»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.
colomon when I run make t/spec/S03-operators/bag.t, I get 00:08
Element shifted from empty list
in method Str at src/gen/m-CORE.setting:13629
BenGoldberg danaj_, he left. 00:19
m: $*IN.gist.say 00:21
camelia rakudo-moar ec6011: OUTPUT«IO::Handle<->(opened, at line 0 / octet 0)␤»
BenGoldberg m: chdir('/usr/share/dict/') 00:22
camelia rakudo-moar ec6011: OUTPUT«Cannot modify an immutable IO::Path␤ in sub chdir at src/gen/m-CORE.setting:15680␤ in block <unit> at /tmp/IBQLjdgpoV:1␤␤»
colomon has a bad feeling that bag.t has been crashing and skipping 8 (or more) tests for a while. :( 00:28
well, the good news is it's probably my fault... 00:31
m: my @d; say ([⊍] @d).perl 00:33
camelia rakudo-moar ec6011: OUTPUT«(Failure.new(exception => X::AdHoc.new(payload => "Element shifted from empty list"))=>1).Bag␤»
colomon m: my @d; say ([⊎] @d).perl 00:34
camelia rakudo-moar ec6011: OUTPUT«().Bag␤»
dalek kudo/nom: 0971a6e | (Solomon Foster)++ | src/core/set_operators.pm:
Add (+)'s check for empty input to (.).
 01:01
colomon has gotten away with his shameless behavior there. :) 01:04
dalek ast: 7994e6b | (Solomon Foster)++ | S03-operators/bag.t:
Fix the plan, now that test 102 isn't crashing.
 01:10
colomon m: my @d; say ([⊍] @d).perl 01:11
camelia rakudo-moar ec6011: OUTPUT«(Failure.new(exception => X::AdHoc.new(payload => "Element shifted from empty list"))=>1).Bag␤»
colomon ooo, Travis-CI liked my patch too. 01:15
and e-mailed me to let me know. So polite!
tony-o m: class a { has $!t; submethod BUILD (:t(:$B<!>t)) { }; }; a.new(:t(5)); a.t.say; 01:58
camelia rakudo-moar 0971a6: OUTPUT«===SORRY!=== Error while compiling /tmp/PS9h3oWjE8␤Unable to parse named parameter; couldnt find right parenthesis␤at /tmp/PS9h3oWjE8:1␤------> ass a { has $!t; submethod BUILD (:t(:$B⏏<!>t)) { }; }; a.new(:t(5)); a…»
masak 早安, #perl6 02:17
j4jackj masak: We don't speak chinese.
masak oh. mille pardoni. :) 02:20
the first character is a sun rising over a horizon, signifying "morning". 02:21
the second character is a woman under a roof, signifying "peace/tranquility/calm". 02:22
masak hence "calm morning", hence "good morning". 02:22
j4jackj masak: are you in china or taiwan, or elsewhere? 02:25
masak yes. 02:26
j4jackj masak: stop being facetious
masak though my exact location on the face of the planet is largely irrelevant, except insofar as it influences the periods of activity during the day. 02:27
though even that can be offset by exotic sleep habits.
j4jackj: I'm currently enjoying the end of a stay in the PRC. but I am originally from elsewhere. 02:28
j4jackj I'd advise legging it. 02:29
masak duly noted. 02:30
j4jackj Why are you in Red China, anyway
bonsaikitten I'm in China because work :) 02:30
j4jackj bonsaikitten: GTFO of that shit. 02:31
bonsaikitten haven't gotten tired of Shanghai yet
j4jackj: hey, at least my taxes are below 100%
j4jackj LEL
china is proof that communism doesn't work
masak makes a wild guess that j4jackj is from the US
j4jackj (they've recently transitioned to a capitalist hybrid)
masak: no, UK but currently residing in Canada
at least the government themselves don't censor political sites 02:32
colomon j4jackj: where in Canada? 02:36
j4jackj colomon: BC 02:36
colomon cool 02:37
colomon 's annual Canadian vacation was in St John's this year. :)
bonsaikitten j4jackj: why are you so upset with China if you aren't even connected to it? 02:40
j4jackj bonsaikitten: because censorship
and Draco
bonsaikitten j4jackj: yeah, I mean, it's almost as bad as Germany
j4jackj LOL
bonsaikitten srsly
j4jackj b-b-but china chops your head off 02:41
bonsaikitten anyway, everywhere is shit, we just choose the subjectively least shitty for us
j4jackj for things germany would put you in prison maybe 15 years
bonsaikitten at least in China there's a legal basis for having my phone tapped
j4jackj LOL
bonsaikitten so that's quite agreeable
j4jackj bonsaikitten: well gtfo of it then
bonsaikitten j4jackj: why?
j4jackj I don't want my phone tapped for ANY reason
;-;
unless i've been naughty 02:42
masak didn't expect his use of Chinese characters would lead to a political discussion, and feels a bit guilty
j4jackj and by naughty, I mean... big-style naughty
bonsaikitten yeah, so I like having a legal basis for it instead of the illegal stuff in other countries
j4jackj bonsaikitten: well you haven't been naughty
bonsaikitten not saying tapping is good, but at least china has some laws that sometimes are followed
j4jackj: but I had my phone tapped in the past ...
j4jackj oh and btw the us' tapping has legal basis (PATRIOT)
bonsaikitten yes, and no :) 02:43
j4jackj kek
bonsaikitten it's legal but unconstitutional
j4jackj bonsaikitten: are you even allowed to leave chongland?
bonsaikitten but who are you going to complain to :)
j4jackj: eh, I can leave anytime I want, foreigners are merely tolerated
everywhere the same, no country wants me for more than squeezing taxes out of me
j4jackj if so, I'd recommend an itinerary of <your current city> -> San Fran -> Taiwan
even though Taiwan is still harsh on drugs, they're a lot more liberal 02:44
inb4 v&
bonsaikitten I won't go over SF ...
j4jackj bonsaikitten: via Vancouver?
bonsaikitten j4jackj: possibly, but I could fly to Taipei directly 02:44
j4jackj o_O
bonsaikitten and I prefer to not enter US airspace
j4jackj they actually have flights from red to blue china?
bonsaikitten of course
j4jackj dafuq.
bonsaikitten even to north korea and japan!
half the real estate in Shanghai is owned by Taiwanese 02:45
j4jackj that's epically backwards
bonsaikitten and no one cares. business is good, no time for hate 02:45
JimmyZ masak: Here are some chinese people actually :)
j4jackj I think just to be fair, Red China should hand over the entire south east of china (and tibet for that matter) to Taiwan
JimmyZ masak: Not just me :P
j4jackj zuz 02:46
bonsaikitten j4jackj: and the US should return New Mexico and parts of Texas
j4jackj i know tibet is southwest...
bonsaikitten heh
j4jackj bonsaikitten: hahaha
but still
JimmyZ masak: 早上好
j4jackj taiwan and prc should have a roughly equal split
masak JimmyZ: 早上好 :)
j4jackj just to even out the bullshit between them
bonsaikitten I don't agree with chinese territorial policy, but they do have enough guns to make a very good case for it
j4jackj LOL
masak .oO( NO LOL ) 02:47
JimmyZ masak: 所以什么时候离开？
j4jackj masak: suriesly
masak .oO( NO PROJECT EULER )
j4jackj Hmm, if <insert China's current head of state> and <insert blue China's current head of state> appoint me as their mediator, should I go to their talk?
bonsaikitten j4jackj: do you like being accidentally suicided? ;) 02:48
j4jackj lol
wtfux.
bonsaikitten: suriesly...
masak JimmyZ: 下个星期三
j4jackj that doesn't even make sense
suicide = killing of self
Huh, I'm thinking of setting up a political blog. 02:49
This one would be specifically about the two Chinas
JimmyZ
.oO(We should speak Esperanto to avoid discussion on politics of any country
btyler on a more perl6-ish note, I saw this project (here, I think, in the backlog?): github.com/carbin/stomp -- commandline password manager in p6 using some p5 modules via Inline::Perl5. looks super neat, has anyone played with it?
j4jackj Hahahahahaha
bonsaikitten j4jackj: oh, there's some great police reports - man shoots himself with a long rifle in the back of the head, then pulls himself up on a rope. Obviously suicide!
JimmyZ so let's speak Esperanto :/ 02:50
bonsaikitten it does make me wonder how the physics of that work :)
masak JimmyZ: mi forlasos la Popolan Respublikon la sekvan merkredon.
j4jackj masak: all greek to me 02:51
bonsaikitten: where were you born?
oh cripes 02:52
i think i got bonsai v&
bonsaikitten j4jackj: vienna
j4jackj bonsaikitten: oh hah
bonsaikitten but that's completely unrelated to the rest anyway
JimmyZ masak: Ĉinio Nacia Tago 02:53
j4jackj all i can say is... china sucks mule balls
it might be a nice place, but you've got the remnants of failed communism and dictatorship everywhere
tony-o m: class a { has $!t; submethod BUILD (:t(:$B<!>t)) { }; }; a.new(:t(5)); a.t.say; 03:01
camelia rakudo-moar 0971a6: OUTPUT«===SORRY!=== Error while compiling /tmp/wkeTTAEKAX␤Unable to parse named parameter; couldnt find right parenthesis␤at /tmp/wkeTTAEKAX:1␤------> ass a { has $!t; submethod BUILD (:t(:$B⏏<!>t)) { }; }; a.new(:t(5)); a…»
tony-o should that work? has it changed since: doc.perl6.org/language/objects ? 03:02
bonsaikitten j4jackj: says someone stuck in UK ;) 03:17
j4jackj bonsaikitten: s/UK/Canada/
bonsaikitten oh right, you managed to escape. 03:18
bonsaikitten hasn't got the best memory
if canada were further south I'd consider it ;)
j4jackj lel
vancouver is p good
bonsaikitten too far north. not enough sunlight
I want to move to a tropical location ... 03:19
j4jackj i would say Hawaii, but you're against the US. Phillipines?
bonsaikitten Carribean is quite nice, but hard to find work 03:20
TimToady is currently in Bowling Green, KY 03:22
j4jackj timotimo: nices
uh
TimToady: nices
Mouq tony-o: That's a formatting error in the docs 03:23
dalek c: b3e9f26 | Mouq++ | lib/Language/objects.pod:
Fix formatto in objects.pod
 03:25
Mouq m: class a { has $!t; submethod BUILD (:t(:$!t)) { }; method tee { $!t } }; a.new(:t(5)); a.tee.say; #tony-o 03:26
camelia rakudo-moar 0971a6: OUTPUT«===SORRY!=== Error while compiling /tmp/E01EibuZHP␤Name t used for more than one named parameter␤at /tmp/E01EibuZHP:1␤------> has $!t; submethod BUILD (:t(:$!t)) { }⏏; method tee { $!t } }; a.new(:t(5)); a.␤ …»
Mouq m: class a { has $!t; submethod BUILD (:$!t) { }; method tee { $!t } }; a.new(:t(5)); a.tee.say; #tony-o
camelia rakudo-moar 0971a6: OUTPUT«Cannot look up attributes in a type object␤ in method tee at /tmp/OGklxvORrc:1␤ in block <unit> at /tmp/OGklxvORrc:1␤␤»
Mouq m: class a { has $!t; submethod BUILD (:$!t) { }; method tee { $!t } }; a.new(:t(5)); a.new.tee.say; #tony-o
camelia rakudo-moar 0971a6: OUTPUT«(Any)␤»
Mouq Huhrm
m: class a { has $.t; submethod BUILD (:$!t)) { }; }; a.new(:t(5)).t.say; 03:28
camelia rakudo-moar 0971a6: OUTPUT«===SORRY!=== Error while compiling /tmp/D18O_KHgit␤Missing block␤at /tmp/D18O_KHgit:1␤------> lass a { has $.t; submethod BUILD (:$!t)⏏) { }; }; a.new(:t(5)).t.say;␤ expecting any of:␤ scoped decla…»
Mouq m: class a { has $.t; submethod BUILD (:$!t) { }; }; a.new(:t(5)).t.say;
camelia rakudo-moar 0971a6: OUTPUT«5␤»
Mouq *sigh* took long enough
dj_goku /win 11
tony-o they autobind now? 03:30
dalek c: 36af6f7 | Mouq++ | lib/Language/objects.pod:
Another formatto and typo

These formattos are really attributable to =config being NYI		 03:33
tony-o thank you mou 03:36
thank you mouq
jnthn .tell pmurias I'm fine with 4c5981159e being put in master 04:17
yoleaux jnthn: I'll pass your message to pmurias.
moritz \o 05:24
but please cherry-pick, not merge :-) 05:25
tony-o ~~~~~~~~~~ 05:26
moritz \o 05:28
dalek kudo/nom: 5022417 | jonathan++ | src/Perl6/Grammar.nqp:
Fix consecutive heredoc parse with Windows \r.

We failed to swallow enough after the stopper in this case, meaning that two consecutive heredocs led to an extra newline appearing in the second one.		 07:44
ast: db49548 | jonathan++ | integration/advent2013-day04.t:
Make a test robust to source having \r.
FROGGS_ does t/spec/integration/weird-errors.t also fail on nom? 08:22
hmmm, it only fails in a spectest run :/ 08:24
(in branch lax)
nine Oh how I hate it when I already have a trivial fix but fail to write a test reproducing the problem... 08:25
FROGGS_ hmmm, and t/spec/S02-types/baghash.rakudo.moar failed test 126 in a spectest run 09:10
dalek kudo/lax: 4d02272 | jonathan++ | docs/ROADMAP:
Remove tr///; it's been done.
 09:44
rakudo/lax: ec60114 | jonathan++ | docs/ROADMAP:
rakudo/lax: Remove a couple of static optimizer tasks.
timotimo o/ 10:42
vendethiel \o 10:46
grondilu Saluton al cxiuj :-) 11:00
dalek ast: 0004afd | (Tobias Leich)++ | S02-names/strict.pl:
add tests for (no/use) strict
 11:17
ast: 292aaeb | (Tobias Leich)++ | S02-names/strict. (2 files):
fix file name (.pl => .t)
 11:18
dalek kudo/nom: 5daa5de | (Tobias Leich)++ | / (5 files):
implement lax mode
 11:23
FROGGS_ .tell daxim Please make sure that patch get manually applied to the libtommath that get bundled by opensuse 11:30
yoleaux FROGGS_: I'll pass your message to daxim.
FROGGS .tell jnthn shall we add detailed information to the ROADMAP entries? like pointing to an RT ticket or rakudo issue with open TODOs for each? 11:32
yoleaux FROGGS: I'll pass your message to jnthn.
lizmat m: say $a 11:33
camelia rakudo-moar 502241: OUTPUT«===SORRY!=== Error while compiling /tmp/FoHQ5cmLqy␤Variable '$a' is not declared␤at /tmp/FoHQ5cmLqy:1␤------> say $a⏏<EOL>␤ expecting any of:␤ postfix␤»
FROGGS lizmat: since that executes a script here the strict mode is turned on 11:34
lizmat yeah, just checking
FROGGS m: 6; say $bar = 42 # that should work in a moment
camelia rakudo-moar 502241: OUTPUT«===SORRY!=== Error while compiling /tmp/cB8QUzxctR␤Variable '$bar' is not declared␤at /tmp/cB8QUzxctR:1␤------> 6; say $bar⏏ = 42 # that should work in a moment␤ expecting any of:␤ postfix␤»
lizmat $ 6 'say $a'
(Any)
FROGGS also "no strict" and "use strict" 11:35
lizmat $ 6 'use strict; say $a'
===SORRY!=== Error while compiling -e
Variable '$a' is not declared
yup :-)
FROGGS :o)
we should talk about general pragma handling in detail at some point
brb 11:36
lizmat yup, in ~ 14 days :-)
grondilu is amazed we finally have no strict
vendethiel is 6; specced to be "only" no strict? 11:37
I'll be reading the synopses again to complete my tutorial anyways 11:38
lizmat FROGGS: just a datapoint: 11:39
$ 6 'no strict "refs"; say $a'
===SORRY!=== Error while compiling -e
Missing semicolon.
lizmat but we'll take just 'use strict' for now :-) 11:40
FROGGS vendethiel / lizmat: the spec is very vague about what no strict is 11:41
dalek kudo/nom: 38453d1 | (Elizabeth Mattijsen)++ | docs/ChangeLog:
Finally lax arrives! FROGGS++
FROGGS so, strict is all about variable declaration at this point 11:42
lizmat yeah... it's a nice to have in -e right now
lizmat but perhaps not wise yet to have your code depend on 'no strict' :-) 11:43
grondilu what would be nice would be an '-e'-like option for no strict. Maybe '-E'? 11:47
lizmat -e *is* 'no strict'
grondilu oh?
lizmat if you pull and rebuild :-)
grondilu it is indeed. 11:48
cool
vendethiel #!/usr/bin/perl 11:51
v6; $x = 1;
the example in S01 is wrong, it seems. 11:52
lizmat hmmm... "But it should be easy to relax with C<-e> or a bare version number:" 11:55
"6" is not a version number, "v6" is
I would argue that FROGGS's implementation is not according to spec
also, I can see some people returning a bare '6' from a sub 11:56
m: sub a { 6 }
camelia ( no output )
lizmat m: sub a { 6 }; say a
camelia rakudo-moar 502241: OUTPUT«6␤»
lizmat m: 6; say $x
camelia rakudo-moar 502241: OUTPUT«===SORRY!=== Error while compiling /tmp/b_7I1gptsi␤Variable '$x' is not declared␤at /tmp/b_7I1gptsi:1␤------> 6; say $x⏏<EOL>␤ expecting any of:␤ postfix␤»
lizmat m: v6; say $x
camelia rakudo-moar 502241: OUTPUT«===SORRY!=== Error while compiling /tmp/CX3gqQEWKn␤Variable '$x' is not declared␤at /tmp/CX3gqQEWKn:1␤------> v6; say $x⏏<EOL>␤ expecting any of:␤ postfix␤»
grondilu in rosettacode.org/wiki/Empty_program#Perl_6, 6; is mentionned as a way to use no strict; 12:04
FROGGS lizmat: "6" does only work when it is the first statement in a compunit 12:05
timotimo i wonder if we should have a benchmark that stresses the use of proxy? 13:18
jnthn timotimo: Can't hurt. We don't do any magic with it in spesh so far, so having a benchmakr to measure improvements when we do so would be worthwhile. 13:30
yoleaux 11:32Z <FROGGS> jnthn: shall we add detailed information to the ROADMAP entries? like pointing to an RT ticket or rakudo issue with open TODOs for each?
timotimo i'm experiencing a writer's block phase that's been going on for a long time now :\ 13:31
jnthn .tell FROGGS hmm, I think the idea of the ROADMAP file is more that you can quickly scan through it... Maybe can find a way to annotate the entries like (1) and then at the botom (1) See omg.its.rt/42 blah blah 13:32
yoleaux jnthn: I'll pass your message to FROGGS.
jnthn timotimo: Having that now with my APW talk title/abstract... 13:38
jnthn hopes that returning home and being less tired will unblock him on Rakudo/Moar stuff 13:39
timotimo jnthn: i was hoping i could build something cute and present it at APW 13:40
oh well.
jnthn timotimo: There's still time... :) 13:53
timotimo i'll end up finishing it on the car ride to APW :P 13:57
or even worse, not get any sleep the night before the car ride ... >_< 13:58
timotimo good lord ... build times on rakudo-parrot were pretty long weren't they ... 14:23
FROGGS[mobile] jnthn: yes, I thought about adding #123 which refers to a github issue 14:30
jnthn Well, except we don't use github issues for Rakudo... 14:33
FROGGS[mobile] true :/ 15:03
jnthn now has a flight to the Austrian Perl Workshop :) 15:04
dalek ast: c8a5e7f | (Solomon Foster)++ | S03-operators/ (2 files):
Rearrange and re-fudge Bag symmetric difference.

Coming up with a definition of Bag symmetric difference changes the answers we are looking for. In addition I've moved the corresponding tests from set.t to bag.t, and fudged them because Rakudo cannot handle them properly yet. (ab5tract++ and I are working on making them work.)
lichtkind what is diff between $*CWD.path and $*CWD.directory? 15:44
tony-o m: say $*CWD.directory.WHAT; say $*CWD.path.WHAT; 15:51
camelia rakudo-moar 502241: OUTPUT«(timeout)»
firefish5000 lichtkind, If I were to guess, path would return the .path and actual working directory (eg. '/home/user'), and .directory like dirname, would return the directory containing CWD '/home'
tony-o id guess one is an IO::Path and the other isnt 15:52
r: say $*CWD.directory.WHAT; say $*CWD.path.WHAT;
weird
camelia rakudo-moar 502241: OUTPUT«(Str)␤(IO::Path)␤»
..rakudo-parrot 502241: OUTPUT«(timeout)»
..rakudo-jvm 502241: OUTPUT«Can't call method "syswrite" on an undefined value at /home/p6eval/jvm-rakudo/eval-client.pl line 32.␤»
tony-o m: say $*CWD.directory; 15:53
lichtkind tony-o, my real question is: why do they diplay a different dir
camelia rakudo-moar 502241: OUTPUT«(timeout)»
tony-o lichtkind: what version of p6 are you running? 15:54
lichtkind just pulled
colomon m: say <a a a b>.Bag (|) <a b c>.Set 15:55
camelia rakudo-moar 502241: OUTPUT«(timeout)» 15:56
colomon !!!
lichtkind i noticed .directory always returdes one dir op 15:59
up, then .path
firefish5000 tony-o, for me, directory returns the containing directory for any path (like dirname, which is aproprate since... oh you noticed) 16:00
tony-o sounds like what firefish5000 sugessted
firefish5000 .directory works on both files and directorys and treats them the same. there should be something like .basename? (dont count on the name) which returns just the files name ('/home/user' -> 'user') 16:01
lichtkind i speak of what return $*CWD 16:02
lizmat_ lichtkind: all of that is currently in flux with my rewrite of S16 and S32/Str 16:09
tony-o i don't get anything returned from $*CWD.directory in moar that i built two days ago
lizmat_ m: $*CWD.directory.say 16:10
tony-o nvm - .directory is a Str and .path is IO::Path
camelia rakudo-moar 502241: OUTPUT«(timeout)»
lizmat_ :-( camelia about to die again ??? 16:11
lichtkind i use rakudo
tony-o because $*CWD is a IO::Path :-)
lichtkind i just was confused that .path . directoy return something different 16:12
lizmat_ anyways, I've *specced* directory as:Returns the directory part of the path, not including the last item. Functions
equivalently to the C<dirname> shell program on Unix-like systems.
tony-o m: say $*CWD.path.path.path.path.path.directory eq $*CWD.directory 16:12
lichtkind lizmat, why? 16:12
camelia rakudo-moar 502241: OUTPUT«True␤»
lizmat because that was more or less what the old spec said 16:13
lichtkind i dont get the sense behind it
lizmat but I'm thinking of renaming it to "dirname"
tony-o well, it works that way, the part that is confusing is $*CWD.path is equivalent to $*CWD
lizmat .path on an IO::Path is a noop
firefish5000 and if you want 2 dirs up, use parent wich returns IO::Path. (directory returns Str.)
dirname would be nice, matches basename 16:14
tony-o lizmat: yea but i think that is more what is causing the confusion here, if it was understood that $*CWD.path IS $*CWD then .directory would make more sense
lichtkind at least ~$*CWD does what i expect
lizmat I'm actually considering making $*CWD a Str
but ~$*CWD should work 16:15
tony-o then we could do $*CWD.IO....
lizmat well, you already can:
tony-o it already IS
lizmat atm, .IO returns an IO::Path
(not a IO::Handle anymore)
tony-o yea and $*CWD is IO::Path :-)
lizmat yes
tony-o i'm smelling what you're stepping in 16:16
lizmat but I've run into a bit of a chicken-egg problem
lizmat as for creating an IO::Path, I need a $*CWD (according to the new spec( 16:16
)
then how can $*CWD be an IO::Path ?? 16:17
tony-o ah
lichtkind i think people expect rathr cwd to be a string
methinks
tony-o i have modules that use $*CWD as a path
IO::Path
lizmat tony-o: how ? 16:18
how do they use $*CWD as an IO::Path ? 16:19
tony-o i use it for finding plugin modules 16:20
dalek ecs: c29d697 | (Elizabeth Mattijsen)++ | S32-setting-library/IO.pod:
IO::Spec.directory -> dirname, for clarity
tony-o or, possible plugin modules 16:20
tony-o oh nvm. i abandoned doing it that way 16:21
dalek kudo-star-daily: d8575cf | coke++ | log/MoarVM-version.log:
today (automated commit)
rl6-roast-data: 567fee9 | coke++ | / (5 files):
today (automated commit)
gtodd argh
lichtkind btw $?FILE could also be complete path of the file :)
gtodd building Inline::Perl5 .... "Memory allocation failed; could not allocate 32 bytes"
sporadic sort of error
oops not Inline::Perl5 I meant "v5" 16:22
it errors trying to build Perl5::Grammar somewhere ... this is using "panda install v5"
lizmat lichtkind: there is nothing specced about the absoluteness of $?FILE 16:25
lichtkind i would find it rather nice obsolete findbin at least this part 16:26
FROGGS[mobile] gtodd: I think a nqp change broke v5 lately... I'll check in a bit 16:32
itz hmm the synopsis use p5 pod not p6 don't they (?) 17:17
lizmat except S26, afaik 17:20
moritz itz: yes, except S26
nine You got to love the P5 API. There's for getting an array's length there's av_len, av_tindex and av_top_index which in spite of their names do exactly the same. And even if these say that there's one element in the array, fetching this element can fail. 17:44
moritz $ perl -wE 'my @a; $#a = 500; say scalar @a' 17:48
501
dunno if that's related to pre-sized arrays somehow
nine Hard to say. I've only ever seen it when invoking 'can' on a non-existing package. It claims to have one result which I store in an array and when I try to convert this array to P6 it dies because av_fetch returns NULL. 17:51
nine SEGFAULTs in the warn: if (av_store(retval, i, next) == NULL) { SvREFCNT_dec(next); } else { warn("%p", *av_fetch(retval, i, 0)); 17:52
nine At least the documentation of av_fetch tells me to check the return value for NULL, so it's still my fault it breaks. But I can clearly see the decades of backwards compatible extending here :) 17:54
dalek blets: 08ad3a6 | (Herbert Breunung)++ | docs/appendix-b-grouped.txt:
update file tests
 17:57
dalek line-Perl5: ea76545 | nine++ | / (3 files):
Don't die if av_fetch returns NULL. Treat it as undef instead.
 18:02
ajr_ How difficult would it be to suppress some of the text output from the rakudo * build process? Simple I/O can dramatically increase run times of processes. 18:04
dalek line-Perl5: 3fad61c | nine++ | / (2 files):
Make "Could not find method" catchable in Perl 6.

When a method of a P6 subclassed P5 object could not be found we croaked. Set ERRSV instead so this can be cought in Perl 6 like every other exception.		 18:16
line-Perl5: 6d1e4ba | nine++ | p5helper.c:
Simplify p5_wrap_p6_object code a bit.
 18:19
dalek line-Perl5: 57a0f3b | nine++ | p5helper.c:
Add a sanity check to p5_call_p6_callable

May shorten a future debugging session. Would have helped in the past...		 18:21
dalek line-Perl5: abea653 | nine++ | / (2 files):
Add p5_sv_reftype method. May be useful for debugging.
 18:23
dalek blets: 24d28d2 | (Herbert Breunung)++ | docs/appendix- (2 files):
renamed file test links and added :p var adverb
 18:31
FROGGS .botsnack 18:42
yoleaux 13:32Z <jnthn> FROGGS: hmm, I think the idea of the ROADMAP file is more that you can quickly scan through it... Maybe can find a way to annotate the entries like (1) and then at the botom (1) See omg.its.rt/42 blah blah
:D
itz perlhacks.com/2014/09/perls-problems/
lizmat not much news there 18:43
dalek line-Perl5/v6: 83f49b2 | nine++ | / (3 files):
Allow modules loaded by Perl 5 to be written in Perl 6.

Support use v6-inline; in Perl 5 code to pass a module's code back to Perl 6 which then creates a Perl 6 module of it.
This may improve use cases where a Perl 5 framework autoloads modules that may now be written in Perl 6.		 18:45
line-Perl5/v6: b823b06 | nine++ | / (3 files):
Fix instancing classes created by v6-inline;

Needs current Rakudo HEAD. Thanks to jnthn++ for implementing GLOBAL::
itz yeah it does retread over well walked earth
moritz 18:50
FROGGS gtodd: I can install v5 using panda just fine... (on latest rakudo@moar)
tony-o FROGGS: i submitted a bug on dbiish because libpq.so doesnt load on osx 18:52
FROGGS tony-o: and I responded
tony-o i see . ill check it out
FROGGS 'libpq' might work though, since NativeCall tries file extensions according to the operating system 18:53
star-m: use NativeCall; sub foo is native('libpq') { ... }; foo()
camelia star-m 2014.08: OUTPUT«(timeout)» 18:54
FROGGS star-m: use NativeCall; sub foo is native('libpq') { ... }; foo()
camelia star-m 2014.08: OUTPUT«(timeout)»
lichtkind in nine a new project?
FROGGS lichtkind: err, what? 18:55
lichtkind oh i misread
nine was a contibuter
FROGGS nine is a human being :o)
lichtkind so Inline-Perl5/v6 is new? 18:59
lizmat lichtkind: you were at YAPC::EU, right? do you remember Stefan Seifert's lightning talk? 18:59
lizmat in which he said "Hello world" from python from perl5 in perl6 ? 19:00
lichtkind lizmat, yes i remember but he doesnt said how old it ise didnt he? 19:04
tony-o FROGGS: it worked for me, want me to submit a PR?
FROGGS tony-o: that would be awesome :o)
dalek Iish: e679f37 | tony-o++ | lib/DBDish/Pg.pm6:
Update Pg.pm6
 19:05
tony-o guess i don't need to
FROGGS let me check on linux...
FROGGS tony-o++ # seems to work 19:07
lizmat lichtkind: it more or less started then and there 19:16
dalek nda/CPAN: d1f446a | (Tobias Leich)++ | / (44 files):
first working but non-portable attempt to fetch from CPAN

We need a lot of additional depenecies here, like HTTP::UA to fetch dists when the server gives us chunks or like Archive::Tar and Compress::Zlib for unpacking the dists itself. These dists are problematic right now when it comes to portability. A real chance that it works is only given on 64bit linux atm.		 19:23
raiph .seen perltricks 19:28
yoleaux I saw perltricks 4 Jul 2014 01:25Z in #perl6: <perltricks> hey EVAL is missing from perl6/doc and I want to add it, but I'm not sure where it should go. Should it be in routines? All the routines seem to be extracted from a parent file - what would the parent of EVAL be? thanks!
dalek Heuristic branch merge: pushed 24 commits to panda/CPAN by FROGGS 19:30
raiph wonders if perltricks sent github.com/sillymoose/xperl6 to the http:://exercism.io project
dalek nda/CPAN: cce7978 | (Tobias Leich)++ | .gitmodules:
replace NativeCall clone with proper submodule
 19:36
dalek kudo/nom: 29de963 | (Elizabeth Mattijsen)++ | / (4 files):
Change order in which we parse IO::Spec c.s.
 19:43
dalek nda/CPAN: 3b16437 | (Tobias Leich)++ | / (5 files):
make all dependencies git submodules
 19:54
nda/CPAN: d5b92ae | (Tobias Leich)++ | ext/File__Tools:
removed File::Tools
 19:56
nda/CPAN: 4bcc5b5 | (Tobias Leich)++ | ext/ (3 files):
update submodules
 20:00
nine v6-inline is currently only in a branch since I want to gather some experience with the interface before merging. It's proven useful so far :) I'm just not sure if generating a class is always the right thing. 20:07
dalek kudo/nom: f02bf8f | (Tobias Leich)++ | src/core/Label.pm:
handle $?FILES being null in Label.new
 20:11
dalek kudo/nom: 6cd83e5 | (Tobias Leich)++ | src/core/Label.pm:
stringify $?FILES in Label.new

There are cases where $?FILES contain an IO::Path object.		 20:30
FROGGS panda --cpan search foo 20:33
==> Fetching www.cpan.org/MIRRORED.BY
Unable to handle source 'www.cpan.org/MIRRORED.BY'
:o(
I guess I'll debug that tomorrow
lizmat FROGGS++ 20:37
gtodd hmm can we use the new "no strict ;" in the REPL?
bartolin I'm unable to build parrot at the moment. make dies with "Can only use nqp_set_sc_for_object with a SixModelObject: pastebin.com/vJyYMCb9
lizmat gtodd: at the moment, only on the one line
bartolin: I'll check 20:38
gtodd ah ok ... or in scripts
bartolin Could that be a recent change to rakudo? (Since I'm running this daily and everything was fine for yesterdays run.)
gtodd lizmat: the repl needs line continuation and wrapping niceness :-)
bartolin lizmat: Thanks
lizmat you may need to reconfigure
gtodd bartolin: on no not parrot on FreeBSD again ! :-)
FROGGS gtodd: EVAL is problematic atm, end the REPL is just EVAL after EVAL 20:39
m: no strict; say $bar = 42
gtodd FROGGS: right ...
bartolin gtodd: Why not? 20:39
lizmat FROGGS: couldn't we EVAL all lines entered as one each time?
camelia rakudo-moar 502241: OUTPUT«===SORRY!=== Error while compiling /tmp/jsmZYzr6Kl␤Variable '$bar' is not declared␤at /tmp/jsmZYzr6Kl:1␤------> no strict; say $bar⏏ = 42␤ expecting any of:␤ postfix␤»
lizmat and start from scratch ?
lizmat each time ? 20:39
gtodd bartolin: it was building fine there for a while
FROGGS camelia got stuck?
lizmat (if a new line compiled ok)
FROGGS: yes, it's borked
FROGGS k 20:40
gtodd bartolin: ... haven't rebuilt parrot since last week
bartolin gtodd: I'm running it for a week or so (using perl6-roast-data). Even p-spectest runs fine ;-) 20:40
gtodd bartolin: :-D 20:41
bartolin gtodd: I'm new, so I'm not aware of problems in the past ...
gtodd bartolin: on FreeBSD it seemed fine until sometime in july or august
bartolin: then you fixed the build ;-)
and it's been good 20:42
bartolin gtodd: I see ;-)
But actually I get build failures for linux as well. (On my linux box it's not automated so it could be a failure on my part as well.) 20:43
(parrot on linux)
gtodd bartolin: I'm using p6 enough now that I notice the speed diff with moar ... I hope to have an 8 core 64 gig machine in the next few months
hope that is fast enough :-P parrot might not be the fastest now but it is interesting 20:44
and current 20:45
lizmat bartolin: confirmed
bartolin gtodd: I use that FreeBSD box only for smoking. So I thought "why not parrot as well".
lizmat++ Thanks again.
dalek nda/CPAN: 76785b6 | (Tobias Leich)++ | lib/Panda/Fetcher.pm:
reorder Fetcher so http rule is picked up first
 20:46
FROGGS panda --cpan search foo 20:47
==> Fetching www.cpan.org/MIRRORED.BY
==> Pinging 253 servers, please be patient
==> Saving top 10 to /home/froggs/dev/nqp/install/languages/perl6/site/panda/mirrors.json
==> Fetching artfiles.org/cpan.org/authors/p6dists.json.gz
Could not fetch artfiles.org/cpan.org/authors/p6dists.json.gz: Will not decode invalid ASCII (code point < 0 found)
okay, that's really something for tomorrow :o)
gnight #perl6
gtodd o/ 20:48
timotimo way cool 20:51
vendethiel yay :) 20:53
gtodd is there an easy shortcutish way to show git diffs the way you can do HEAD^ HEAD ... but to roll back more than one commit?
errm I mean compare earlier than most recent commit 20:54
geekosaur more ^s 21:00
HEAD^^ is 2 commit back, etc.
perhaps more interesting: 'HEAD^{/pattern}' looks for the most recent commit containing pattern in its description 21:02
lizmat parrot breakage seems some commits old already :-( 21:15
lizmat looks like 5daa5de7e is to blame for parrot breakage (which gave us lax) 21:19
.tell FROGGS seems the 5daa5de broke the parrot build
yoleaux lizmat: I'll pass your message to FROGGS.
lizmat JVM build also seems broken, but that could be my fault 21:20
gtodd geekosaur: whoa ... 21:30
geekosaur: ^^HEAD would make more sense but git has plenty o' options it's like exploring a cave 21:31
lizmat seems 6cd83e532e93 has broken the JVM build 21:51
trying to fix that one now 21:52
testing fix on moar 22:00
dalek kudo/nom: b13069b | (Elizabeth Mattijsen)++ | src/core/Label.pm:
Fix JVM build breakage
 22:04
[Coke] some java test hung for hours, had to kill it by hand 23:40
somewhere in S32-str
S32-str/capitalize.rakudo.jvm 23:41

Powered by Cro and the Raku® Programming Language.

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

Thank you!