»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 moritz on 3 May 2013.
00:00 jerome__ joined 00:19 kurahaupo left
dalek ecs: 1dbea27 | lue++ | S02-bits.pod:
[S02] Fix small formatting code typo.
 00:24
00:29 REPLeffect left 00:57 PacoAir left 00:58 anuby joined 01:05 imIKARi left 01:06 berekuk left 01:08 berekuk joined 01:10 berekuk left 01:16 REPLeffect joined 01:29 snearch left 01:32 jerome__ left 01:44 jerome__ joined 01:49 FROGGS left 01:57 lizmat left 02:03 FROGGS joined 02:07 lustlife joined
pmichaud FROGGS: 404 Not Found 02:17
FROGGS: also, you shouldn't send out the email announcement until the files are actually available :-/ 02:18
that looks... bad. 02:19
pmichaud double-checks the release guide docs.
dalek kudo/nom: 8bf24a8 | pmichaud++ | docs/release_guide.pod:
Don't send out email announcements before tarballs are uploaded.
 02:21
02:23 lizmat joined 02:32 japhb_ joined
japhb_ Erm, Rakudo commit f7bfee6cd6259be7463b458a78a67efa639af12a looks rather wrong to me .... 02:33
Was Rakudo already released like that? 02:34
02:49 raiph joined 03:13 kurahaupo joined 03:15 rom1504 joined 03:33 hypolin left 03:37 fildon__ joined 03:40 hypolin joined
dalek rl6-bench: ead65e6 | (Geoffrey Broadwell)++ | / (2 files):
Factor compiler definitions out of timeall into compilers.pl
 03:40
03:40 fildon_ left 03:55 preflex_ joined 03:56 preflex left, preflex_ is now known as preflex
pmichaud japhb_: you're correct, f7bfee6cd is incorrect. 04:05
there hasn't been a tarball published yet, though, so.... I dunno.
I'm guessing we might need some qc for first-time-release-managers 04:06
although in this case FROGGS isn't a first-timer. So, I dunno.
04:12 kurahaupo left
dalek kudo/nom: 17b9fe4 | pmichaud++ | docs/release_guide.pod:
Release announcements now have .md extension.
 04:12
04:14 kurahaupo joined 04:20 IKARi__ joined 04:24 census joined 04:25 Psyche^ joined 04:29 Patterner left, Psyche^ is now known as Patterner 04:30 zby_home joined 04:34 adu left 04:36 spider-mario joined 05:03 kurahaupo left, spider-mario left 05:07 prevost left 05:08 dmol joined
FROGGS pmichaud: ó.ò sorry 05:10
these are the correct links: froggs.de/perl6/rakudo-2013.05.tar.gz froggs.de/perl6/nqp-2013.05.tar.gz 05:11
dalek kudo/nom: f97991c | (Tobias Leich)++ | VERSION:
[release] bump VERSION to 2013.05
 05:14
kudo/nom: 9743c6b | (Tobias Leich)++ | docs/release_guide.pod:
Merge branch 'nom' of github.com:rakudo/rakudo into nom
05:21 snoopy joined
FROGGS I've rebuild the tarball to have the right version fwiw 05:24
I'm really sorry that I've messed up the release, will start the release next time a few ours earlier to not get into midnight sleepiness again 05:26
and I'll double check what I'm going to commit, I believe I made the same mistake in november already
05:32 raiph left 05:40 FROGGS[mobile] joined 05:44 zby_home left, FROGGS left 05:45 tgt left 05:47 census left 05:49 kurahaupo joined 05:53 atroxaper joined, FROGGS joined
atroxaper Hello guys! 05:54
FROGGS hi atroxaper 05:55
atroxaper Does anybody know when new Rakudo Start will release?
FROGGS atroxaper: there will be a release when all modules pass, see github.com/coke/rakudo-star-daily/...odules.log 05:57
atroxaper: currently are three modules affected 05:58
06:00 FROGGS[mobile] left
atroxaper Ok. Thank you. 06:02
06:06 atroxaper left 06:08 SamuraiJack joined 06:09 dmol left 06:14 sqirrel joined
cognominal__ r: say('a' ~~ /a|aa/) 06:16
camelia rakudo 9743c6: OUTPUT«｢a｣␤␤»
cognominal__ I thought Perl 6 was using LTM. How is that compatible with that result?
06:24 arlinius left 06:26 lizmat left 06:28 SamuraiJack left
TimToady I don't see an 'aa' to match, so it falls back to 'a' 06:33
06:34 raiph joined
tadzik hmm. Stratopan is about having custom, CPAN-like repos in a cloud. Custom. CPAN-like. Something for panda, maybe 06:35
I seem to have a life-long account too 06:36
cognominal__ oops
indeed 06:37
r: say('aa' ~~ /a|aa/)
camelia rakudo 9743c6: OUTPUT«｢aa｣␤␤»
cognominal__ everything is find and dandy indeed.
and the term "falls back" is probably inappropriate for declarative matching. 06:38
r: say('aa' ~~ m:p5/a|aa/) 06:39
camelia rakudo 9743c6: OUTPUT«===SORRY!===␤Adverb p5 not allowed on m␤at /tmp/ppi6jWJRG8:1␤------> say('aa' ~~ m:p5/a|aa/⏏)␤»
cognominal__ r: say('aa' ~~ m:P5/a|aa/)
camelia rakudo 9743c6: OUTPUT«｢a｣␤␤»
cognominal__ P5 | is the P6 || :) 06:40
06:40 lizmat joined
cognominal__ good morning lizmat. 06:40
…and other #perl6 denizens 06:41
lizmat morning cognominal! 06:42
and #perl6!
cognominal__ learning that denizen comme from old french deinz 'within' which is probably the ancestor of the modern "dans"
indeed, both are coming from late latin deintus 06:44
hoelzro cognominal__: that's cool! 06:47
cognominal__ everything is cool in #perl6 except when I am too tired to think straight 06:49
but thinking straight maybe incompatible with lateral thinking, or diagonal, or fractal... 06:50
tadzik morning lizmat! 06:51
lizmat I usually only think out of the box
morning tadzik!
still only a 7 hour drive… :-)
with a few stops on the way
cognominal__ about fractals : imgur.com/gallery/lWkhr 06:52
06:52 woolfy joined
timotimo r: =item First␤=item Second␤=item Third;␤ .Str.say for $=pod; 06:52
camelia rakudo 9743c6: OUTPUT«Pod::Item<678915223>␤Pod::Item<-386447946>␤Pod::Item<-398726255>␤»
timotimo should the .Str method output a stringified version of those items instead?
there's not terribly much about this in the spec, i fear
also, good * people :) 06:55
06:55 lizmat left
timotimo also, i thought about using Konsole for a bit, but it seems like it has major trouble with that nl character >_< 06:55
06:55 lizmat joined
timotimo depending on wether the cursor is left of it or right of it, text seems to be too far to the left, so i can't see what i'm typing 06:56
FROGGS PerlJam++ # next rakudo star smoke run will have only two failing modules instead of three
timotimo oooh, what did you do? :)
FROGGS he fixed the tests for Grammar-Profiler-Simple 06:58
timotimo neat
timotimo builds the newest release 06:59
07:00 kurahaupo left
timotimo hmm. 07:01
FROGGS zavolaj (NativeCall) and perl6-lwp-simple still failing 07:02
timotimo can i just write a sub that takes an invocant and invoke that sub on some object without having to put it into the class explicitly?
tadzik yes 07:03
just do $foo.&bar instead of bar($foo) 07:04
bar needn't be special
07:04 kaleem joined
tadzik r: sub plusone($a) { $a + 1 }; say(plusone 5); say (5.&plusone) 07:04
camelia rakudo 9743c6: OUTPUT«6␤6␤»
timotimo r: enum Frob <FOO BAR>; sub not-special(Frob:D: $num) { say self; say $num }; FOO.&not-special(10); 07:05
camelia rakudo 9743c6: OUTPUT«===SORRY!===␤'self' used where no object is available␤at /tmp/H3eDyXwyGj:1␤------> R>; sub not-special(Frob:D: $num) { say ⏏self; say $num }; FOO.&not-special(10);␤ expecting any of:␤ argument list␤ prefix or te…
timotimo r: enum Frob <FOO BAR>; method not-special(Frob:D: $num) { say self; say $num }; FOO.&not-special(10);
camelia rakudo 9743c6: OUTPUT«Useless declaration of a has-scoped method in mainline␤===SORRY!===␤Undeclared name:␤ &not-special used at line 1␤␤»
timotimo not like that, obviously?
07:05 SamuraiJack joined 07:06 brrt joined
tadzik well, it's not really passed as an invocant then, I s'pose 07:06
it's a regular function call with a fancy syntax
timotimo mhm
well, i could always enhance the values of the enum with a role, no? 07:07
r: enum Frob <FOO BAR>; role Frobable { method frob() { return self +^ 42 }; FOO does Frobable; BAR does Frobable; BAR.frob.say;
camelia rakudo 9743c6: OUTPUT«===SORRY!===␤Unable to parse expression in block; couldn't find final '}'␤at /tmp/c4_e1qoy82:1␤------> obable; BAR does Frobable; BAR.frob.say;⏏<EOL>␤ expecting any of:␤ method arguments␤ statement end␤ s…
timotimo r: enum Frob <FOO BAR>; role Frobable { method frob() { return self +^ 42 } }; FOO does Frobable; BAR does Frobable; BAR.frob.say; 07:08
camelia rakudo 9743c6: OUTPUT«No such method 'mixin' for invocant of type 'Perl6::Metamodel::EnumHOW'␤ in sub infix:<does> at src/gen/CORE.setting:13509␤ in sub infix:<does> at src/gen/CORE.setting:13503␤ in block at /tmp/_Ymduy6jF9:1␤␤»
timotimo ah, not for those types i suppose
tadzik what are you trying to do? 07:09
timotimo just idly toying with language features that may or may not exist 07:13
then i'll have a look at why lwp-simple may be stumbling
07:17 REPLeffect left
timotimo www.software-path.com/was-cgi/json_...r_test.cgi - that's what's causing the lwp test failure 07:20
it's just a 500 error
probably expired domain?
moritz then either disable the test, or find some other URL to test against 07:21
lizmat rakudo.org ?
tadzik uh 07:22
lizmat have it send a request to an URL to rakudo.org ? 07:23
*at
timotimo it seems like it needs a cgi that will echo the command sent to it via a header
no, actually from the POST content 07:24
FROGGS timotimo: there are other problems too, already send a pull request to fix one
07:24 fhelmberger joined
lizmat that should not be too hard to make? anyways, just a suggestion 07:24
lizmat goes on the road…
see you all later, on irc or in meatspace :-) 07:25
FROGGS o/ :o)
woolfy is joining lizmat... :-)
07:25 lizmat left, woolfy left
tadzik oh 07:26
timotimo the other pull request might be good to apply, too
FROGGS uhh, moritz++
tadzik fun fact: if you write 'Build' which calls panda under the hood, you get a perl 6 module installable with cpanm
timotimo cosimo: do you have time to do it in the coming days?
tadzik: would you think a panda command to clone the repository for the module you specify to the CWD/repo-name would fit panda? 07:27
tadzik timotimo: I don't see why not
diakopter panda climb 07:28
tadzik but wait, isn't that what 'panda-fetch' is about?
timotimo not sure, i thought panda fetch fetches the code to a folder hidden away somewhere
dalek nda: b641a4d | tadzik++ | bin/panda:
Restore a fix which was lost during panda2 merge
tadzik not anymore
timotimo cool
tadzik I mean, 'panda-fetch', not 'panda fetch'
everything's a lot more stupid these days 07:29
timotimo ah, ok
tadzik but for what you say you'd want some kind of a wrapper for
fetch is apparently too stupid :)
so yeah, that'd make sense
timotimo tadzik: what's the difficulty in making p6prove, btw? 07:31
tadzik timotimo: nothing specific. Tuits, mostly, I think 07:33
timotimo care to point me at what's missing?
i'd like test runs to take longer, you know? :P
tadzik :P 07:34
timotimo: well, what needs to be done is: do a module smoketest, then make panda run p6prove instead of prove, then do a module smoketest again, compare differences 07:35
that'll give you new testcases for p6prove
timotimo where do i find the super-convenient "smoketest" command?
tadzik then you just need to fix bugs :0
github.com/tadzik/emmentaler here 07:36
there's a difftool which gives you a colourful summary of what has changed
so when it says "LWP::Simple now passes its tests", you'll know that p6prove is wrong about LWP::Simple
timotimo aaw :) 07:38
should i use the projectsfile from star?
07:39 arlinius joined
timotimo hm, no, apparently it should be the regular modules file? 07:39
tadzik I think the regular one, with all the things
feather.perl6.nl:3000/projects.json
timotimo ah, i was about to get the META.list instead 07:40
dalek nda: 9f6e824 | tadzik++ | / (2 files):
Fix panda-install
tadzik this is generated from META.list
the regular module smoketest is ran from this file 07:41
(the one from which feather.perl6.nl/~sergot/modules/ has the results)
I think gist.github.com/tadzik/5641922 is quite awesome :) 07:42
07:42 kresike joined
kresike hello all you happy perl6 people 07:42
tadzik you just need a fancy Build.PL (which can be a no-op), and Build (which just rans panda under the hood)
so now I can write 'panda cpanize' which does that for you 07:43
s/rans/runs/ 07:44
FROGGS cool! tadzik++
timotimo oh, neat
tadzik it should be possible to extract Perl 6 Pod from the files, convert it to Perl 5 Pod and store all in .pod
then metacpan and friends should be able to read it too
timotimo something's wrong with the smoker: No such method 'project-list' for invocant of type 'Any' - in line 40: for $panda.ecosystem.project-list -> $p { 07:45
tadzik /o\
I cannot reproduce that. Hrm 07:46
timotimo and indeed, dumping $panda gives me :ecosystem(▶Any)
this is panda2, btw
tadzik oh
panda2 is merged
you may want to use master again :) 07:47
merging fixed some bugs and introduced other
but it seems to work here
timotimo oh, ok
tadzik sorry for the mess 07:48
07:48 tgt joined
timotimo AFK for 30 07:49
tadzik heh. I can install panda with cpanm :D 07:58
08:00 konundra left
FROGGS O.o 08:00
dalek nda: 8bdbe7b | tadzik++ | bin/panda-cpanize:
Add panda-cpanize
 08:01
tadzik it doesn't solve any bootstraping problem, sadly
cpanm or not, it's still panda code doing the heavy lifting underneath
now I need to wait for stratopan to show up and put our entire ecosystem there 08:02
that's going to be fun :0
08:04 arlinius left
labster good morning 08:08
tadzik hi labster
FROGGS \o 08:09
08:10 berekuk joined
labster So I noticed this line in the specs for &open: A file opened with :bin may still be processed line-by-line, but IO will be in terms of Buf rather than Str types. 08:11
So, wait, we're supposed to do this .lines as strings, and then .encode to a Buf? And how are you supposed to figure out what whitespace is if it's binary? 08:12
*horizontal whitespace
timotimo i suppose it will rely on the tried-and-true cr and/or lf "standard" 08:17
08:17 TheAthlete joined
labster I think it makes more sense to still do stringy operations on a file opened as binary, but allow the compiler to assume a single-byte encoding. 08:18
Or maybe just get rid of the :bin switch altogether on open. If you want to do binary reading, use .read or .slurp(:bin). 08:19
jnthn morning, #perl6
brrt \o jnthn
08:20 sqirrel left
timotimo .o( method slur(:racial?, :ethnical?) { ... } ) 08:20
labster this isn't the yapc email list :P 08:21
tadzik :D
morning jnthn 08:22
jnthn tadzik: hey, guess where I'm flying to today? :)
tadzik \o/ 08:23
-o-
\o/
-o-
\o/
-o-
sorry, couldn't resist
jnthn You're making it look like hard work :P 08:24
tadzik I'm always excited when a plane takes off
someone pulls a lever, and a giant metal piece starts falling upwards
I'm always like "Yeah! Science, bitches!"
jnthn I'm always pretty relieved when it lands safely too :) 08:25
tadzik (no sexism and/or coc offense intended)
08:25 TheAthlete left
tadzik that too :) 08:25
There was sort of a Polish tradition of clapping hands when the pilot lands safely
cosimo timotimo: what should i do? 08:26
yoleaux 17 May 2013 01:41Z <[Coke]> cosimo: can you get perl6-lwp-simple passing its tests so we can cut a star release? looks like some sub signatures have changed.
cosimo oh :-)
FROGGS cosimo: is fixed already :o) 08:27
cosimo great then
tadzik jnthn: will you be at the pre-conf beer?
cosimo i'm a bit out of touch with p6 lately 08:28
have been
timotimo yeah, there's two pull requests waiting for you; also, one of the tests fails because a host is giving a 500 internal server error, maybe the domain it was on expired
jnthn tadzik: My flight lands at 16:55, so yeah, I should make it.
tadzik awesome :) 08:29
jnthn tadzik: Just looking at where I'm staying, where that is, how to get from airport, etc. :)
FROGGS cosimo: can you apply my three pull requests please? github.com/cosimo/perl6-lwp-simple/pulls
diakopter what's the pre-conf beer
tadzik jnthn: have you seen act.yapc.eu/plpw2013/wiki?node=TheH...eToWarsaw?
er, remove trailing ?
diakopter: a cheapskate version of "welcome dinner party" ;) 08:30
diakopter wait, which conf
tadzik Polish Perl Workshop
diakopter oh ok :)
'cuz I hadn't written down the pre-conf beer place Saturday the 1st 08:31
well I did in a side channel
tadzik :)
diakopter :D
08:32 raiph left 08:33 IKARi__ left 08:34 imIKARi joined, arlinius joined
timotimo tadzik: i don't understand how the smoker is supposed to work like that; it seems like the class Panda requires an Ecosystem instance to be passed, yet the only thing that the smoker passes is the projectsfile 08:36
tadzik hm 08:37
maybe I didnt push my fixes :o
oh, *facedesk*
sorry
08:37 berekuk left
dalek mentaler: 0d67c41 | tadzik++ | smoker:
Make it panda2-compatible
 08:37
tadzik timotimo: there, that should fi it 08:38
timotimo :) 08:40
timotimo smokes
cosimo FROGGS: done! 08:42
timotimo er ... 08:43
cosimo please don't assume I can read everything GH throws at me, a mail would be great if I'm blocking you
timotimo i wonder how long that'll take
probably more ignoring than blocking, no? 08:44
FROGGS cosimo: 'am currently testing... thanks!
timotimo i feel mildly uneasy about just running code from github like that :|
cosimo timotimo: ruby folks do it all the time, must be cool ;-) 08:45
timotimo heh.
FROGGS perl 5 folks too
08:45 berekuk joined
FROGGS but having a moderated app-store like 6pan is crap too 08:46
timotimo removes projects from github users he doesn't recognize
FROGGS *but having a moderated app-store as 6pan is crap too
okay, only zavolaj left with failling tests 08:49
timotimo: it took like 6 hours at my (slow) server at home 08:51
hmmm, zavolaj doesnt fail if I run it :/
timotimo yikes 08:53
i should fire up my somewhat mighty desktop
maybe i could do something to parallelize test runs in smoker?
tadzik: is it dangerous to run multiple panda installs at the same time? 08:54
FROGGS ahh, [Coke]'s failing rakudo perl6 binary is 2013.04, a working one is 2013.04-97-gf7bfee6
[Coke]: after upgrading to rakudo 2013.05 all modules should be fine 08:55
08:56 domidumont joined
tadzik timotimo: at this point it may be 08:56
because they all write to '.work/'
I should make it '.work.PID' or something, that'd be safe 08:57
FROGGS [Coke]: btw, the S02-types/bool.t failure does not happen using 2013.04!
I guess bisecting 97 commits is rather pain on feather -.- 08:58
timotimo i'm considering adding a -j or SMOKER_JOBS parameter that would flatten the whole tree (dependency resolution and all that) and collect the outputs into an ordered kind of way or something
so that each run will give the same output ordering for correct diffs 08:59
08:59 fgomez left
FROGGS nice 09:00
timotimo tadzik: are you going to add the .PID right now? 09:01
09:01 bbkr joined
FROGGS my not have each module in its own TAP file? so a file is only changed if its tests change somehow 09:01
timotimo hm, that would be sensible, too. 09:02
09:07 xinming left 09:12 rindolf joined 09:14 atroxaper joined, atroxaper left
timotimo i have to ./smoker > output, don't i? 09:14
FROGGS I think so 09:15
or wait, is there a result.json?
but to get the TAP you need to capture stdout 09:16
timotimo right.
tadzik timotimo: not right now, this evening, maybe 09:20
09:27 berekuk left 09:28 berekuk joined 09:29 araujo joined, araujo left, araujo joined 09:36 imIKARi left 09:39 xinming joined 09:49 SamuraiJack left 09:50 SamuraiJack joined 09:51 berekuk left 09:57 sqirrel joined
FROGGS arrgghh!! Unsupported serialization format version 5 (current version is 4) 10:04
lunch &
10:13 mrlo joined 10:22 SamuraiJack left, raiph joined 10:25 jest joined
timotimo tadzik: i'm not sure why, but the $statefile doesn't properly get passed to the ecosystem's BUILD method :| 10:27
what's keeping the tarballs from being released? does it still need testing? 10:31
10:33 anuby left 10:35 brrt left
tadzik timotimo: you're talking about smoker? 10:41
timotimo yes 10:42
10:42 rindolf left
tadzik hm 10:43
that's bad :(
do you have a testcase? 10:44
10:46 raiph left
timotimo yes, hold on 10:47
use Panda; my $projectsfile = "projects.json"; my $e = Panda::Ecosystem.new( statefile => "{cwd}/state", projectsfile => $projectsfile,); my $panda = Panda.new( ecosystem => $e, installer => Panda::Installer.new(destdir => "{cwd}/installlib")); 10:48
tadzik looks similar to smoker :)
I have quite a busy day, if you could open a bug for me, I'll probably manage to fix it tomorrow 10:49
masak oh hai, #perl6
tadzik: guess where I too am going today? :D
unlike jnthn though, I need to stay behind a bit, and take the next flight. so I'll miss any pre-conf stuff. 10:50
masak is still in teaching mode
gonna go home in a few hours and write my presentations :P
FROGGS masak: can you upload this to rakudo.org? froggs.de/perl6/rakudo-2013.05.tar.gz froggs.de/perl6/nqp-2013.05.tar.gz 10:51
masak FROGGS: 'fraid not. lunch right now, but I can't swap in non-$dayjob tasks :/
FROGGS k
masak someone else is bound to be able to help, though. 10:52
FROGGS does somebody know what that means when trying to build nqp? Unsupported serialization format version 5 (current version is 4) 10:54
tadzik masak: \o/ 10:55
masak: guess what arrived with the post today :)
FROGGS: clean ALL THE THINGS maybe
masak tadzik: nice! happy reading! 10:56
FROGGS tadzik: k, will do 10:57
masak away for a few more hours
11:04 brrt joined
FROGGS tadzik: even after removing and cloning nqp again, I cant build 2013.04 :/ 11:06
tadzik FROGGS: hm :|
I have no clue
FROGGS maybe parrot was configured using other flags... 11:07
11:11 mtk left
jnthn Poland & 11:13
FROGGS no, even rebuilding parrot doesnt work
11:15 mtk joined, PacoAir joined 11:18 jest left
FROGGS looks like it didnt rebuild parrot even when passing -gen-parrot and removing its folder 11:25
Ulti has the release schedule for R* changed to a longer period? or is everything on hiatus for JVM stuff?
FROGGS I removed the install folder now
Ulti: there were failing modules
Ulti ah right
FROGGS I am currently trying to hunt down the last bug 11:26
Ulti cool
timotimo so, r* may even come this weekend?
FROGGS well, first we have to fix that bug and find someone who can make a release (and has tuits) 11:28
moritz back 11:29
FROGGS moritz: can you upload this to rakudo.org? froggs.de/perl6/rakudo-2013.05.tar.gz froggs.de/perl6/nqp-2013.05.tar.gz 11:31
moritz seems I can :-) 11:33
FROGGS \o/
moritz done.
bbkr is testing release 11:34
FROGGS bbkr++ 11:35
cool, looks like my nqp-2013.04 got compiled
12:01 rindolf joined 12:02 dayangkun joined
timotimo hm. when i want to override only a subset of the candidates of a multi method in a class i derive from, could i get at those other candidates within a proto i write for the derived class? 12:05
huh, i just noticed in this advent post about LTM, there's a rule { +? } which i'm rather sure should be .+? instead or something similar 12:06
and the rule text is empty?
oooh, angle brackets perhaps?
tadzik: huh, with current panda i can't install rakudo-debug; it will just copy the rakudo-debug.nqp file to my bin folder?! 12:14
tadzik timotimo: uh-oh, I must've broken something :( 12:15
timotimo no, happens with ufo, too
tadzik ufo was never capable of running Build.pm I think
timotimo oh, right 12:16
tadzik I'll try to fix it this weekend, if you could open me a bug
timotimo since it's blocking r* (or should block r*), i'll try to fix it myself first
yeah, there's a FIXME around the Build.pm part 12:17
tadzik ah, there we go 12:18
thank you
12:20 dayangkun left
pmichaud (R*) if someone will tell me when modules are no longer failing, I can find tuits to cut the R* release 12:21
moritz afaict timotimo++ just fixed LWP::Simple 12:23
timotimo timotimo-- didn't, FROGGS++ did
moritz oh right
your nicks both contain an 'o', no wonder I keep confusing you :-)
timotimo if R* is made, maybe revert to panda1
tadzik is it made now? 12:24
oh
timotimo probably not immediately
tadzik good
I'll try to fix everything during plpw
or at least during yapcna
timotimo so, Resources got axed from panda2; Rakudo::Debugger sets my $workdir = $.resources.workdir, but doesn't use the variable at all 12:25
tadzik yeah, I killed Resources with firne
and for a reason :0 12:26
timotimo it didn't seem very useful, maybe it was supposed to grow past its form at that time?
i wonder how many Build.pm there are in R*
tadzik I think there's one, and not just in R* :)
timotimo but i suppose i can find out with Task::Star 12:27
tadzik yep
new build($x) should probably just chdir to $x (justin case) and do the same thing that it does now
timotimo ah, so no more Pies::Project argument 12:28
tadzik yep
timotimo should it be a Str or an IO? 12:29
tadzik er, I think Str
timotimo ok
tadzik like in github.com/tadzik/panda/blob/maste...der.pm#L54
I think the topmost 'return' there is wrong too 12:30
arnsholt pmichaud: I hope to get Zavolaj fixed in Austin, but I'm not terribly optimistic about getting it fixed before that, sadly :/
timotimo tadzik: is .IO.d not preferable to .IO ~~ :d? 12:32
tadzik timotimo: timtowtdi, I guess 12:33
timotimo mhm
tadzik the former seems cleaner to me
timotimo but a few lines down you use .IO.f ;)
tadzik hehe
timotimo what's wrong about the return? should it be moved down after looking fora Build.pm?
tadzik I think it can be removed altogether 12:34
timotimo u r boss! ;)
tadzik :)
module can have a custom-built bin/foo, but no lib
it just never happened before
I'm terrified about being the boss. So many things break because of me
timotimo as long as we can backpaddle a few commits and things work again ... ;) 12:35
should i add a chdir after the custom Build is called, so that the custom Build can't wreak havoc? 12:37
moritz yes, sounds sane
12:39 kaleem left
dalek rl6-roast-data: 15eafd8 | coke++ | / (4 files):
today (automated commit)
 12:44
timotimo i'm having trouble requiring the Build file 12:45
[Coke] ah, hell, looks like the data collection bit at the end misfired there.
also looks like rakudo started failing a bunch of tests yesterday. 12:46
timotimo require 'Build' will give "Could not find file 'Build' for module Build", require "Build.pm" will give a huge stacktrace resulting apparently from not being able to stringify the exception that flies
[Coke] (all in S32-io/IO-Socket-INET.t)
FROGGS [Coke]: rakudo was clean on my box 12:47
[Coke]: btw there are some notes in the scrollback
[Coke] FROGGS: the test was run with rakudo 25bfa5f and roast 298edbd 12:48
If you want to bump the version rakudo * is building with before the automated run this afternoon, perhaps that will help the star module run. 12:49
dalek kudo-star-daily: 34b4847 | coke++ | log/ (5 files):
today (automated commit)
[Coke] no change in star test module runs. 12:50
(those are both from yesterday's run, not pushed until now)
dalek ar: c79a520 | pmichaud++ | modules/ (10 files):
Pull latest modules for a possible Rakudo Star 2013.05 release.
12:51 bluescreen10 joined
tadzik timotimo: (re chdir) you may want to use indir 12:51
it's in Panda::Common or somewhere
it's pretty polished
timotimo there's something else problematic 12:52
it seems like $where is a relative directory
how do i turn it into an absolute?
tadzik uhhh
fix Panda.pm, I suppose :/
[Coke] pmichaud++
timotimo Failed to change the working directory to '.work/1369399823_0': chdir failed: No such file or directory <- this is what happens if i chdir $where in my Build.pm/method build
tadzik timotimo: or don't touch in an actual Build.m
it worked without it before
pmichaud hmmm... Rakudo's release announcement doesn't note the change in ?-semantics.
FROGGS [Coke]: I have to update the modules versions too for a clean star, right? 12:54
pmichaud: true, that is in the changelog only 12:55
timotimo perl6-debug seems to install properly now 12:57
tadzik yeah! timotimo++ 12:58
timotimo i'll have a pull request for you shortly
dalek ar: fc649f7 | pmichaud++ | docs/announce/2013.0 (2 files):
Rename 2013.04.md announcement to 2013.05.md.
 12:59
ar: d1cb981 | pmichaud++ | / (2 files):
Some updates for a 2013.05 release.
ar: 9fabc39 | pmichaud++ | / (2 files):
Update star release number in README/Makefile.in .
 13:00
[Coke] FROGGS: pmichaud just did that.
[Coke] will try to track down the roast-spec failure from yesterday (the meta-failure, not hte rakudo socket failures) and maybe rewrite history. 13:01
dalek nda: 6b2cc0d | (Timo Paulssen)++ | lib/Panda/Builder.pm:
make custom Build.pm files work again.
nda: 23ff9b7 | tadzik++ | lib/Panda/Builder.pm:
Merge pull request #46 from timo/master

make custom Build.pm work again
tadzik I have to admit, I didn't even test that :) 13:02
timotimo hm. i have a slight feeling that rakudo::debugger has suffered from a bit of bitrot
jnthn: you'll find a pull request for Rakudo::Debugger, too 13:04
it no longer seems to show what part of text was matched so far by regexes 13:05
dalek kudo/nom: 45ae2d2 | pmichaud++ | docs/release_guide.pod:
Bump copyright year as suggested by bbkr++.

Closes #159.		 13:06
13:08 fhelmberger left
bbkr FROGGS: 2013.05 release works fine on OS X 10.6 and 10.7 13:10
brrt you know what would be awesome? json for actual graphs
13:13 jaldhar left, ajr joined, ajr is now known as Guest73822
timotimo where do i get the current cursor from again? $/.CURSOR? 13:13
13:14 Guest73822 is now known as ajr_
timotimo i think something's wrong about the way rakudo::debugger tries to get the cursor object 13:14
my $cur = eval_in_ctx($ctx, q[DYNAMIC::<$¢>]); 13:15
13:15 cognominal__ left
pmichaud current cursor is in $¢ 13:16
13:16 cognominal__ joined
timotimo This type does not support positional operations is the error that line is causing 13:16
pmichaud $/.CURSOR will get there, but builds a Match object in the process.
timotimo ^H^H^H^H^H^H^H^H^H^Hok
timotimo tries with "..." instead of q[...]
13:17 Pleiades` left
timotimo i don't quite understand how i get that error :| 13:18
the same with eval_in_ctx($ctx, '1 + 1'), so maybe the error is in eval_in_ctx 13:19
ah, that's probably it, yes. 13:20
it's a ForeignCode object, that's probably what broke it. recently ForeignCode was introduced, right? 13:21
i don't really know how to get at an object inside that to set the outer context :| 13:24
pmichaud timotimo: I don't know either
timotimo well, just removing the line with set_outer_ctx seems to fix it
13:24 Pleiades` joined
timotimo maybe it's now enough that the outer context was given as :outer_ctx in the $compiler.compile call 13:25
interesting! if the last line doesn't end in a semicolon, it won't be highlighted 13:31
but that's just a minor papercut 13:33
13:37 SmokeMachine joined
timotimo feels useful for a change :3 13:40
i should create a few "slides"
Ulti i.minus.com/iH8aOpXstddxg.gif 13:45
^ not perl6 related
timotimo wow, she is not enjoying that
i imagine i would be amused at least
FROGGS bbkr: cool! good to know 13:46
Ulti I laughed out loud at the top of my voice in the office :/ people came to see what was wrong
timotimo heh.
diakopter the audio isn't working for me 13:47
Ulti the audio on the .gif?
timotimo so, i'm trying to write code that runs if Text::Levenshtein exists, but doesn't explode at compile-time if it doesn't. "require Text::Levenshtein <distance>" doesn't seem to do what i want: Trying to import from 'Text::Levenshtein', but the following symbols are missing: distance
but there's a sub "distance" in there that is marked "is export"
Ulti instead of Levenshtein implement Bitap :] en.wikipedia.org/wiki/Bitap_algorithm 13:48
timotimo levenshtein is already implemented. i'd just have to use it 13:49
but not with "Use"
Error while compiling block (source text: "require Text::Levenshtein <distance>;\nsay distance(\"foo\", \"boo\");\n"): Error while compiling op call: Error while compiling block : Error while compiling op call (source text: "say distance(\"foo\", \"boo\")"): Error while compiling op call (source text: "\"foo\", \"boo\""): Null PMC access in find_method('handle') 13:51
that's better! ... not :(
FROGGS arnsholt: ping 13:52
13:54 bruges left, bruges joined
timotimo tadzik: do you think ecosystem.suggest-project is a good place to put a suggestion generating method for when a project name has been misspelt? 13:58
FROGGS what about ecosystem-typo-projectname? 14:00
suggest-project sounds like advertisement to me
PerlJam good $localtime #perl6 people ! 14:01
brrt hi PerlJam
timotimo hehehe
PerlJam What's new in the world?
brrt nothing, ever
we decay by the second law 14:02
but that is not what you meant, is it? :-)
PerlJam brrt: I dunno what I meant because whoever answers is meant to supply the missing context ;)
brrt good point 14:03
timotimo it's like when an interviewer keeps holding the microphone to someones face after they stopped talking
Ulti brrt maybe if you are a physicist and not a biologist 14:04
brrt huh, i actually study biology ;-) 14:05
Ulti life isn't really heading towards maximum entropy even after a billion years
timotimo hm, there's bioperl, but is there physiperl?
PerlJam timotimo: In p5-land that's called PDL ;)
Ulti PDL is physiperl I guess 14:06
brrt ulti, and why is that, really?
Ulti more astro/image analysis perl
brrt its true
but its weird isn't it?
timotimo is that like numpy for python?
PerlJam really PDL is kind of like mathperl (which we all know is the basis for physiperl (which, in turn, is the basis for bioperl)) 14:07
:-)
timotimo you forgot the chemiperl in the middle 14:08
Project Text::Levenshtein not found in the ecosystem. Maybe you meant Text-Levenshtein?
Ulti brrt the best sorts of intersections between life and thermodynamics are papers like these www.ncbi.nlm.nih.gov/pubmed/16408074
timotimo (a little bit of an autopun?)
tadzik: do you like that? :)
brrt that is bizarre… thanks for that 14:09
Ulti entropy motors are p. neat this is why I love molecular biology evolution at this level is just hacking reality 14:10
PerlJam Ulti: all of what we do is "hacking reality" 14:11
we imagine a future, then hack reality to match our imagination.
Ulti sure but this is a blind natural process 14:12
like imagine if the ocean just started doing something weird like teleporting around, its that level of wtf some of what happens in your cells
14:12 PacoAir left
Ulti that kinesin motor I just linked to, is usually paired with another protein that is biased to "walk" in the opposite direction so it switches footing when it bumps into something in its way 14:13
14:13 dayangkun joined 14:14 dayangkun left
PerlJam And Conway's Game of Life comes full circle ;) 14:14
Ulti :) yeah basically gliders on a lattice :) 14:15
they even look a bit like them too
www.youtube.com/watch?v=y-uuk4Pr2i8 14:16
timotimo i am bummed out that no new video about that stuff has been made since that one came out 14:17
Ulti yeah
timotimo: have you seen this book www.amazon.com/The-Machinery-Life-D...0387849246 might be of interest 14:18
14:19 xinming left 14:20 PacoAir joined 14:21 xinming joined 14:22 PacoAir left, PacoAir joined 14:23 PacoAir left, PacoAir joined
tadzik timotimo: sounds good :) 14:27
timotimo: there was a panda ticket with some discussion about this, have you seen it?
the implementation is nice and elegant too 14:29
timotimo i read it now
the implementation i made?
tadzik yep 14:32
timotimo thank you :D
PerlJam Does panda use Levenshtein edit distance to offer helpful advice on possible module names when it can't find what the user has asked for?
tadzik not yet
but see timotimo++'s pull request
timotimo no, it doesn't, because i couldn't get Text::Levenshtein in as a runtime-only-dependency
PerlJam oh! 14:33
Nice.
(I was just asking because I typed "panda install Text::Levenstein" and was surprised that it couldn't be found :)
FROGGS hehe
timotimo 16:07 < timotimo> Project Text::Levenshtein not found in the ecosystem. Maybe you meant Text-Levenshtein? 14:34
;)
PerlJam okay, even when I type it correctly, I get "Project Text::Levenshtein not found in the ecosystem" What gives? 14:35
timotimo "panda update"?
PerlJam no dice. 14:36
FROGGS thinks he has found the core.pm of all evil
actually src/core/terms.pm 14:37
timotimo er. that's not how you type correctly, PerlJam
Maybe you meant Text-Levenshtein?
PerlJam Hrm. that's icky. 14:38
FROGGS it is
PerlJam panda install Text::CSV worked fine.
What does it use to determine the module name?
FROGGS the value in META.info
PerlJam I think I feel itchy 14:39
maybe panda could try Module::Name and then Module-Name ? 14:40
14:40 Guest73688 joined
timotimo don't think so 14:40
see the discussion in the ticket about case insensitivity 14:41
PerlJam reads
timotimo++ btw
14:41 Guest73688 left
timotimo :3 14:42
kresike bye folks 14:45
14:45 kresike left 14:46 kaleem joined 14:51 spider-mario joined 14:55 sjn left 14:57 sjn joined 14:58 domidumont left
FROGGS [Coke]: btw, in rakudo/src/core/terms.pm, if you comment out line 8 and line 117, S02-bool.t passes 14:59
will bisect zavolaj this evening more
bbl
14:59 FROGGS left 15:00 PacoAir left 15:01 PacoAir joined 15:04 FROGGS[mobile] joined
timotimo i wonder if the many object systems for perl5 will work atop rakudo 15:05
FROGGS[mobile] we will find out 15:08
brrt off 15:12
15:12 brrt left 15:13 kaleem left 15:22 IKARi__ joined 15:23 risou_awy is now known as risou
timotimo i wonder if it would make sense to introduce a way to say $a is always exactly a Dog, never has roles applied on it or contains a subclass of Dog 15:25
but that's probably too specific
PerlJam timotimo: you mean as a constraint or a compiler hint? 15:26
timotimo constraint would be sanest 15:28
i wouldn't like to hint at the compiler "this is always a Dog" and then accidentally put a subclass of Dog in there 15:29
pmichaud maybe a "is <mumble>" trait
timotimo is exactly[Dog] 15:30
pmichaud my Dog $a is pure;
timotimo maybe "purely"
15:31 Chillance joined
timotimo is "methods on Cool exist as sub, too" a rule without exceptions? 15:31
pmichaud I don't think so.
some methods can't easily be subbed 15:32
timotimo mhm
pmichaud grep is an example. 15:33
(although grep is really defined on Any)
timotimo how incomplete is :P5 support for regexes in rakudo? 15:34
pmichaud "30"
15:35 alester left
PerlJam Where's the regex feature matrix for :P5 ? :) 15:35
15:35 rindolf left
pmichaud r: say 101.base(16) 15:36
camelia rakudo 45ae2d: OUTPUT«65␤»
pmichaud r: say base(101, 16); 15:37
camelia rakudo 45ae2d: OUTPUT«===SORRY!===␤Undeclared routine:␤ base used at line 1␤␤»
pmichaud there's an example of a method in Cool that isn't a sub.
15:46 IKARi__ left
timotimo right. that 30 was to what base? :) 15:46
i feel like i need to take a nap :/ 15:47
my slides are not progressing
colomon does p5 have a FIRST phaser? 16:12
16:15 zby_home joined
colomon worked around it... 16:16
16:16 raiph joined 16:17 REPLeffect joined
masak oh hai, #perl6 16:32
FROGGS[mobile] o/ 16:35
16:39 FROGGS[mobile] left, FROGGS[mobile] joined 16:49 bluescreen10 left, fgomez joined 16:51 domidumont joined 16:55 apejens joined 16:59 ajr_ left 17:05 FROGGS joined 17:11 Timbus left 17:14 Timbus joined 17:16 risou is now known as risou_awy, risou_awy is now known as risou 17:32 prevost joined 17:34 risou is now known as risou_awy 17:36 ajr joined, ajr is now known as Guest99846 17:37 Guest99846 is now known as ajr_ 17:49 imIKARi joined 17:54 linguini joined 17:55 raiph left 17:58 berekuk joined 18:10 kaare_ joined 18:16 japhb_ left, kaleem joined 18:17 kaleem left, saxx11 joined
PerlJam #perl6 is too quiet today. 18:20
colomon It's because we're scheming.
18:21 dukelet0 is now known as dukeleto
geekosaur shouldn't you be perl6ing? 18:21
18:21 dukeleto left, dukeleto joined
colomon has been working on 71.10.146.50:3001/report # now with sorting and much better individual project reports 18:22
geekosaur: technically speaking I've been perl5ing.
PerlJam How do people get nominated for the White Camel? jnthn and pmichaud deserve nominations.
18:23 SamuraiJack joined
PerlJam (And really, TimToady deserves some kind of perseverence award or something too :-) 18:24
dalek osystem: 93b0b25 | colomon++ | META.list:
Transfer RFC back to ihrd

Module is fixed now.
colomon Should have said URI instead of RFC. 18:25
sorear HIII 18:29
sorear bangs pans
PerlJam colomon: That sounds like META.list contains the One True Version of modules 18:30
colomon PerlJam: well, it contains the versions that panda looks at. That's pretty close to the One True Version. 18:31
18:32 zby_home left 18:37 bluescreen10 joined
colomon can see someone is checking out Test::Corpus... 18:39
flussence waves
colomon \o
18:40 domidumont left
flussence aw nuts, my modules are all acting up... 18:40
colomon moritz++ # the core of his code for analyzing the JSON smoke logs is still in this Dancer project 18:41
flussence ok, that was easy enough. Test-Corpus and Text-Tabs-Wrap should be fixed now 18:45
flussence &
colomon \o/
flussence
.oO( well, apart from that one TTW file that never did work properly... )
18:46 dagurval joined
flussence gah, TTW still needs work. I'll have a look later... 18:47
18:52 bluescreen100 joined 18:54 dukeleto is now known as parrot_dukeleto 18:58 dmol joined
[Coke] in the middle of a star build, will let you know if it resolves any test failures. 19:06
colomon is thinking calling panda 175 times in a row might not be the fastest way to render a web page. :\
[Coke] colomon: 71.10.146.50:3001/report isn't responding here.
colomon [Coke]: may have something to do with the script trying to call panda 175 times in a row... 19:07
hmmmm, may have really broken it. 19:09
dalek rl6-roast-data: 4d6c3d3 | coke++ | / (4 files):
today (automated commit)
 19:10
rl6-roast-data: 7a366a2 | coke++ | perl6_pass_rates:
fix yesterday's rakudo summary.
 19:12
colomon [Coke]: seem to have it working again.
dalek ast: b40d7fe | dagurval++ | S32-io/io-handle.t:
Added tests for IO flushing when going out of scope
FROGGS [Coke] / arnsholt: what I said earlier about commenting out two lines in src/core/terms.pm: this fixes both S02/types.t and zavolaj O.o 19:15
19:15 SmokeMachine left
FROGGS so I just need to get that right, and I think it is about build order 19:15
[Coke] ... do any spectests or other roast tests fail if you do that? O_o 19:16
FROGGS I believe so, but didnt tested it
19:17 kaare_ left
FROGGS tests about $*ARGFILES and $*TMPDIR must fail though, because these aren't defined 19:17
[Coke] colomon: nope, still just loading.
colomon [Coke]: weird. I dunno what's going on here, I'm going to take it offline for now... 19:19
19:20 SmokeMachine joined
colomon huh, it's working well enough for me locally. :\ 19:22
19:26 parrot_dukeleto is now known as dukeleto 19:47 wooden joined
colomon finally put it in github: github.com/colomon/SmokeResults 19:49
20:03 REPLeffect left 20:08 REPLeffect joined
flussence oh, no wonder Text-Tabs-Wrap wasn't fixed by that commit... it'd help if it actually *used* Test-Corpus instead of an internal module that does the same thing. 20:08
20:29 snearch joined 20:31 SamuraiJack left 20:39 Chillance_ joined 20:41 berekuk left, saxx11 left
colomon flussence: :) 20:41
20:42 saxx11 joined, Chillance left 20:45 berekuk joined 20:47 FROGGS[mobile] left, FROGGS[mobile] joined 20:48 FROGGS[mobile] left 20:55 japhb_ joined 20:57 dmol left 21:01 [0xF] joined
[0xF] perl6: sub foo(Int $x is copy) { $x = 5 } foo(1); 21:02
camelia niecza v24-51-g009f999: OUTPUT«===SORRY!===␤␤Strange text after block (missing comma, semicolon, comment marker?) at /tmp/84TjVIvMIR line 1:␤------> sub foo(Int $x is copy) { $x = 5 }⏏ foo(1);␤␤Parse failed␤␤»
..rakudo 45ae2d: OUTPUT«===SORRY!===␤Two terms in a row␤at /tmp/UAuJm9SJJ0:1␤------> sub foo(Int $x is copy) { $x = 5 } ⏏foo(1);␤ expecting any of:␤ postfix␤ statement end␤ statement modifier␤ statement modifier loop␤ …
[0xF] perl6: sub foo(Int $x is copy) { $x = 5 }; foo(1);
camelia rakudo 45ae2d: OUTPUT«Cannot modify an immutable value␤ in block at /tmp/DKANrufkoO:1␤␤»
..niecza v24-51-g009f999: ( no output )
[0xF] perl6: sub foo($x is copy) { $x = 5 }; foo(1);
camelia rakudo 45ae2d: OUTPUT«Cannot modify an immutable value␤ in block at /tmp/MSeAbyoUK5:1␤␤»
..niecza v24-51-g009f999: ( no output )
21:03 bluescreen100 left, bluescreen10 left, [0xF] left
labster r: say $*ARGFILES 21:05
camelia rakudo 45ae2d: OUTPUT«IO::ArgFiles.new(args => [], filename => Any, ins => Any, ins => Any, chomp => Bool::True, path => Any)␤»
FROGGS hi labster 21:06
labster hi FROGGS
FROGGS btw, setting $PROCESS::TMPDIR breaks two tests on feather
and I have no idea why
labster I don't either. I saw that, so I'm seeing if using nqp::bindkey instead of direct assignment makes a difference 21:07
FROGGS no, checked that
still crashes
labster huh... strange. which tests in particular? 21:08
FROGGS it crashes already when calling IO::Spec.tmpdir()
labster whoa, okay.
21:08 bluescreen100 joined, bluescreen10 joined
FROGGS I'm making another test now that just does: nqp::say( IO.Spec.tmpdir() ) within that terms.pm 21:10
I think it will print '/tmp/' when precompiling modules like lib.pm, but will crash then too
labster maybe it's the use of state variables? the only other place it's used in the setting is in eval(). 21:12
FROGGS hmmm, maybe, yes 21:13
labster $*TMPDIR isn't specced, so you can take it out... but if we have a $*TMPDIR, there's really no need to have state variables in there anyway.
21:14 linguini left
FROGGS okay, I will test that next (having a my-scoped variable in that class instead of a state variable) 21:15
as I said, my earlier run printed '/tmp' and crashed after the test run 21:16
labster actually... it's three classes... why don't I just rewrite it really fast? 21:17
FROGGS I just care about Unix atm
okay, compiling and waiting 21:18
labster me too... it looks like that section could stand to be refactored. 21:26
FROGGS maybe it dies because you're using %*ENV within tmpdir? maybe we should use nqp::at_key 21:27
because it looks like the state variable wasn't the problem
now I just return a string to just see that the method's body must be it 21:28
labster sure. Why is feather different, I wonder?
FROGGS it is running in a qemu vm, I guess that's why 21:29
labster if it's not that, it could be the my $submodule line (line 16) in Spec.pm 21:31
FROGGS ohh 21:32
yeah, maybe...
okay, it works when returning '/tmp' directly 21:36
labster oh good, I didn't want to think about the $submodule case breaking 21:37
FROGGS :o) 21:39
labster can you try my nom branch: github.com/labster/rakudo 21:42
FROGGS I believe %*ENV is the problem, because it gets declared later in the setting 21:44
will test you fork in a bit, but this will take at least 30mins 21:45
21:45 rindolf joined
FROGGS I'll test now that method with only %*ENV<TMPDIR> commented out, and in parallel I'll clone your fork 21:47
labster eh, more than likely, you're right. Although it's kind of annoying because %ENV is in scope when it's called. 21:52
r: nqp::say(nqp::atkey(nqp::getenvhash(), nqp::unbox_s('TMPDIR'))) 21:53
camelia rakudo 45ae2d: OUTPUT«␤»
labster huh.
FROGGS nqp: nqp::say(nqp::atkey(nqp::getenvhash(), nqp::unbox_s('TMPDIR')))
camelia nqp: OUTPUT«Can only use repr_unbox_str on a SixModelObject␤current instr.: '' pc 46 ((file unknown):149089253) (/tmp/gTj7WhgZbF:1)␤»
FROGGS nqp: nqp::say(nqp::atkey(nqp::getenvhash(), 'TMPDIR')) 21:54
camelia nqp: OUTPUT«␤»
labster is $TMPDIR not defined on feather?
FROGGS nqp: nqp::say(nqp::atkey(nqp::getenvhash(), 'TERM'))
camelia nqp: OUTPUT«screen␤»
labster that works locally
21:54 spider-mario left 21:55 berekuk left, berekuk joined
FROGGS it works atm 21:56
I'll replace the commented out $*ENV... by nqp::atkey(nqp::getenvhash(), nqp::unbox_s('TMPDIR')) 21:57
22:01 dagurval left 22:02 lustlife left
dalek kudo-star-daily: cff7958 | coke++ | log/ (5 files):
today (automated commit)
 22:10
[Coke] ^^ all green! 22:11
jnthn evening, #pelr6
[Coke] jnthn: HIIIIIOOOOOOO 22:12
jnthn has safely made it to Poland, attended the pre-conf beer, etc. :)
22:12 sqirrel left
sorear YO 22:15
22:16 bluescreen10 left, bluescreen100 left
jnthn hacks slides while waiting for masak to show up 22:16
FROGGS [Coke]: how can it be green already?? 22:18
timotimo jnthn: did you idly look at my Rakudo::Debugger pullrequests? 22:21
jnthn timotimo: I didn't actually get to my email to se they exist yet :) 22:27
22:28 lizmat joined
jnthn o/ lizmat 22:29
lizmat hi jnthn, long time no see :-)
we made it back to the hotel ok :-)
jnthn yay :) 22:30
Glad the taxi driver understood what we wanted. :)
lizmat will look at backlog tomorrow
jnthn :) 22:32
timotimo jnthn: i'm wondering, would cat-ing a file into perl6-debug and comparing it to "known good" output be an acceptable way to have a test case for Rakudo::Debugger?
dalek kudo-debugger: 3aff7eb | (Timo Paulssen)++ | Build.pm:
make Build.pm correspond with changes in current Panda.
 22:33
kudo-debugger: 49eea87 | jonathan++ | Build.pm:
Merge pull request #9 from timo/master

Make Build.pm work with current panda.
jnthn timotimo: I'm thinking the other patch helps but only so much 22:34
Like, it probably doesn't fix variable examination
Instead of being removed that line probably wants to be nqp::forceouterctx($pbc, $vm_ctx) or so 22:35
lizmat calls it a day after having enough cider 22:36
timotimo indeed, it doesn't work with variable examinanitions
i'll try that next.
ajr_ You may remember that I was trying to find a Perl 6 equivalent to $. According to S28, there should be a method $*IN.input_record_number() 22:37
but there isn't.
Looking at IO.pm, I see a variable $!.ins that appears to contain a record count. 22:38
timotimo $!.? that's aggressive on the twigils
ajr_ There's a test in s16-filehandle/io.t that appears to reference that ($in4.ins) 22:39
flussence r: say $*IN.^methods
camelia rakudo 45ae2d: OUTPUT«open close eof get getc lines read seek tell write opened t print slurp spurt copy chmod Str flush encoding d e f l r s w x z modified accessed changed say <anon> <anon> <anon>␤»
flussence n: say $*IN.^methods
camelia niecza v24-51-g009f999: OUTPUT«Unhandled exception: Unable to resolve method methods in type ClassHOW␤ at /tmp/E8GTCwRqXQ line 1 (mainline @ 4) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4331 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4332 (module-CORE @ 582) ␤ at… 22:40
flussence blargh.
ajr_ I've tried $*IN.ins, but that returns 0 every time. How confused am I?
timotimo jnthn: that doesn't help much; apparently nqp::forceouterctx just calls .set_outer_ctx and fails to do so just like the code that was there previously 22:41
jnthn timotimo: yes, note I got rid of the atpos
timotimo oh, right 22:42
but it still doesn't work :D
and yeah, it's true, ForeignCode *nor* Callable have that method. and i don't know what's going on anyway
jnthn oh, right... 22:43
ajr_ Sorry, that variable is $!ins
jnthn yeah, it needs some more tweaks
But I'm writing my talk for tomorrow right now :) 22:44
timotimo so, will my not-quite-actually-a-fix make it in for R*? 22:45
22:45 tgt left 22:47 woolfy joined
jnthn When will R* be? 22:48
I can probably do the right fix tomorrow afty.
timotimo probably this weekend
22:48 ajr_ left 22:52 tgt joined, jercos left 22:59 massimino joined 23:12 tgt left 23:25 rindolf left
masak oh hai #perl6, from a hotel room in Warsaw 23:41
tadzik hai 23:44
I somehow got home. I think I'm going to be sick :| 23:45
masak oh noes 23:46
too much beer, or just jungle fever?
tadzik too much rain, to little coats, too long on the bus stop 23:47
maybe it'll be better tomorrows, but I can already feel my throat hurting 23:48
timotimo that's unlucky timing :(
tadzik I wanted to go home earlier and get some rest, but it's good that I stayed, there were some monetary fuckups 23:50
23:50 SmokeMachine left
sorear hi masak 23:50
sorear wonders if that excludes Austin 23:51
tadzik sleep & 23:55
sorear & well 23:56
23:57 fgomez left
masak S03:2280: "the C<Any> type is not a real instantiable type" 23:58
rn: say Any.new
camelia niecza v24-51-g009f999: OUTPUT«Any.new(...)␤»
..rakudo 45ae2d: OUTPUT«Any.new()␤»
masak is too!
sorear I would like to make Any not instantiable 23:59
masak I've instantiated it now and then.
it's a nice little nondescript object to instantiate.
sorear so have I, but never on purpose and it always seems to make things worse
niecza's compiler contains an augment that prevents Any.new

Powered by Cro and the Raku® Programming Language.

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

Thank you!