07
11:08 jeff_s1 left 11:10 jeff_s1 joined 11:13 hypolin left 11:29 dagerik joined
dagerik use v6;my @f=1,1,{$^a+$^b}...*;say @f[42]~".html"; 11:29
what does this do?
moritz r: use v6;my @f=1,1,{$^a+$^b}...*;say @f[42]~".html" 11:31
camelia rakudo 7cb364: OUTPUT«433494437.html␤»
moritz dagerik: produces the string "433494437.html" :-)
dagerik: where 433494437 is the 42s fibonacci number
tadzik 43rd, is 0 is the 1st one
I never know :)
moritz ... given the appropriate definition of fibonacci numbers that make my statement true :-) 11:32
*makes
tadzik yep :) 11:34
11:35 abnorman left, abnorman joined 11:42 average joined 11:43 rindolf left 11:46 dagerik left, pernatiy joined
lizmat jnthn: looking at handling the my %h{Str} of Int case 11:52
I think I can fix the problem, but with a leak
arnsholt The entry for Perl 5 on ian.hixie.ch/programming/ under the Paradigm heading: "Yes" =D 11:53
lizmat container_type_info creates the improper (Any,Str) type
11:53 denisboyun left
tadzik :D 11:53
lizmat I can probably tiwddle the container description afterwards, with the proper (Int,Str) tyoe 11:54
but the (Any,Str) type would still be created
so I'm looking at moving create_container_descriptor / install lexical container until after the trait handling 11:55
with the $var.compose method doing all the work, and the trait-mods just setting up stuff to be done later 11:57
does that make sense ?
FROGGS r: test 11:58
camelia rakudo 7cb364: OUTPUT«===SORRY!=== Error while compiling /tmp/pDWpO_9KYu␤Undeclared routine:␤ test used at line 1␤␤»
jnthn lizmat: Yeah, I think so
lizmat ok, then I'll try to do that 11:59
jnthn lizmat: You may want to use container_type_info partially before the trait mod to have some early bits to populate Variable with.
lizmat but it's container_type_info setting up the wrong type object
using the $shape
anyways, need to pack now for FrosCON 12:00
no more hacking for me today, probably 12:01
dalek ast: 2647c16 | (Tobias Leich)++ | S32-str/sprintf.t:
add sprintf tests from RT tickets
 12:03
12:06 anon3242 joined 12:12 jnap joined 12:13 abnorman left, abnorman joined 12:14 kaleem joined 12:18 locsmif left, _jaldhar joined, jaldhar left 12:19 denisboyun joined
moritz ok, I need help with git submodules 12:29
I've added modules/PSGI to star 12:30
but now that directory is empty
what do I need to do to fill it?
12:30 skids left
donaldh maybe git pull --recurse-submodules, or somesuch 12:31
lee_ git submodule update iirc
12:31 colomon_ left
donaldh what lee_ said 12:31
moritz lee_: that says fatal: reference is not a tree: 240f8d025a72dfd9695290dde45981f238287d79 12:32
Unable to checkout '240f8d025a72dfd9695290dde45981f238287d79' in submodule path 'modules/uri'
how do I recover from that?
hoelzro sweet, the lady approves
moritz feels very n00bish
tadzik hoelzro: \o/
lee_ moritz: hmm, maybe you need to edit .gitmodules ? i haven't seen that one before
hoelzro how do I register?
tadzik poke lizmat
hoelzro lizmat: poke 12:33
donaldh moritz: stackoverflow.com/questions/1476850...odule-path
timotimo .gitmodules doesn't specify what tree is supposed to be checked out, but maybe the 240f... is missing because the sub-repository wasn't pulled from correctly? 12:34
moritz can I tell it to simply use 'master' in modules/uri? 12:35
or phrased differently, can I make it not try to checkout 240f8d? 12:37
dalek ar: e2359a1 | moritz++ | modules/uri:
track modules/uri master
 12:38
moritz ah, I do a 'git checkout master' in that directory, and then 'git add' it from the parent repo 12:39
jnthn I think there' a git submodule foreach than can be useful for automating synchronizing all to latest master, if that's something we want to do...
donaldh moritz++ n00b--
jnthn *that 12:40
donaldh hmm, the return type for chdir is not specced and cwd is not specced at all.
timotimo i used to git submodule foreach 'git fetch && gitk --all' and then i could use the gitk gui to reset the local master branch to any commit i wanted 12:41
donaldh should the return type be Str, or IO::Path ?
dalek ar: 54b9d4c | moritz++ | / (2 files):
update modules/PSGI
moritz donaldh: IO::Path
ok, my modules/PSGI is still empty
wtf?
donaldh moritz: thx - I achieved that by 'accident' ;-) 12:42
moritz ok, re-cloning from scratch helped 12:43
moritz feels like a 1st day git user again
donaldh is cwd not specced because it is not wanted, or because of a spec omission
moritz r: say cwd 12:44
camelia rakudo 7cb364: OUTPUT«/home/p6eval␤»
moritz donaldh: it feels a wee bit redundant do have both $*CWD and cwd 12:45
donaldh: I personally would prefer chdir/cwd over $*CWD, but that's just opinion
donaldh well, $*CWD is the state for chdir emulation
donaldh will add cwd to the spec for now 12:47
FROGGS jnthn: I'm going to outline what I think how slang could be made pluggable this weekend, would be cool if you have time to review it next week or so 12:48
moritz donaldh++ 12:49
jnthn FROGGS: OK. I've gotta go to Oslo for the first chunk of next week to teach, but may still find time on evenings...
FROGGS jnthn: no hurry :o), I just wanted to tell what I am up to 12:50
12:50 ajr joined 12:51 ajr is now known as Guest14547 12:54 Guest14547 is now known as ajr_
daxim what's RTTI? ian.hixie.ch/programming/ 12:54
tadzik Run Time Type Introspection, I think
daxim ahh :|
FROGGS Run-time type information?
according to google's #1 12:55
12:56 benabik left, anon3242 left
moritz next problem with star: PSGI depends on HTTP::Status, but PSGI is compiled first 12:58
jnthn moritz++ # workin gon tstar 12:59
uh, working on star 13:00
timotimo moritz the rakudo tzar? 13:02
lizmat hoelzro: I consider myself poked and on the list for RaNIW
well, you of course :-)
FROGGS that even might be valid klingon: workin gon tstar
you just have to shout it
do we have something in the Perl 6 world that is a collection of classes? 13:03
timotimo everything shouted is either valid klingon or valid german
FROGGS RATZNFATZ! 13:04
yeah :/
hoelzro \o/
just need to book my flight then
timotimo you can tile the whole plane (actually i guess multidimensional space?) of things humans can express while shouting with german and klingon words. only grammar leaves some holes 13:05
FROGGS \o/, our pyg-Man is coming too!
13:06 dalek left 13:07 dalek joined, ChanServ sets mode: +v dalek 13:08 fridim__ joined
hoelzro timotimo: I will not tile the plane! 13:10
they'll kick me off!
lizmat you know, there's an app for that
www.thetileapp.com ;-) 13:11
13:14 fhelmberger_ left, denisboyun left
timotimo lizmat: "has anybody seen my plane? it's yea tall." 13:24
cognominal that's nice, I will buy in when it's 2$ a tile. 13:27
13:28 bluescreen10 joined
dagurval that tile product is the exact same idea that the winners of Startup Weekend Trondheim has over a year ago 13:28
too bad for them that they didn't market it quicker :) 13:29
cognominal My regular phone only use is to find my iPhone, next I will use my IPhone to find my (tiled) keys set. How recursive these searchs can be? 13:31
lizmat fetching the van & 13:32
cognominal I am using public transportation. :) 13:33
… socialist europa 13:34
13:35 jnap left 13:37 gshank joined 13:41 dmol left 13:44 jnap joined, pernatiy_ joined
Juerd lizmat: Wow, that tile thing seems like an incredibly stupid thing, environmentally. 13:44
lizmat: A non-replacable battery that lasts only a year.. 13:45
Throwaway electronics :(
sjn jnthn: ping 13:46
jnthn: how long are you staying in Oslo? 13:47
13:47 bluescreen10 left
jnthn sjn: Arrive Sun, leave Wed midday 13:47
13:47 pernatiy left
jnthn sjn: Tuesday is the good night for me to go out for beer. Sun and Mon I'll have teaching to get up for the next day. :) 13:48
13:49 lowpro30 joined
nwc10 so, Monday in oslo for breakfast and coffee? :-) 13:49
sjn jnthn: right 13:52
tuesday is good \o/ 13:53
13:55 fhelmberger joined 13:58 thou joined 14:01 skids joined
FROGGS jnthn: first draft, note that the S02 pod is taken unmodified from specs: gist.github.com/FROGGS/fbb4455f544a35cf265d 14:02
jnthn: the cool thing is that it is just a about a few lines, nothing special to invent for it...
TimToady: would be awesome if you take a glimpse at gist.github.com/FROGGS/fbb4455f544a35cf265d 14:03
14:05 dmol joined, fhelmberger left 14:07 lowpro30_ joined 14:08 lowpro30 left, dmol left 14:09 dmol joined 14:10 xinming left 14:11 lowpro30_ left, xinming joined
moritz today I learned about en.wikipedia.org/wiki/Auto-antonym 14:13
14:14 lowpro30 joined
timotimo FROGGS: that new method is pseudocode, no? 14:15
14:18 PacoAir joined 14:19 pmurias joined
pmurias nqp::shell() is not implemented on parrot? 14:19
is there a way to express it using pir ops? 14:21
14:22 lowpro30 left
dalek ar: afc39e0 | moritz++ | modules/MODULES.txt:
build HTTP::Status before PSGI
 14:23
ar: 69626a8 | moritz++ | tools/build/Makefile.in:
Rakudo::Debugger does not need perl6_group
14:24 kaleem left 14:25 sidus left
moritz ok, star build is finally cleans 14:26
*clean
moritz doesn't even want to look at the tests...
zavolaj!
timotimo oh zavolaj :( 14:27
pmurias donaldh: ping 14:28
donaldh pmurias: hi
moritz panda!
t/01-file-find.t ...... Failed 1/10 subtests
erm, File::Tools
timotimo yeah, is that the continue-after-exception test? 14:29
moritz dunno, I just see the passing test scroll by
moritz considers TODOing the zavolaj tests
timotimo ah, right.
moritz DBIish seems to work with it
Testing modules/jsonrpc...
t/client.t .. Failed 1/39 subtests
jnthn If DBIish works, zavolaj can only be so busted... 14:30
moritz oh, and LWP::Simple fails most of its tests too :(
14:30 SamuraiJack left
donaldh what is the status of the nio File ops on JVM ? 14:32
work in progress? 14:33
jnthn Yeah 14:34
async slurp and async lines work so far 14:35
14:36 benabik joined 14:37 denis_boyun joined
dalek ar: 7b365da | moritz++ | modules/perl6-lwp-simple:
Use a fixed LWP::Simple
 14:38
moritz anybody wnat to look into that jsonrpc failure?
pmurias donaldh: I'm planning to port nqp::shell to parrot
donaldh pmurias++ 14:39
14:39 Psyche^_ joined
donaldh jnthn: looks like FileHandle.write is needed too 14:39
pmurias is it possible to easily express the one argument version in terms of the 3 argument one 14:40
jnthn donaldh: Oh, wait, you mean for binary I/O
donaldh: That is todo :)
donaldh: In the synchronous case also :)
donaldh k
14:40 rindolf joined
donaldh pmurias: The one argument version is going away... 14:40
tadzik moritz, moritz 14:41
donaldh pmurias: now that rakudo uses the three argument version
tadzik File::Tools needs to be replaced with File::Find and Shell::Command
pmurias and how is nqp code using the 1 argument version supposed to be translated?
tadzik I think there's even a starticket for that
also, open($a, "p") is necessary for IO::Socket::INET tests 14:42
donaldh jnthn: I tried to do spurt('hello.t', 'use v6; say "hello";') and got IO::Handle.write NYI on this backend
14:43 Psyche^ left
donaldh jnthn: so, yeah, I'll try and fill the gaps. 14:43
jnthn donaldh++
tadzik don't we have binary file IO on nqp level?
we have binary sockets IO, should be easy enough to port it
donaldh tadzik: not for JVM.
jnthn tadzik: We don't on JVM yet, no
donaldh looks pretty trivial though
tadzik also, I consider merging the existing sockets code after all
pmurias is there a way to get the cwd on nqp-jvm?
tadzik it's not finished, but it's Something [tm] 14:44
pmurias so that nqp::shell('echo hi',nqp::cwd(),nqp::getenvhash()) can work?
tadzik and thus better than nothing
donaldh pmurias: nqp::cwd does that
pmurias: that works for me 14:47
pmurias I got confused by the definition not including a thread context
donaldh pmurias: github.com/perl6/nqp/commit/4a311e...490773499b 14:48
dalek ar: 0ce8b0d | moritz++ | / (5 files):
replace File::Tools with File::Find and Shell::Command
 14:50
moritz stackoverflow.com/questions/1260748...le#1260982 oh how I love submodules 14:51
pmurias donaldh: how do we get things like the exit code from nqp::shell?
14:52 ajr_ left
donaldh nqp::shell returns the process exit value 14:53
on JVM anyway
14:53 ajr joined
donaldh or 255 if it fails to get as far as executing a shell at all 14:53
FROGGS timotimo: it is a proposal, yes
14:54 ajr is now known as Guest8695
pmurias donaldh: ok 14:54
FROGGS pmurias: grep for spawn
14:55 Guest8695 is now known as ajr_
donaldh On JVM it's handled by the Process class. We do Process.waitFor(); retval = Process.exitValue(); 14:55
14:56 bluescreen10 joined
donaldh pmurias: github.com/rakudo/rakudo/blob/nom/...ol.pm#L202 14:57
FROGGS jnthn: so my plan is to fix :from<NQP>, and then to provide a sample pure Perl 6 slang, using my proposed Slang class 14:59
15:04 _jaldhar left
dalek kudo/nom: e9e99b3 | (Geoffrey Broadwell)++ | tools/build/Makefile-JVM.in:
Support rakudo-jvm builds on smaller systems

Tune JVM memory options to allow successful compile on a 32-bit/2GB RAM system, while still remaining <2% slower to compile the setting on a larger 64-bit system than the current settings.
Max heap size was tested every 100m; -Xmx1100m was minimal for compiles on 32-bit, and -Xmx1300m was minimal on 64-bit, but noticeably slower than current settings. -Xmx1600m seemed to be the sweet spot for minimal slowdown on 64-bit while still being buildable on a 32-bit/2GB RAM system.
Also explicitly set stack size, as the automatic tuning seemed to work fine on the larger 64-bit system, but set a too-small stack on the smaller system, causing the build to fail.
  -Xss1m seems to work well on both.		 15:04
pmurias hmm, how should we test nqp::shell()? 15:06
benabik Carefully. 15:07
FROGGS dunno, is `dir` cross platform enough?
geekosaur nope 15:08
pmurias would special casing windows-- be enough?
nwc10 is `echo` cross platform enough? 15:09
FROGGS I'd say we use `echo` as long we dont have a reason to do something else 15:10
15:14 colomon joined
pmurias is it possible to run nqp-jvm without perl5? 15:17
tadzik run, yes
build, no
15:19 crab2313 joined
pmurias pastie.org/8262991 - would something like that be portable? 15:20
FROGGS I think it works on the platforms I know 15:26
nwc10 pmurias: I think so. You might want to make your regex a bit more robust. I think that /^Hello$/ should be safe
donaldh pmurias: would or should ?
nwc10 um, not portable to VMS. But nothing you have is yet. Cross that bridge when you come to it 15:27
as I have a window open on a VMS machine I tested
donaldh :)
15:29 zakharyas left, lowpro30 joined 15:33 benabik left, lowpro30 left 15:48 dmol left
tadzik hahaha 15:49
"A QA engineer walks into a bar. Runs into a bar. Crawls into a bar. Dances into a bar. Tiptoes into a bar. Rams into a bar. Jumps into a bar"
jnthn Doesn't that make the barman testy? :P 15:50
skids Considering the number of drinks he is about to order, no.
TimToady A QA engineer crawls out of a bar... 15:51
15:52 bbkr_ joined 15:53 woolfy left
lizmat on our way to FrosCON 15:54
&
15:54 lizmat left 15:55 benabik joined
bbkr_ is checking jsonrpc bug right now 15:55
15:58 benabik_ joined 15:59 benabik left, benabik_ is now known as benabik 16:03 daxim left
donaldh Feedback welcome: is this moving in a sane direction for chdir emulation github.com/donaldh/rakudo/commit/8...c88440a9db 16:04
bbkr_ by the way - I confirm that LWP::Simple now works
16:07 btyler joined 16:09 xinming left 16:10 xinming joined
pmurias is there a portable way to call pwd? (so I can test the directory parameter)? 16:11
and something that would use an env variable?
JimmyZ nqp::cwd 16:15
pmurias JimmyZ: from a shell command 16:16
do I need to worry about portability to something other than windows/unix variants? 16:19
huf pwd is a shell builtin usually. what do you want to get out of it?
diakopter pmurias: DOS?
huf i dont think pwd is for anything but interactive use though... 16:20
bbkr_ r: my $x = Nil; $x.WHAT.say # what is going on here?
camelia rakudo e9e99b: OUTPUT«(Any)␤»
donaldh pmurias: perl ?
16:23 cbk left
TimToady bbkr_: you're setting $x back to its defualt of Any by assigning Nil to it 16:23
Nil means "there's a missing value here" 16:24
but a typed variable can't hold that; at its most undefined, it can hold only an undefined type object like Any or Int 16:25
donaldh pmurias: the rakudo ENV tests use is_run which is defined in t/spec/packages/Test/Util.pm 16:26
bbkr_ TimToady: thanks for explanation. this new (or newly implemented) feature breaks JSON::RPC in Rakudo star. now I know how to fix it :)
TimToady r: my Complex $c = 1+2i; say $c; $c = Nil; say $c;
camelia rakudo e9e99b: OUTPUT«1+2i␤(Complex)␤»
bbkr_ r: my Nil $x = Nil; $x = Nil; $x.WHAT.say 16:27
camelia rakudo e9e99b: OUTPUT«Nil␤»
TimToady r: my Mu $x = Nil; say $x
camelia rakudo e9e99b: OUTPUT«(Mu)␤»
TimToady my $x := Nil; say $x 16:28
r: my $x := Nil; say $x
camelia rakudo e9e99b: OUTPUT«Nil␤»
TimToady that works
16:38 fridim__ left
FROGGS r: say Nil 16:39
camelia rakudo e9e99b: OUTPUT«Nil␤»
FROGGS why is it missing the () ?
n: say Nil
camelia niecza v24-88-g1f87209: OUTPUT«Nil␤»
geekosaur it's a value, not a type object? 16:40
flussence Stage parse : 2151.679 # maybe I should'a invested in a new laptop... 16:42
16:42 panchiniak joined 16:44 djrb joined 16:47 djrb left
tadzik wow 16:47
bbkr_ jsonrpc fixed for Star release 16:49
TimToady geekosaur: yes, it's a value, though Nil.WHAT probably returns Nil :) 16:50
r: say Nil.WHAT 16:51
camelia rakudo e9e99b: OUTPUT«Nil␤»
16:51 bbkr_ left
geekosaur yeh, I was wondering how the value vs. type went 16:51
TimToady well, type objects are values too, so Nil is also a missing type object :)
16:54 ajr joined, ajr_ left, ajr is now known as Guest19022, Guest19022 is now known as ajr_ 16:58 average left 16:59 panchiniak left 17:02 spider-mario joined, ajr_ left 17:06 Mouq joined 17:20 betterworld joined
masak o/ 17:20
pmurias \o 17:21
17:21 Targen joined 17:24 Mouq left, Mouq joined 17:25 jeff_s1 left
sjohnson （　｀ー´） 17:26
masak survived the Python teaching :)
sjohnson: hey -- nice to see you around.
sjohnson you too.
admittedly way behind in perl6 news unfortunately :/
flussence stuff happened! 17:27
masak sometimes more than once! 17:28
flussence is only slightly more up to date on things
masak being up to date is overrated.
17:29 BenGoldberg joined
flussence well at least I've got an up to date `perl6` today, I've been neglecting that for a while... 17:29
masak .oO( in Soviet Russia, an up-to-date `perl6` neglects YOU ) 17:31
flussence I attempted the JVM thing, but that only reminded me my laptop's still running java 6... 17:32
17:32 pernatiy_ left
flussence (and I see a java *8* in these package repos now!) 17:32
17:33 jeff_s1 joined 17:35 kaare_ joined 17:40 dmol joined, btyler left 17:41 denis_boyun left, jnap left 17:44 crab2313 left
BenGoldberg rn: for (gather { take 1; take 2; say 3 }) { .say; last } 17:44
camelia niecza v24-88-g1f87209: OUTPUT«1␤»
..rakudo e9e99b: OUTPUT«3␤1␤»
BenGoldberg Good afternoon
masak BenGoldberg: o.O 17:45
where's... where's that 3 coming from?
BenGoldberg The say 3 17:46
masak right.
...oh :(
I bet it's that bloddy buffering thing again. 17:47
BenGoldberg r: gather {...}.infinite
camelia rakudo e9e99b: ( no output )
BenGoldberg r: gather {...}.infinite.say
camelia rakudo e9e99b: OUTPUT«Nil␤»
17:48 btyler joined
BenGoldberg I'm pretty sure it's because for() on rakudo reifys lists for which .infinite is not True 17:48
17:48 jnap joined
BenGoldberg And of course, Nil is not true 17:49
That bug is what led me to discovering the gather {...} but role {...} bug... since I wanted to try: 17:54
rn: for (gather { take 1; take 2; say 3 } but role { method infinite {True} } ) { .say; last } 17:55
camelia rakudo e9e99b: OUTPUT«This type does not support elems␤ in method REIFY at src/gen/CORE.setting:7531␤ in method reify at src/gen/CORE.setting:6757␤ in method reify at src/gen/CORE.setting:6744␤ in method reify at src/gen/CORE.setting:6849␤ in method reify at src/gen/CORE.setting:67…
..niecza v24-88-g1f87209: OUTPUT«3␤1 2␤»
BenGoldberg blinks 17:56
Is that a niecza bug? 17:57
17:58 rindolf left
BenGoldberg n: for (gather { take 1; take 2; say 3 }) { .say; last } 17:58
camelia niecza v24-88-g1f87209: OUTPUT«1␤»
BenGoldberg n: for (gather { take 1; take 2; say 3 } but role { method foo {True} }) { .say; last }
camelia niecza v24-88-g1f87209: OUTPUT«3␤1 2␤»
pmurias is changing nqp::getenvhash() supposed to change the env variables? 18:00
jnthn no
pmurias it does on parrot 18:01
donaldh pmurias: env is immutable on JVM 18:03
jnthn pmurias: OK, can it be *relied on* to do it? No 18:05
pmurias jnthn: on parrot it can be *relied on* 18:06
jnthn pmurias: You asked about the nqp:: op. 18:07
Which I take to mean "in general" :)
dalek ar: 6d48160 | moritz++ | modules/jsonrpc:
update to newer jsonrpc
 18:08
pmurias jnthn: I was asking on the "specced" behavior, not the de facto one ;)
18:09 grondilu left
diakopter de jure 18:09
du jour
18:10 SamuraiJack joined
jnthn That's just bribing me to write a spec doc for nqp:: ops :P 18:11
(yes, I know it needs doing :))
donaldh o O ( and more tests for nqp, not just test-by-rakudo ) 18:12
pmurias jnthn: bribing being "If you spec it I'll implement it correctly" ;)
jnthn ;)
donaldh realises thoughts like that are tantamount to volunteering
jnthn donaldh: Agree; pmurias++ has been adding more of late :)
18:13 lizmat joined
lizmat arrived at Siegburg 18:13
18:14 awwaiid joined 18:15 btyler left, vk joined
pmurias jnthn: is it possible to implement nqp ops as chunks of nqp code? 18:16
jnthn pmurias: See nqp::sprintf 18:18
pmurias which vm?
so I should hand compile my code to QAST? 18:19
18:19 btyler joined
pmurias figured out 18:20
jnthn pmurias: I think all VMs now...but certainly JVM
No, I'm sure sprintf was done in nqp
18:20 woolfy joined
pmurias I could do that or figure how to create and traverse a nqp hash in parrot C code 18:23
jnthn I'd do that :) 18:24
18:24 jnap left
jnthn (as in, not write more C code) 18:24
lizmat NMCC !
diakopter 17:54 < dalek> MoarVM: review: github.com/MoarVM/MoarVM/commit/59a7bb838d 18:25
17:57 < not_gerd> so, on to the next controversial subject: header re-organization
17:57 < yoleaux> 15:47Z <JimmyZ> not_gerd: please change the defaults to linenoise unless --use-readline ;)
17:57 < yoleaux> 15:54Z <JimmyZ> not_gerd: please see backlog
17:58 < not_gerd> [x] done
17:58 < not_gerd> I'm thinking something along these lines:
gist.github.com/gerdr/990ebfa6ef040b31b5bb
18:01 < jnthn> BenGoldberg: Probably when NQP runs nicely on it :)
18:02 < BenGoldberg> So, not until it self-hosts?
18:02 < BenGoldberg> Err, self-hosts NQP?
18:03 < diakopter> not_gerd: something will generate the public header file?
18:04 < not_gerd> diakopter: could be done, sure
18:05 < diakopter> just need to detect that marker and backslash-line-endings
18:05 < diakopter> (and follow includes, I guess)
18:05 < jnthn> BenGoldberg: Right
18:06 < jnthn> BenGoldberg: Well, self hosts NQP better than crashing before it even reaches the parser :)
18:06 < jnthn> Which is the current state of the self-host :)
aw f
jnthn fail 18:26
18:26 jnap joined
diakopter well, at least it wasn't something personal 18:27
lizmat NMWW !
jnthn yes, I wonder who he was pasting that to :P
diakopter no one 18:28
accidental selection (putty copy)
accidental right-click (putty paste, due to new computer and touchpad I'm not used to) 18:29
<-fail
18:29 spider-mario left
dalek p: 001cf1d | (Pawel Murias)++ | t/nqp/59-nqpop.t:
Test that nqp::istrue works on the result of nqp::null().
 18:35
p: 606da15 | (Pawel Murias)++ | / (3 files):
Partially implement nqp::shell() on parrot (ignores it's third argument).
masak diakopter: irssi has this nice "are you sure?" feature for big pastes. 18:36
diakopter masak: yes. I have it enabled.
sometimes it doesn't work
18:36 ajr joined 18:37 ajr is now known as Guest84619, Guest84619 is now known as ajr_ 18:42 benabik left, hummeleBop left
thou diakopter: congratulations on the new computer! :-) 18:42
diakopter thou: heh. 18:43
masak .oO( best of luck with the copy/pasting in the future! ) :P
diakopter easy fix (change putty to not paste on right-click) 18:44
18:47 denisboyun_ joined 18:49 sidus joined
masak FROGGS++ # closing a bunch of sprintf tikkits 18:50
18:50 woolfy left
FROGGS was a pleasure :o) 18:50
lizmat dinner& 18:51
18:51 lizmat left 18:53 donaldh left 18:57 donaldh joined
donaldh jnthn: are nqp test files just added with the next sequential number? 19:03
FROGGS donaldh: yes 19:04
19:08 Teratogen joined 19:09 Rotwang joined 19:14 awwaiid left 19:19 bluescreen100 joined 19:20 awwaiid joined 19:21 kaare_ left 19:22 bluescreen10 left
jnthn donaldh: yeah 19:27
donaldh: Unless you spot a gap, in which case fill it... :)
19:34 pernatiy joined 19:37 abnorman left
Mouq r: my $foo="+";&infix:($foo)(3,4).say 19:42
camelia rakudo e9e99b: OUTPUT«===SORRY!=== Error while compiling /tmp/hMrxNDLskK␤You can't adverb that␤at /tmp/hMrxNDLskK:1␤------> my $foo="+";&infix:($foo)⏏(3,4).say␤»
Mouq r: my $foo="+";&(infix:($foo))(3,4).say
camelia rakudo e9e99b: OUTPUT«===SORRY!===␤No such method 'has_compile_time_value' for invocant of type 'NQPMu'␤»
flussence (interesting thing I've noticed about that irssi paste detecting mechanism is that it seems to be entirely timer-based; I've had it questioning my actions just from typing fast over a slow ssh) 19:44
FROGGS r: BEGIN my $foo="+";&(infix:($foo))(3,4).say 19:47
camelia rakudo e9e99b: OUTPUT«===SORRY!===␤No such method 'has_compile_time_value' for invocant of type 'NQPMu'␤»
Mouq I think the &() should call it anyway. But it's bug ##68024 19:49
s/#//
19:50 Mouq left
BenGoldberg r: my $foo = "+"; say &(infix::($foo))(3,4); 19:52
camelia rakudo e9e99b: OUTPUT«===SORRY!=== Error while compiling /tmp/IbW58yTNBL␤Combination of indirect name lookup and call not (yet?) allowed␤at /tmp/IbW58yTNBL:1␤------> my $foo = "+"; say &(infix::($foo)⏏)(3,4);␤ expecting any of:␤ argument li…
BenGoldberg Which is less incorrect, infix:($foo) or infix::($foo)?
moritz r: say &::("infix:<+>").(3, 4) 19:53
camelia rakudo e9e99b: OUTPUT«7␤»
BenGoldberg Aah 19:54
19:57 abnorman joined 19:58 denisboyun_ left
moritz ok, R* is down to the one test failure in zavolaj 19:58
jnthn \o/ 19:59
moritz I'm going to fudge that one, and release.
nwc10 \o/
jnthn Looks like we'll make an R* this month then
moritz++
nwc10 doesn't R* usually happen a week after R?
moritz (unless somebody promises to fix the zavolaj failure in less than a week, then I'm going to wait for it)
jnthn moritz: Please (if there ain't one already) make a ticket for the failure
moritz nwc10: R* usually happens, if at all, between R* release and end of month
nwc10 ah OK
moritz jnthn: github.com/jnthn/zavolaj/issues/28 exists 20:00
[Coke]++
slava arnsholt: jnthn: I hope I am not being an annoyance, but I am running into some trouble with NativeCall. I started a project on github(github.com/theslava/git) to implement a perl6 interface to libgit2. I am completely lost as to how to represent a pointer to a pointer to a struct (passed in as an argument to git_repository_open). 20:02
jnthn slava: The typical trick for that is to use a one-element array 20:03
(CArray)
moritz CArray[CArray[YourStruct]]
slava jnthn: that's what I did ... but the other thing is that I could not find anywhere where they define the actual struct ...
jnthn I suspect just one level, given YourStruct is alredy a reference time...
slava: Are you just meant to treat it as an opaque pointer rather than looking into it?
slava: If so, use the OpaquePointer type 20:04
slava jnthn: I think so 20:07
github.com/libgit2/libgit2/blob/HE...h#L105-105 20:08
dalek volaj: 61224fa | moritz++ | t/02-simple-args.t:
skip failing test to make R* releasable
 20:09
slava that's there the git_repository is typedefed, I tried recursively searching for all structs and couldn't find anything ...
so repr('OpaquePointer') and then just CArray[OpaquePointer] ?
dalek ar: 876236e | moritz++ | modules/zavolaj:
use updated zavolaj with failing test skipped
 20:10
moritz slava: just use OpaquePointer as a type constraint in the function signature 20:12
20:13 denisboyun joined
moritz slava: if you look at github.com/perl6/DBIish/blob/maste...ish/Pg.pm6 for example, it uses OpaquePointers for all the structs that libpq uses 20:13
slava: and it never unpacks them, just passes them to other native functions 20:14
dalek ar: 7fbe884 | moritz++ | README:
bump version in README
 20:17
moritz jnthn: can you please add a .gitignore with perl6-debug* to rakudo::debugger?
jnthn: (or give me a commit bit, and I'll do it)
jnthn sorry, was away for a moment 20:18
moritz jnthn: thing is, 'git diff' in star is non-empty if there are untracked (and unignored) files in a submodule, which is kinda annoying
jnthn moritz: you got a commit bit
moritz jnthn: thanks
slava the fact that there is a variable name doesn't matter?
moritz slava: no, it doesn't 20:19
jnthn slava: OpaquePointer is kinda like a void*
slava I see, it also appears that git_repository it has to be allocated by the user
moritz that is annoying.
r: 'a' ~~ /(b)?/; say $0.perl 20:23
camelia rakudo e9e99b: OUTPUT«Any␤»
dalek kudo-debugger: af2f675 | moritz++ | .gitignore:
ignore generated files
 20:29
colomon Is rakuod-debugger working? 20:31
dalek ar: dcfcdaf | moritz++ | docs/announce/2013.08.md:
add announcement for 2013.08 release
colomon rakudo
moritz colomon: jnthn unbusted it yesterday night, iirc 20:32
colomon \o/
moritz moritz.faui2k3.org/tmp/rakudo-star-...RC0.tar.gz 20:35
20:35 lizmat joined
masak r: sub op($op) { &::("infix:<$op>") }; say op('+')(3, 4) 20:37
camelia rakudo e9e99b: OUTPUT«7␤»
masak \o/
dalek kudo-star-daily: 01b8bea | coke++ | log/ (5 files):
today (automated commit)
moritz please test and proof-read the release announcements
s/s$//
masak waitwait, is that a star I see shining? 20:38
moritz++
tadzik Rising!
moritz RCRising
timotimo coooooke!
you upset the configure script! :)
moritz on perl6.org/compilers/features Rakudo still has a +- on "Buf, buf and utf8" 20:40
do you think the support for Buf, Blob and utf8 is good enough now to give it a straight + ?
timotimo moritz: it says "Unary hyper ops currently descend into nested arrays and hashes. This will change to make them equivalent to a one-level map", but that's not accurate
actually, a "is nodal" trait will be applied to some methods and operators
jnthn moritz: The main missing thing is that Buf isn't mutable yet. 20:41
moritz timotimo: can you please fix it in the announcement? :-)
jnthn: I thought that's what Blob is for?
timotimo uh, i don't think i have access, but i can write a patch for you, or would you accept a replacement paragraph?
jnthn moritz: Blob is the immutable one
moritz oh
timotimo: I think I can give you commit access to star 20:42
slava also, is that auth() and ver() part of the standard and work? or is it still a wishlist type of item?
timotimo cool, thanks :)
jnthn whoa. Something is very hosed here. "git diff" doesn't work :/ 20:43
masak o.O
20:44 Rotwang left
tadzik you run it, but it makes no difference 20:45
bad one, I know
moritz so what? I'm inDIFFerent :-) 20:46
masak .oO( making puns on this one proved to be difficult ) 20:47
20:47 vk left
[Coke] (nqp:shell) is jvm only - there is already a shell in rakudo. could be refactored so the difference is hidden in nqp,aye. 20:48
jnthn Gah. Reboot... :/ 20:49
20:52 woolfy joined
moritz gah. Now I get t/02-simple-args.t (Wstat: 0 Tests: 11 Failed: 0) Parse errors: Tests out of sequence. Found (10) but expected (7) 20:53
jnthn Well. That didn't help... :S
20:55 lizmat left, lizmat joined
FROGGS O.o 20:56
jnthn and it's only for one repo that things don't work!
FROGGS well, clone afresh, and copy the files (except .git) over 20:57
jnthn aye
Oddly, I could commit...just not diff :S
wtf, it only happens when I have the submodule... 20:59
moritz would gladly abandon submodules immediately 21:01
jnthn heh
A git version update seems to have fixed it
but still...wtf
moritz git subtree is an interesting approach for something like R*
you just put the module in a subfolder under normal version control, and then do a subtree merge whenever you update a modulel 21:02
jnthn subtree may work out better for R*
masak +1 21:03
21:04 skids left
PerlJam +1 to getting rid of submodules :) Dunno if subtree merge will end up being better though 21:04
timotimo hm. i'm trying to come up with a description of nodality shorter than 4 lines :| 21:05
moritz timotimo: "the extend to which hyper operator recurse will be determined by properties of the methods involved" 21:06
timotimo i'd like to include the name "nodal" in it
moritz ... will be limited by nodal methods"
timotimo do you want to do the release now or tomorrow? 21:07
moritz plus maybe a pointer to the specs
[Coke] the rakudo-star build is busted.
moritz timotimo: not today
[Coke]: how so?
timotimo ok good
i'll go grab some food :)
[Coke] github.com/coke/rakudo-star-daily/...figure.log
looks like it happened in the last 24 hours. 21:08
moritz huh. That check is in there since 2012-08-28 21:09
and can be circumvented by callling Configure.pl --force
pmurias jnthn: should the qast_* tests from nqp-jvm be moved in into the standard nqp test suit? 21:10
[Coke] I've been doing builds with a fresh build every day for a month, first time this has happened.
jnthn pmurias: Given they shell out, I'm not sure... 21:12
pmurias: The point of them has been to exercise cross-compilers in development... 21:13
pmurias: Before they can actually handle having the code generator attached to the NQP frontend
masak 'najt, #perl6
pmurias jnthn: hmm, it could be possible to port them over to avoid shelling out when bootstraped 21:18
21:18 btyler left
pmurias so we don't pay the price of starting the JVM over and over 21:19
FROGGS jnthn: if nqp::getattr fails, because the requested attr doesnt exist, is there an op to check for existence? 21:20
jnthn FROGGS: no, you should know :)
moritz you can always introspect
jnthn FROGGS: I mean, you can always ask the MOP for a list of attributes
FROGGS moritz: in nqp?
moritz NQP has a MOP too 21:21
nqp: say(NQPMu.^methods)
camelia nqp: OUTPUT«Confused at line 2, near "say(NQPMu."␤current instr.: 'panic' pc 14738 (src/stage2/gen/NQPHLL.pir:5238) (src/stage2/gen/NQPHLL.nqp:279)␤»
pmurias jnthn: instead of QAST::Op.new(:op('say'),...) we would have an output(...), which could be turned into a say or something else for running in process
jnthn pmurias: Sure, if we want them in the NQP test suite that makes sense.
moritz nqp: say(NQPMu.HOW.methods(NQPMu))
camelia nqp: OUTPUT«10␤»
FROGGS cool
jnthn pmurias: They just were never written with the intent of becoming that, is all.
moritz nqp: say(nqp::join(' ', NQPMu.HOW.methods(NQPMu)))
camelia nqp: OUTPUT«get_string() not implemented in class 'NQPRoutine'␤current instr.: '' pc 61 ((file unknown):43) (/tmp/_BWRDBv4B1:1)␤»
jnthn pmurias: So they expect to have a main method run, etc. 21:22
FROGGS nqp: say(nqp::isstr($_) ?? $_ !! '') for NQPMu.HOW.methods(NQPMu)
camelia nqp: OUTPUT«␤␤␤␤␤␤␤␤␤␤»
jnthn pmurias: With the right munging they can end up in the NQP test suite though.
FROGGS: It's a list of method objects, not names. 21:23
FROGGS yeah, I recognized that :o)
21:23 woolfy left
moritz nqp: say(nqp::join(' ', map { $_.name }, NQPMu.HOW.methods(NQPMu))) 21:23
camelia nqp: OUTPUT«Confused at line 2, near "say(nqp::j"␤current instr.: 'panic' pc 14738 (src/stage2/gen/NQPHLL.pir:5238) (src/stage2/gen/NQPHLL.nqp:279)␤»
FROGGS nqp: say($_.name) for NQPMu.HOW.methods(NQPMu)
camelia nqp: OUTPUT«CREATE␤bless␤BUILDALL␤new␤Str␤Numeric␤Int␤defined␤ACCEPTS␤isa␤»
moritz nqp: say(nqp::join(' ', map({ $_.name }, NQPMu.HOW.methods(NQPMu))))
camelia nqp: OUTPUT«Could not find sub &map␤current instr.: '' pc 81 ((file unknown):48) (/tmp/eZo0OqZSlV:1)␤»
FROGGS nqp: say($_.name) for NQPMu.HOW.attributes(NQPMu) 21:24
camelia nqp: ( no output )
FROGGS nqp: say($_) for NQPMu.HOW.attributes(NQPMu)
camelia nqp: ( no output )
FROGGS ahh
nqp: grammar G { }; say($_) for G.HOW.attributes(G) 21:25
camelia nqp: OUTPUT«get_string() not implemented in class 'NQPAttribute'␤current instr.: '' pc 768 (src/stage2/gen/NQPCORE.setting.pir:355) (src/stage2/NQPCORE.setting:665)␤»
FROGGS nqp: grammar G { }; say($_.name) for G.HOW.attributes(G)
camelia nqp: OUTPUT«$!shared␤$!from␤$!pos␤$!match␤$!name␤$!bstack␤$!cstack␤$!regexsub␤$!restart␤»
FROGGS \o/
r: grammar G { }; say($_.name) for G.HOW.attributes(G)
camelia rakudo e9e99b: OUTPUT«$!ast␤$!shared␤$!from␤$!pos␤$!match␤$!name␤$!bstack␤$!cstack␤$!regexsub␤$!restart␤»
moritz gist.github.com/moritz/6324200 can anybody tell me what I'm doing wrong here? 21:26
donaldh hmm, the problem with writing tests is that you find bugs. 21:27
FROGGS moritz: the last one has no number, is that the problem?
donaldh r: my Mu $fh := nqp::open('CREDITS', 'r'); say nqp::readlinefh($fh); say nqp::tellfh($fh);
camelia rakudo e9e99b: OUTPUT«Unable to open filehandle from path 'CREDITS': No such file or directory(2)␤ in block at /tmp/mLqmroe5qJ:1␤␤»
donaldh that's not the bug :) 21:28
21:31 bruges joined
moritz FROGGS: maybe, but the error message says "Found (10) but expected (7)" which doesn't make any sense to me 21:32
FROGGS moritz: I don't understand it either
21:34 pmurias left
moritz the error doesn't go away when I add the number to the last test 21:35
FROGGS the strip tests from the end until it goes away
moritz oh funny
when I run prove with --verbose, the skipped test is indeed printed after no. 6 21:36
seems like a buffering/flushing issue :/
donaldh 'night #perl6 o/ 21:37
FROGGS newline missing?
moritz only happens when writing into the pipe that prove opens 21:38
oh, the tests have some fflush calls, but not everywhere it seems
21:39 donaldh left 21:41 btyler joined
dalek volaj: d2392e1 | moritz++ | t/02-simple-args.c:
guard tests against suffering from buffering
 21:41
ar: 8e50fcd | moritz++ | modules/zavolaj:
use newer zavolaj

should fix out-of-order tests		 21:42
moritz I'll do another RC tomorrow
lizmat moritz++ 21:44
moritz and bbkr++ for promtly fixing jsonrpc tests 21:47
and jnthn++ for fixing Rakudo::Debugger
it's not a one-man show here :-)
diakopter or woman! 21:48
moritz certainly isn't a one-woman show :-) 21:50
21:59 panchiniak joined 22:00 denisboyun left 22:02 abnorman left 22:06 ajr_ left
lizmat I would hope not :-) 22:06
22:06 kbaker left
lizmat I just pulled and spectested: one failure: 22:06
t/spec/S32-str/sprintf.rakudo.parrot (Wstat: 0 Tests: 135 Failed: 1)
Failed test: 132
not ok 132 - RT #116280# TODO sprintf printf numbers before NaN 22:07
this is on parakudo
FROGGS it is a todo O.o
22:09 fhelmberger joined, btyler left
lizmat not ok 132 - RT #116280# TODO sprintf printf numbers before NaN 22:11
# got: '9223372036854775808.00NaN'
# expected: ' NaN'
22:11 fhelmberger left, bluescreen100 left
lizmat I have no idea why this TODO test is not being ignored 22:11
diakopter twoo many # ?
lizmat yup 22:12
FROGGS lizmat: if you are going to fix it, an you fix the typo too?, it should be: sprintf prints numbers before NaN 22:13
can*
diakopter printf in early english
erm.
dalek ast: 6998cb4 | (Elizabeth Mattijsen)++ | S32-str/sprintf.t:
Remove confusion in fudger, so that TODO test is really a TODO
diakopter 1700s english
lizmat oops, saw that too late
FROGGS lizmat: thanks anyway :o) 22:14
dalek ast: f9ed594 | (Elizabeth Mattijsen)++ | S32-str/sprintf.t:
Fix typo, as per FROGGS++
 22:15
FROGGS lizmat++
22:22 xenoterracide_ is now known as xenoterracide 22:24 xenoterracide left, xenoterracide joined 22:25 btyler joined 22:32 sivoais joined 22:34 panchiniak left, denisboyun joined 22:35 btyler left 22:36 btyler joined 22:38 grondilu joined
flussence r: say [R<] 4,3,2,1 22:44
camelia rakudo e9e99b: OUTPUT«Not enough positional parameters passed; got 0 but expected 2␤ in block at src/gen/CORE.setting:15456␤ in sub at src/gen/CORE.setting:15584␤ in block at /tmp/ZJAWpJErOm:1␤␤»
flussence blah.
TimToady looks like a bug to me
r: say [>] 4,3,2,1 22:45
camelia rakudo e9e99b: OUTPUT«True␤»
TimToady True, it's a bug
n: say [R<] 4,3,2,1 22:46
camelia niecza v24-88-g1f87209: OUTPUT«True␤»
TimToady niecza++
in general, niecza still seems to handle composition of metaops a bit better than rakudo
diakopter what happens if you give that an inf list 22:47
TimToady it takes a long time
22:47 abnorman joined
diakopter the reversed one I mean 22:47
TimToady it doesn't reverse the list
diakopter ... 22:48
TimToady it still does it left to right
which is why I asked which "reverse" you was referrin' to
diakopter oh. I meant the [R thingamagic
TimToady yes, and I thought of at least 3 ways you might have meant that 22:49
jnthn is too tired to look into it righ tnow
Time for some rest...'night o/
FROGGS gnight jnthn
TimToady [<] -> any( [>], [R<], reverse [<] ) 22:50
o/
diakopter we need irc clients that channel-hop like securer radios channel-hop
TimToady er, [<] reverse
diakopter l
n: say [>] 1,2...* # y u no fastfail 22:53
camelia niecza v24-88-g1f87209: OUTPUT«(timeout)»
22:53 btyler left
diakopter r: say [>] 1,2...* # y u no fastfail 22:53
camelia rakudo e9e99b: OUTPUT«False␤»
diakopter r++ 22:54
TimToady that's a nieczabug from being a bit too pure about its HOP :)
why it can do [R<] though, since it's just composing higher-order functions 22:55
but it doesn't know to propagate short-circuitness
nr: say [&&] 0, 1 xx * 22:56
camelia rakudo e9e99b: OUTPUT«(timeout)»
..niecza v24-88-g1f87209: OUTPUT«===SORRY!===␤␤This macro cannot be used as a function at /tmp/Spk4OstQsX line 1 (EOF):␤------> say [&&] 0, 1 xx *⏏<EOL>␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1502 (die @…
TimToady and rakudo doesn't always do it either :)
we will probably need to work out the details of that someday... 22:57
grondilu r: say reduce {$^a && $^b}, 0, 1 ** * 23:05
camelia rakudo e9e99b: OUTPUT«0␤»
TimToady maybe we need to do it with a Y combinator... 23:16
lizmat r: use MONKEY_TYPING; augment class Any { method nth ($n=1) { my $seen=0; self.grep({ ++$seen %% $n }) } }; (1..100).nth(10).say 23:21
camelia rakudo e9e99b: OUTPUT«No such method 'nth' for invocant of type 'Range'␤ in block at /tmp/4PlkF0OUW0:1␤␤»
lizmat would expect this to work, apparently stuff in augmented classes does not inherit yet 23:22
fwiw, why don't we have a method nth() like that in Any ?
feels like a nice fit with the :nth modifier 23:23
23:24 btyler joined
diakopter just override postcircumfix [] :} 23:25
TimToady why not just use postcircumfix []? 23:26
lizmat this is returning *every* nth element from the list
how would that work with postcircumfix ? (/me is about to learn) 23:27
diakopter oh
heh
TimToady we both misunderstood, which shows how clear "nth" is :)
lizmat it was inspired by :nth 23:28
and I didn't think that one up :-)
TimToady :nth doesn't do that
say "thequickbrownfoxjumpsoverthelazydog" ~~ m:5th/./ 23:29
lizmat reads up on nth again
TimToady nr: say "thequickbrownfoxjumpsoverthelazydog" ~~ m:5th/./
camelia rakudo e9e99b, niecza v24-88-g1f87209: OUTPUT«｢u｣␤␤»
lizmat I guess my sub should be called "everynth" then 23:30
diakopter nr: say "thequickbrownfoxjumpsoverthelazydog" ~~ m:999999999999999999999999999999999999999999999999999999th/<:LETTER>/ 23:31
camelia rakudo e9e99b, niecza v24-88-g1f87209: OUTPUT«False␤»
lizmat :-)
ok, I've learnt something again :-) 23:32
grondilu use MONKEY_TYPING; augment class Any { method nth ($n=1) { my $seen=0;
lizmat I'll sleep on it
grondilu oop
lizmat gnight, #perl6, sleep&
TimToady ('aa'...'zz').pairs.map({ .value if .key %% 5 })[^10].say 23:33
23:33 _jaldhar joined
TimToady r: ('aa'...'zz').pairs.map({ .value if .key %% 5 })[^10].say 23:33
camelia rakudo e9e99b: OUTPUT«aa af ak ap au az be bj bo bt␤»
grondilu r: ('aa'...'zz').grep({ ++(state $) %% 5 })[^10].say 23:34
camelia rakudo e9e99b: OUTPUT«===SORRY!=== Error while compiling /tmp/B6pQTU0gFF␤Unsupported use of $) variable; in Perl 6 please use $*EGID␤at /tmp/B6pQTU0gFF:1␤------> ('aa'...'zz').grep({ ++(state $)⏏ %% 5 })[^10].say␤»
grondilu r: ('aa'...'zz').grep({ ++(state $ ) %% 5 })[^10].say 23:35
camelia rakudo e9e99b: OUTPUT«ae aj ao at ay bd bi bn bs bx␤»
grondilu r: ('aa'...'zz').grep({ (state $ )++ %% 5 })[^10].say
23:35 btyler left
camelia rakudo e9e99b: OUTPUT«aa af ak ap au az be bj bo bt␤» 23:35
TimToady r: ('aa'...'zz').pairs.map(-> $,$,$,$,$x { $x })[^10].say 23:36
camelia rakudo e9e99b: OUTPUT«4 => "ae" 9 => "aj" 14 => "ao" 19 => "at" 24 => "ay" 29 => "bd" 34 => "bi" 39 => "bn" 44 => "bs" 49 => "bx"␤»
TimToady r: ('aa'...'zz').map(-> $,$,$,$,$x { $x })[^10].say
camelia rakudo e9e99b: OUTPUT«ae aj ao at ay bd bi bn bs bx␤»
TimToady r: ('aa'...'zzzzz').map(-> $,$,$,$,$x { $x })[^10].say
camelia rakudo e9e99b: OUTPUT«ae aj ao at ay bd bi bn bs bx␤»
grondilu would have expected ('aa'...'zz').map(-> $,$,$,$,$x { $x })[^10].say to work :/ 23:37
oh
diakopter r: 'alive'.say; my \googolplex = 10**10**100; say(googolplex.WHAT);
camelia rakudo e9e99b: OUTPUT«alive␤(Num)␤»
grondilu ohr: ('aa'...'zz').map(-> $x,$,$,$,$ { $x })[^10].say
r: ('aa'...'zz').map(-> $x,$,$,$,$ { $x })[^10].say
camelia rakudo e9e99b: OUTPUT«aa af ak ap au az be bj bo bt␤»
TimToady r: ('aa'..'zzzzz' Zxx 0,0,0,0,1 xx *)[^10].say 23:43
camelia rakudo e9e99b: OUTPUT«ae af ag ah ai aj ak al am an␤»
23:44 mtk left
TimToady r: ('aa'..'zzzzz' Zxx (0,0,0,0,1) xx *)[^10].say 23:46
camelia rakudo e9e99b: OUTPUT«ae aj ao at ay bd bi bn bs bx␤»
TimToady r: (0,0,0,0,1 XRxx 'aa'..'zzzzz')[^10].say 23:47
camelia rakudo e9e99b: OUTPUT«(timeout)»
TimToady n: (0,0,0,0,1 XRxx 'aa'..'zzzzz')[^10].say
camelia niecza v24-88-g1f87209: OUTPUT« ␤»
23:47 lue joined 23:48 mtk joined
TimToady n: (0,0,0,0,1 ZRxx 'aa'..'zzzzz')[^10].say 23:48
camelia niecza v24-88-g1f87209: OUTPUT« ae (Any) (Any) (Any) (Any) (Any)␤»
TimToady say what? 23:49
oh, ^10
r: (0,0,0,0,1 ZRxx 'aa'..'zzzzz')[^10].say
camelia rakudo e9e99b: OUTPUT«ae␤»
TimToady niecza doesn't do ranges off the end of subscripts yet
TimToady is guessing 23:50
r: (0,0,0,0,1 XRxx 'aa'..'zz')[^10].say
camelia rakudo e9e99b: OUTPUT«aa ab ac ad ae af ag ah ai aj␤»
TimToady r: (0,0,0,0,1 XRxx ('aa'...'zz'))[^10].say
camelia rakudo e9e99b: OUTPUT«aa ab ac ad ae af ag ah ai aj␤» 23:51
23:51 _jaldhar left, _jaldhar joined
TimToady r: say 0 Rxx "aa' 23:51
camelia rakudo e9e99b: OUTPUT«===SORRY!=== Error while compiling /tmp/qJ_tnXh_6L␤Unable to parse expression in double quotes; couldn't find final '"'␤at /tmp/qJ_tnXh_6L:1␤------> say 0 Rxx "aa'⏏<EOL>␤ expecting any of:␤ postfix␤ infix stoppe…
TimToady r: say 0 Rxx "aa"
camelia rakudo e9e99b: OUTPUT«␤»
TimToady r: say 0 XRxx "aa" 23:52
camelia rakudo e9e99b: OUTPUT«␤»
TimToady r: say 0 XRxx "aa".."zz"
camelia rakudo e9e99b: OUTPUT«␤»
TimToady r: say 1 XRxx "aa".."zz"
camelia rakudo e9e99b: OUTPUT«aa ab ac ad ae af ag ah ai aj ak al am an ao ap aq ar as at au av aw ax ay az ba bb bc bd be bf bg bh bi bj bk bl bm bn bo bp bq br bs bt bu bv bw bx by bz ca cb cc cd ce cf cg ch ci cj ck cl cm cn co cp cq cr cs ct cu cv cw cx cy cz da db dc dd de df dg dh di dj d…
TimToady r: say 0,1 XRxx "aa".."zz"
camelia rakudo e9e99b: OUTPUT«aa ab ac ad ae af ag ah ai aj ak al am an ao ap aq ar as at au av aw ax ay az ba bb bc bd be bf bg bh bi bj bk bl bm bn bo bp bq br bs bt bu bv bw bx by bz ca cb cc cd ce cf cg ch ci cj ck cl cm cn co cp cq cr cs ct cu cv cw cx cy cz da db dc dd de df dg dh di dj d…
TimToady r: say "aa".."zz" Xxx 0,1 23:53
camelia rakudo e9e99b: OUTPUT«aa ab ac ad ae af ag ah ai aj ak al am an ao ap aq ar as at au av aw ax ay az ba bb bc bd be bf bg bh bi bj bk bl bm bn bo bp bq br bs bt bu bv bw bx by bz ca cb cc cd ce cf cg ch ci cj ck cl cm cn co cp cq cr cs ct cu cv cw cx cy cz da db dc dd de df dg dh di dj d…
TimToady r: say ("aa"..."zz") Xxx 0,1
camelia rakudo e9e99b: OUTPUT«aa ab ac ad ae af ag ah ai aj ak al am an ao ap aq ar as at au av aw ax ay az ba bb bc bd be bf bg bh bi bj bk bl bm bn bo bp bq br bs bt bu bv bw bx by bz ca cb cc cd ce cf cg ch ci cj ck cl cm cn co cp cq cr cs ct cu cv cw cx cy cz da db dc dd de df dg dh di dj d…
TimToady something is wacky there...
r: say 1..10 X* 0,1 23:54
camelia rakudo e9e99b: OUTPUT«0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0 10␤»
TimToady oh, right
nevermind, braino 23:55
grondilu r: say 1..10 Z* ((0,1) xx *) 23:56
camelia rakudo e9e99b: OUTPUT«0 2 0 4 0 6 0 8 0 10␤»
TimToady yes, that works better :)
grondilu r: say ("aa"..."zz") Zxx ((0,1) xx *)
23:56 _jaldhar left
camelia rakudo e9e99b: OUTPUT«ab ad af ah aj al an ap ar at av ax az bb bd bf bh bj bl bn bp br bt bv bx bz cb cd cf ch cj cl cn cp cr ct cv cx cz db dd df dh dj dl dn dp dr dt dv dx dz eb ed ef eh ej el en ep er et ev ex ez fb fd ff fh fj fl fn fp fr ft fv fx fz gb gd gf gh gj gl gn gp gr gt g… 23:56
TimToady r: say 'aa'..'zz' Zxx (0,1) xx * 23:57
camelia rakudo e9e99b: OUTPUT«ab ad af ah aj al an ap ar at av ax az bb bd bf bh bj bl bn bp br bt bv bx bz cb cd cf ch cj cl cn cp cr ct cv cx cz db dd df dh dj dl dn dp dr dt dv dx dz eb ed ef eh ej el en ep er et ev ex ez fb fd ff fh fj fl fn fp fr ft fv fx fz gb gd gf gh gj gl gn gp gr gt g…
TimToady don't need all those parens
grondilu indeed 23:58
TimToady r: say 'aa'..'zz' Zxx <0 1> xx *
camelia rakudo e9e99b: OUTPUT«ab ad af ah aj al an ap ar at av ax az bb bd bf bh bj bl bn bp br bt bv bx bz cb cd cf ch cj cl cn cp cr ct cv cx cz db dd df dh dj dl dn dp dr dt dv dx dz eb ed ef eh ej el en ep er et ev ex ez fb fd ff fh fj fl fn fp fr ft fv fx fz gb gd gf gh gj gl gn gp gr gt g…
TimToady don't need any of 'em :) 23:59
however, I'll bet it's numifying every single time
(this is a spot that p5's allomorphic caching beats p6)

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!