»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.
00:17 stevan_ left 00:18 ajr_ left 00:20 stevan_ joined 00:24 dayangkun joined
dalek ecs: 346540c | labster++ | S32-setting-library/IO.pod:
add IO::Handle.IO noop
 00:26
00:41 aindilis joined
grondilu wonders why there is so much spam on RC 00:52
00:55 lustlife joined 00:58 jlaire left 00:59 jlaire joined 01:07 anuby joined 01:22 dayangkun left
dalek rl6-bench: b5ff52a | (Geoffrey Broadwell)++ | timeall:
Increase default for ENOUGH_TIME to 5.0 seconds; increase count of scale 0 compile/override time runs to same as startup_runs
 01:22
01:27 [particle]1 joined 01:30 [particle] left, japhb_ left 01:38 snoopy joined 01:57 ingy left, ingy joined 01:59 ingy left, ingy joined 02:02 ingy left 02:05 grondilu left 02:06 snearch_ joined 02:09 snearch left 02:11 adu joined 02:17 ingy joined 02:19 ingy left, ingy joined 02:51 xinming joined 02:57 snearch_ left 03:03 Vlavv left 03:08 arlinius left 03:15 Vlavv joined 03:27 leontopod_ is now known as Teratogen 03:42 RandalSchwartz joined
RandalSchwartz Uh... why is there a perl6 release, but not a rakudo release? 03:42
I'm *very* confused
rakudo seems not updated since febuary 03:43
adu hmm
RandalSchwartz did I miss a memo? :)
adu maybe it's one of those 2.0.0.0.0.0.1 releases 03:44
RandalSchwartz well - rakudo star had been good about a monthly release, and now it stoppped
very curious
adu maybe someone's on vacation
RandalSchwartz the project can't stop for one person 03:45
and for 3 months
Teratogen hey Randal =)
RandalSchwartz Hey Teratogen 03:46
do you know anything about this release interruption?
Teratogen I'm waiting for Perl 6 1.0
no, no I don't
RandalSchwartz where transaction_type in ('SALE', 'PAY') 03:50
03:50 arlinius joined
RandalSchwartz oops 03:50
03:50 preflex_ joined 03:52 preflex left, preflex_ is now known as preflex, Pleiades` left
sorear o hai 03:56
04:01 fhelmberger joined 04:02 fhelmberger left 04:06 atroxaper joined
RandalSchwartz still wondering why the rakudo thing is broken 04:09
pmichaud RandalSchwartz: no star release because we've had some errors in the modules 04:11
we _think_ we have those cleaned up now, and expect a 2013.05 release this week.
but since one of the purposes of Rakudo Star is to provide some stability across releases, it hasn't seemed appropriate to release when we know that key modules are broken. 04:12
RandalSchwartz ahh... was just used to the regular release 04:13
so it seemed odd that they stopped
pmichaud agreed.
RandalSchwartz was there a blog post to say that?
pmichaud no, there wasn't, and probably should have been. 04:14
another reason the releases have been delayed is because I've been essentially without tuits since mid-March.
RandalSchwartz ok, as an interested consumer, that might have been useful :)
pmichaud normally we'd find someone else to cut the release in that case, but with the module errors we didn't have people with the right skill set, apparently. 04:15
I think I'll add a note to the Star release guide that says that if a regularly scheduled release is missed, we make a blog post to that effect. 04:16
04:16 zby_home joined, zby_home left
dalek ar: cabee9d | pmichaud++ | modules/ (3 files):
More module updates for 2013.05 release.
 04:22
ar: 1631f37 | pmichaud++ | tools/star/Makefile:
Bump to use Parrot 5.3.0 (supported) instead of 5.2.0 (devel).
 04:23
04:26 Psyche^ joined 04:29 Patterner left, Psyche^ is now known as Patterner 04:39 anuby_ joined 04:43 anuby left 04:44 yakshavr left 04:51 avuserow joined
avuserow hi perl6ers 04:51
avuserow is giving weechat a spin
diakopter hi
avuserow hi diakopter, how goes? 04:52
diakopter differently than I'd expected the last 2 months, but interestingly awesome
sorear hello avuserow
diakopter (.. along with some large fail along the way)
avuserow so what have you been working on recently? 04:53
diakopter helping w/ yapc the last 2 months 04:55
avuserow oh cool. I'm looking forward to it :)
diakopter you're going? 04:56
or will watch the livestreams?
sorry if I don't know your name :( 04:57
avuserow I'll be going 04:58
I don't know if I ever posted much?
sorear will also be going
avuserow :D
sorear has not done much with p6 for months
04:58 birdwindupbird joined
avuserow yeah me neither. I did GSoC in 2008 for Perl 6 tests under an ancient handle 04:58
diakopter ah, there you are 04:59
sorear if you don't mind me asking, what handles have you used with perl 6 recently? 05:00
I joined the project in feb 2010, if you did nothing since then then I wouldn't know you :(
05:00 adu left
avuserow nope, nothing since then 05:00
university work kept me very busy
I'm hoping to get back into contributing occasionally, either here or on CPAN. I have a neglected module there :\ 05:01
diakopter Auzon 05:03
avuserow :)
the internet never forgets
oh right I forgot, this is when all the happy European hackers are sleeping :) 05:05
diakopter yeah it's just all us blasé far westerners 05:07
avuserow r: say "hello"
camelia rakudo 8a0859: OUTPUT«hello␤»
avuserow sweet, unicode seems to work 05:08
sorear n: say "hello"
camelia niecza v24-51-g009f999: OUTPUT«hello␤»
avuserow niecza++ :)
TimToady p: say "hello" 05:09
camelia pugs: OUTPUT«hello␤»
sorear nqp-jvm: say "moo"
nqpjvm: say("moo")
camelia nqp-jvm: OUTPUT«Confused at line 2, near "say \"moo\""␤ in panic␤ in comp_unit␤ in TOP␤ in parse␤ in parse␤ in <anon>␤ in compile␤ in eval␤ in evalfiles␤ in command_eval␤ in command_line␤ in MAIN␤ in <anon>␤ in <anon>» 05:10
Teratogen p: say "Hello, world!"
camelia pugs: OUTPUT«Hello, world!␤»
sorear nqp-jvm: say("moo")
pmichaud feather is thrashing at the moment for some reason
camelia nqp-jvm: OUTPUT«moo␤»
05:10 fhelmberger joined
sorear pmichaud: 1 or 3? 05:10
pmichaud 1, I guess.
feather.perl6.nl
avuserow is that where camelia is these days?
TimToady std: say "ok 00:00 45m"
camelia std 0336087: OUTPUT«ok 00:00 42m␤»
05:10 prevost left
pmichaud yay, all module tests pass with current R* 05:11
sorear avuserow: last I heard, camelia is split between 1 (compiler builds) and 3 (execution)
TimToady \o/ quick add more modules
pmichaud [pmichaud@feather ~]$ uptime 05:11:05 up 111 days, 9:49, 28 users, load average: 27.60, 23.51, 17.28
sorear Lovely
Teratogen moar module
s
that is a hefty load average, sorear 05:12
TimToady we'll have moar modules pretty soon now
diakopter quantum shifts a little
sorear Teratogen: you meant pmichaud 05:14
pmichaud pmichaud.com/sandbox/rakudo-star-20...rc1.tar.gz # release candidate for R* 2013.05
Teratogen oops 05:15
pmichaud I'll look for comments overnight and publish a R* release tomorrow, .msi to follow shortly after that.
sorear \o/
pmichaud anyway, bedtime here -- be back tomorrow morning 05:19
diakopter pmichaud: 'nite 05:21
05:51 dmol joined 06:01 FROGGS joined 06:11 s1n left 06:15 RandalSchwartz left 06:26 s1n joined 06:29 kaleem joined 06:49 dmol left 06:51 yves_ left 06:58 dmol joined 06:59 dmol left 07:00 domidumont joined 07:03 dalecooper joined
dalek ecs: 4f0895b | larry++ | S32-setting-library/Numeric.pod:
add lsb and msb methods to Int
 07:04
ar: abacc10 | moritz++ | docs/announce/2013.05.md:
[announce] minor formatting update
 07:17
07:20 yves_ joined
moritz I just tried to build an R* tarball myself, and somehow it didn't download rakudo and parrot 07:20
Teratogen that does not sound good 07:21
07:23 domidumont left 07:24 domidumont joined 07:28 avuserow left
sorear um 07:30
FROGGS moritz: looks like you have to do that manually 07:32
moritz I'm pretty sure that's now how it's supposed to be
sorear that is not the standard, least-surprising definition of msb and lsb 07:33
equivalently: those are not the standard, least-surprising names for ctz and clz
moritz wouldn't expect to be able to find least-surprising defintion for msb on an arbitrary-size type like Int 07:34
07:34 preflex left 07:35 preflex_ joined, preflex_ is now known as preflex
sorear +(* < 0) 07:36
(if you think 2-adically...) 07:37
07:37 DarthGandalf joined
sorear that would however still be rather problematic because my uint8 $x; $x.msb # you're calling Int.msb here even if you don't realize it because uint8 is only a a container type 07:38
07:39 sqirrel joined 07:40 rindolf joined, rindolf left 07:45 DarthGandalf left 07:47 salv0 left 07:48 salv0 joined 07:49 DarthGandalf joined 07:53 anuby_ left, anuby joined 07:54 raiph left 07:59 rindolf joined 08:03 FROGGS left 08:13 raiph joined 08:16 FROGGS joined 08:18 rindolf left 08:21 FROGGS left 08:25 mikemol left, Juerd left 08:29 anuby_ joined 08:31 Juerd joined 08:32 anuby left
masak the CSP program from yesterday runs about 2 seconds faster if I multiply all the coefficients by 100 and solve the problem with Int variables. 08:35
08:36 dakkar joined
labster morning masak 08:37
sorear feh. I started to read that and was all OOO WE CAN HAS MULTITHREADINGS?
o/ masak
08:48 mikemol joined 08:49 rindolf joined 08:50 FROGGS joined 08:54 arlinius left
bbkr pmichaud: HTTP::Easy or JSON::RPC does not work in star RC (after request is sent to server it does not return response). I'm debugging it right now 09:04
09:05 jaldhar left
bbkr can you please run following code on Star RC: use HTTP::Easy::PSGI; my $app = sub (%env) { print %env.perl; return [ 204, [ ], [ ] ]; }; my $psgi = HTTP::Easy::PSGI.new( host => "", port => 8080); $psgi.app( $app ); $psgi.run; 09:12
and then wget localhost:8080 ?
In my case wget awaits response and $app sub receives and prints it after killing wget 09:13
tadzik is it just me, or is feather a bit laggy recently/ 09:14
sorear tadzik: 22:11 < pmichaud> [pmichaud@feather ~]$ uptime 05:11:05 up 111 days, 9:49, 28 users, load average: 27.60, 23.51, 17.28 09:17
tadzik that'll explain it 09:19
azawawi's perlbrew is milling, and so are sergot's cronjobs
sorear kill them all 09:20
tadzik I don't have root here
sorear bah
09:20 risou_awy is now known as risou
sorear has root 09:20
sorear attempts to log in
sorear is having trouble... 09:21
09:21 saxx11 joined, saxx11 left
masak yeah, logging in is slow today. 09:21
lizmat you probably need patience on a box that is that loaded
sorear WOOH I HAVE THE MOTD 09:22
09:30 espadrine left
sorear I've killed the offending processes and disabled sergot's crontab 09:31
tadzik probably sergot's jobs should be running on some big niceness 09:32
sorear .tell sergot_ Please rewrite your crontab so that when the load average is 25 and single perl6 jobs take hours, you don't start MORE JOBS
yoleaux sorear: I'll pass your message to sergot_.
tadzik that'll help too
sorear guy was running five tasks per hour
Teratogen reboot 09:33
tadzik yeah, sure
sorear I forget, do I ban him on sight? 09:34
Teratogen who me?
no
09:35 eternaleye left
sorear do we even USE gitit? 09:36
there's a git-backed wiki engine running on feather1, and it's the biggest consumer of memory now
moritz it powers wiki.perl6.org/
yes, we use it 09:37
sorear :(
09:37 eternaleye joined
moritz but I'm sure you can restart it, and it'll use less memory afterwards 09:37
maybe we should start it with an ulimit
FROGGS (and/or restart it by cron.weekly) 09:38
moritz fwiw I've restartet gitit now 09:39
pmichaud maybe also see if it's being hit hard by spiders
I had trouble with that with my wiki engines at one time
sorear there's a farabi process running that's consumed a huge amount of resources; any way to check if that's actually being used?
moritz yes: kill it, and see if anybody complains 09:40
pmichaud nowadays I have .htaccess entries that refuse to serve pages to bots when the server load level > 5.0
Yahoo's Slurp bot is one of the worst offenders by far. 09:41
bonsaikitten caching, proper robots.txt to throttle robots ...
moritz pmichaud: funny, IME it's bing :-)
pmichaud: want to share your .htaccess?
pmichaud well, I haven't looked in several years. But when I did some log evaluations several years ago, Slurp was making 1500+ requests for the same page per day. 09:42
moritz pmichaud: ah, I think yahoo stopped crawling about a year ago, and now uses microsoft's index
pmichaud hmmm, I still find Slurp in my log entries. anyway, getting .htaccess
09:44 raiph left
sorear did one of you kill the azawawi stuff? 09:47
pmichaud oh, looks like I do the load-level check in the wiki software itself, not in .htaccess 09:48
sorear judging from how long the find is taking we clearly have too many files on feather...:D 09:49
shachaf What is feather? 09:50
sorear feather.perl6.nl
crappy VM that runs stuff for a bunch of people, perpetually broken 09:51
ehhh I need to sleep
shachaf Aha.
♞
sorear s/crappy/small/ mostly
but I shouldn't complain, it's donated hardware 09:52
masak aye. Juerd++
sorear goes to bed before he says any more stupid things
pmichaud what is oidentd and why are there 122 of them running on feather? ;-)
Teratogen oidentd is auth port 113 09:54
used by irc to "identify" users
there should be one (1) oidentd running
09:55 arlinius joined
Teratogen how do I know? I just installed oidentd on my amazon vps 09:55
oidentd is an identd daemon/server
there should only be one running on a box
but check oidentd.conf
maybe you have something interesting going on
there should be only one running on an instance of an operating system, I should say 09:56
09:57 ponbiki is now known as ohi3 10:07 daxim joined
dalek : ebdb97d | (Tobias Leich)++ | / (3 files):
added primitive support for subroutine signatures

It can basically handle only $ and @-tokens, where $-tokens are (wrongly) treated as optionals. Calling a $$-sub with three args will die already.		 10:11
daxim developers.slashdot.org/story/13/05...ll-runtime 10:12
pmichaud afk, sleep (hopefully) 10:16
10:20 espadrine joined 10:25 rindolf left 10:28 rindolf joined 10:35 eternaleye left
lizmat started a new spec issue "What to do with 2+ modules at UNIT level" at github.com/perl6/specs/issues/46 10:55
afk while doing preparations for trip to Austin TX 10:59
moritz daxim: thanks, looks quite interesting 11:00
11:03 espadrine left
colomon msb and lsb? 11:04
11:05 jerome__ left 11:07 jaldhar joined 11:12 perl6 joined
perl6 Qwerty 11:14
Teratogen uiop
11:14 perl6 left
lizmat r: {module A { sub a is export { say "a" } }}; import A; a # should this work? 11:15
camelia rakudo 8a0859: OUTPUT«a␤»
lizmat I would assume A would not be visible in the outer scope, yet we can import from it?
afk again 11:16
11:16 fhelmberger left 11:18 jerome__ joined 11:19 atroxaper-m joined
moritz lizmat: modules are our-scoped by default 11:19
masak was gonna say. 11:20
moritz lizmat: so it's visible from the outer scope
11:20 fhelmberger joined, eternaleye joined 11:22 atroxaper left
lizmat ok, thanks for the reality check 11:22
r: {my module A { sub a is export { say "a" } }}; import A; a # indeed fails 11:23
camelia rakudo 8a0859: OUTPUT«===SORRY!===␤Could not find module A to import symbols from␤at /tmp/3C3gQnQees:1␤------> sub a is export { say "a" } }}; import A⏏; a # indeed fails␤ expecting any of:␤ postfix␤ statement end␤ statement …
lizmat wonders what the use of an "is export" sub in a "my module" 11:24
is there a way to import that sub then at all? If not, maybe it should generate a warning ?
11:24 atroxaper joined
masak & 11:25
lizmat it being " using is export on a sub in a lexically bound module"
11:26 atroxaper-m left, atroxaper left
masak TimToady: it is not clear to me from github.com/perl6/specs/commit/4f08...736108f061 what lsb and msb actually return. specifically, isn't "the least significant bit position" always 0? 11:27
timotimo i'm kind of bummed that we can't make set(1, 2, 3, 4) ~~ set(*, *, 2, 3) work properly (or can we!) 11:28
11:29 SamuraiJack joined
colomon TimToady: what masak just said. 11:32
timotimo: don't think so
11:33 SamuraiJack_ joined 11:35 SamuraiJack left
colomon timotimo: note what you're testing is relatively easily checked using cardinality and subset, something like $set == 4 && set(2, 3) (<) $set 11:35
dalek : 87b409c | (Tobias Leich)++ | lib/Perl5/Grammar.nqp:
fix issue with space after a "use NAME" statement
 11:39
lizmat hmmm… t/spec/S11-modules/nested.t is failing for me with "make sometests", but not directly 11:50
I mean, calling the test file directly gives all ok's
11:50 sqirrel left
timotimo colomon: but then you have to write a closure for the smartmatch ;) 11:51
FROGGS lizmat: maybe this make target doesnt involve fudging?
lizmat I'm also getting TODO passed on t/spec/S32-temporal/DateTime.rakudo and t/spec/S32-temporal/calendar.rakudo
FROGGS: indeed, no fudging 11:52
FROGGS well, that explains it :o) 11:53
lizmat huh? 11:54
it didn't use to fail
FROGGS the goal was that all tests should pass without fudging, but I guess this will never happen
Teratogen 41 => perl6 11:55
bash: perl6: command not found
when will perl6 start being included in the major linux distros?
timotimo when the linux kernel gets rewritten in perl6
11:56 markstacey joined, markstacey is now known as gudahtt
FROGGS lizmat: I expected that nested.t has a #?skip in it, but it has not 11:56
Teratogen I think the linux kernel, like Perl 5, is pure C
timotimo that's why they'll have to rewrite it :)
Teratogen baw
that's no answer, man!
timotimo no, i was making a lame joke
Teratogen I know, I know 11:57
*sigh*
timotimo i know, too ;)
Teratogen I think perl6's coming of age will be when it starts getting included in major linux distros
like perl 5 and python are today
timotimo you mean "in a default installation", right?
FROGGS Teratogen: I guess they will bundle it when there is that moment when somebody decides it is ready
Teratogen and, apparently, php
timotimo er, what? php isn't on my major linux distro
Teratogen type php 11:58
and see what happens
timotimo zsh: command not found: php
Teratogen wow, it comes with Slackware
I don't know what distro you have but it must suck!
timotimo it's debian
Teratogen oh, well, I don't know what to say!
SLACKWARE UBER ALLES!
FROGGS debian++ # that should be said
11:59 anuby_ left, snearch joined
colomon timotimo: if you really want to do "that", you can define a new class which is something like "a set plus N whatevers" and write the appropriate ACCEPTS for it. 11:59
timotimo why would php be included by default? no system-related utilities use it (don't start talking about PHPGtk) and you can hardly make use of it on the commandline either
dalek ast: 265ab23 | (Elizabeth Mattijsen)++ | S32-temporal/ (2 files):
Unfudge 2 tests that are passing now
lizmat dalek seems slow: that was 2+ minutes ago
FROGGS lizmat: do you have a A.pm file somewhere around? 12:01
lizmat grrr… actually I do
FROGGS that would explain why nested.t failed
lizmat perhaps the test should unshift, rather than push ?
FROGGS had that "problem" two weeks ago, and then jnthn++ pointed him to it 12:02
sounds sane
timotimo Teratogen: on my mint, it'll say "php will not execute and completion _php exists."
Teratogen try php -v
12:03 SamuraiJack_ left
dalek ast: 92e7ba5 | (Elizabeth Mattijsen)++ | S11-modules/nested.t:
Make sure that these test's external files are found first
 12:03
FROGGS I get an installer hint for rakudo, nqp and parrot on my linux box, even if these installable packages are at least a year old or so
timotimo same - it's what zsh says when the program is not installed, but there is a completion script available 12:04
also, when i hit return a few times, it'll say zsh: correct 'php' to '_php' [nyae]? n
zsh: command not found: php
lizmat really afk for a few hours now
12:04 SamuraiJack_ joined
FROGGS how to pronounce "nyae"? 12:04
nüää?
timotimo "naja, äh" 12:05
FROGGS *g*
huf the nya part is fairly straightforward 12:06
Teratogen weird
huf like lin nyan cat, without the n
Teratogen which php
see what that says =)
sounds like you have a partial installation of php, somehow
timotimo php not found
no, it's a "no installation of php at all"
Teratogen why for the love of heaven are you using zsh? 12:07
timotimo all those completions come with zsh
because i don't feel comfortable with fish
Util FROGGS: I say it as "No, Yes, Abort, Edit". :^)
12:07 xinming left
FROGGS :P 12:07
FROGGS now takes a walk with his coffee mug after this php install-party :o) 12:08
timotimo more like "php is already installed"-party 12:09
12:09 xinming joined
timotimo anyway, i need to tidy up and clean my apartment, and that could take a while :) 12:09
FROGGS .oO( is it about a girl or his mom? ) 12:11
timotimo two male friends are going to stay over for GPN 12:13
FROGGS then I see no reason for cleaning *g*
maybe I can make it to the GPN too some day, sounds fun 12:15
12:16 bowtie left 12:18 bowtie joined 12:22 jaldhar left, SamuraiJack_ left 12:28 SamuraiJack_ joined
[Coke] (regularly scheduled R* releases) I think part of the problem is that they were not regularly scheduled. they just happened each month. (I thought R* was NOT intended to be monthly) 12:33
pmichaud R* is intended to be regularly scheduled. We declared they'd be monthly again because we expected a fair bit of monthly improvement in the compiler. 12:35
which is indeed what happened, but then the releases were held up due to module issues, which we didn't expect.
s/expect/plan for/
then they stopped being monthly, without any real explanation of why. 12:36
12:36 fhelmberger_ joined
[Coke] ok; I missed the announcement that they turned monthly, or I'd have tried harder to get it out the door. whoops. 12:37
not much harder, mind. :)
12:38 fhelmberger left 12:41 PacoAir joined 12:43 konundra joined 12:44 SamuraiJack_ left, snearch left 12:46 ajr joined 12:47 ajr is now known as Guest447, Guest447 is now known as ajr_
daxim rakudo, 'foo.bar.baz' ~~ '\w' 12:50
rakudo, ['foo.bar.baz' ~~ '\w'].perl.say
rakudo: ['foo.bar.baz' ~~ '\w'].perl.say
camelia rakudo 8a0859: OUTPUT«[Bool::False]␤»
daxim rakudo: 'foo.bar.baz' ~~ '\w'
camelia rakudo 8a0859: ( no output )
FROGGS daxim: what do you expect? 12:51
daxim rakudo: 'foo.bar.baz' ~~ m/\w/ 12:53
camelia rakudo 8a0859: ( no output )
daxim rakudo: say 'foo.bar.baz' ~~ m/\w/ 12:56
camelia rakudo 8a0859: OUTPUT«｢f｣␤␤»
daxim rakudo: say 'foo.bar.baz' ~~ m:g/(\w+)/ 12:59
camelia rakudo 8a0859: OUTPUT«False␤»
daxim rakudo: ['foo.bar.baz' ~~ m/(\w+)/].perl 13:02
camelia rakudo 8a0859: ( no output )
daxim rakudo: ['foo.bar.baz' ~~ m/(\w+)/].perl.say
camelia rakudo 8a0859: OUTPUT«[Match.new(orig => "foo.bar.baz", from => 0, to => 3, ast => Any, list => (Match.new(orig => "foo.bar.baz", from => 0, to => 3, ast => Any, list => ().list, hash => EnumMap.new()),).list, hash => EnumMap.new())]␤»
daxim rakudo: ['foo.bar.baz' ~~ m:g/(\w+)/].perl.say 13:04
camelia rakudo 8a0859: OUTPUT«[Bool::False]␤»
daxim this is like I have to learn to walk all over again 13:05
timotimo what's with the [ ]? 13:07
FROGGS r: say 'foo.bar.baz' ~~ /\w+/ 13:08
camelia rakudo 8a0859: OUTPUT«｢foo｣␤␤»
FROGGS r: say 'foo.bar.baz' ~~ /(\w+)/
camelia rakudo 8a0859: OUTPUT«｢foo｣␤ 0 => ｢foo｣␤␤»
FROGGS r: say 'foo.bar.baz' ~~ /(\w+) % '.'/
camelia rakudo 8a0859: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter % (must be quoted to match literally)␤at /tmp/ThX9xcVWB_:1␤------> say 'foo.bar.baz' ~~ /(\w+) ⏏% '.'/␤Unable to parse regex; couldn't find final '/'␤at /tmp/ThX9xcVWB_:1␤------> [3…
FROGGS r: say 'foo.bar.baz' ~~ /[\w+] % '.'/
camelia rakudo 8a0859: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter % (must be quoted to match literally)␤at /tmp/Evum0_Ej3f:1␤------> say 'foo.bar.baz' ~~ /[\w+] ⏏% '.'/␤Unable to parse regex; couldn't find final '/'␤at /tmp/Evum0_Ej3f:1␤------> [3…
FROGGS -.- 13:09
r: say 'foo.bar.baz' ~~ /(\w+)+ % '.'/
camelia rakudo 8a0859: OUTPUT«｢foo.bar.baz｣␤ 0 => ｢foo｣␤ 0 => ｢bar｣␤ 0 => ｢baz｣␤␤»
FROGGS it really is better after coming around the first Stolperstein 13:11
timotimo daxim: did you perhaps want 13:12
r: say "foo.bar.baz".comb(/\w+/)
camelia rakudo 8a0859: OUTPUT«foo bar baz␤»
timotimo r: "foo.bar.baz".comb(/\w+/).perl.say
camelia rakudo 8a0859: OUTPUT«("foo", "bar", "baz").list␤»
[Coke] r: say "foo.bar.baz".comb(/\w+/).perl
camelia rakudo 8a0859: OUTPUT«("foo", "bar", "baz").list␤»
[Coke] yah, that. :)
13:14 denisboyun_ joined
daxim why can't a match simply return a list? 13:15
colomon r: say "foo.bar.baz".words.perl 13:16
camelia rakudo 8a0859: OUTPUT«("foo.bar.baz",).list␤»
13:16 xilo left
colomon daxim: a list of what? 13:17
daxim of match results, like in js and perl5
13:17 espadrine joined 13:19 xilo joined
colomon daxim: you mean of captures? 13:19
daxim yes 13:20
FROGGS r: 'foo.bar.baz' ~~ /(\w+)+ % '.'/; say $0
camelia rakudo 8a0859: OUTPUT«｢foo｣␤ ｢bar｣␤ ｢baz｣␤␤»
colomon r: say ('foo.bar.baz' ~~ /(\w+)+ % '.'/).list.perl 13:21
camelia rakudo 8a0859: OUTPUT«((Match.new(orig => "foo.bar.baz", from => 0, to => 3, ast => Any, list => ().list, hash => EnumMap.new()), Match.new(orig => "foo.bar.baz", from => 4, to => 7, ast => Any, list => ().list, hash => EnumMap.new()), Match.new(orig => "foo.bar.baz", from => 8, to => 1…
FROGGS the are still Match-es btw
colomon r: .Str.say for ('foo.bar.baz' ~~ /(\w+)+ % '.'/).list; 13:22
daxim r: say ('foo.bar.baz' ~~ /(\w+)+/).list.perl
camelia rakudo 8a0859: OUTPUT«foo␤bar␤baz␤»
rakudo 8a0859: OUTPUT«((Match.new(orig => "foo.bar.baz", from => 0, to => 3, ast => Any, list => ().list, hash => EnumMap.new()),),).list␤»
13:23 markstacey joined
colomon r: .Str.say for ('foo.bar.baz qix' ~~ /(\w+)+ % '.' (\w+)/).list; 13:23
camelia rakudo 8a0859: OUTPUT«foo␤bar␤baz␤ba␤z␤»
13:23 gudahtt left, markstacey is now known as gudahtt
colomon r: .Str.say for ('foo.bar.baz qix' ~~ /[(\w+)+ % '.'] (\w+)/).list; 13:23
camelia rakudo 8a0859: OUTPUT«foo␤bar␤baz␤ba␤z␤»
FROGGS ? 13:24
colomon n: .Str.say for ('foo.bar.baz qix' ~~ /[(\w+)+ % '.'] (\w+)/).list;
camelia niecza v24-51-g009f999: OUTPUT«foo bar ba␤z␤»
colomon FROGGS: I've got nothing.
oh, I've got something
rn: .Str.say for ('foo.bar.baz qix' ~~ /[(\w+)+ % '.'] \s+ (\w+)/).list;
camelia rakudo 8a0859: OUTPUT«foo␤bar␤baz␤qix␤»
..niecza v24-51-g009f999: OUTPUT«foo bar baz␤qix␤»
pmichaud r: my @a = ('foo bar baz' ~~ /(\w+) (\w+) (\w+)/).list 13:25
camelia rakudo 8a0859: ( no output )
pmichaud r: my @a = ('foo bar baz' ~~ /(\w+) (\w+) (\w+)/).list; say @a.perl;
FROGGS ahh, of course
camelia rakudo 8a0859: OUTPUT«Array.new(Match.new(orig => "foo bar baz", from => 0, to => 1, ast => Any, list => ().list, hash => EnumMap.new()), Match.new(orig => "foo bar baz", from => 1, to => 2, ast => Any, list => ().list, hash => EnumMap.new()), Match.new(orig => "foo bar baz", from => 2,…
13:25 sanug joined
pmichaud r: my @a = ('foo bar baz' ~~ /(\w+) (\w+) (\w+)/).list; .say for @a; 13:25
camelia rakudo 8a0859: OUTPUT«｢f｣␤␤｢o｣␤␤｢o｣␤␤»
colomon FROGGS: it's still not "ah, of course" to me. I mean, I understand now why qix didn't match. But why does baz appear to match twice?
pmichaud ooc, what are you all trying to do/achieve? 13:26
FROGGS colomon: that "baz ba z" is wrong there, niecza is right at this 13:27
the "ah, of course" was just because of the needed \s
colomon FROGGS: gotcha. 13:28
PerlJam good $localtime all
colomon Is this a known rakudobug?
daxim: anyway, I believe the quick answer to your question is that returning a Match for everything carries more information, and it is easily converted to a list of captures if desired. 13:29
pmichaud r: my @a = ('foo bar baz' ~~ /:ss (\w+) (\w+) (\w+)/).list; .say for @a; 13:31
camelia rakudo 8a0859: OUTPUT«===SORRY!===␤Unrecognized regex modifier :ss␤at /tmp/OZ3oiBSyAB:1␤------> my @a = ('foo bar baz' ~~ /:ss⏏ (\w+) (\w+) (\w+)/).list; .say for @a;␤»
pmichaud r: my @a = ('foo bar baz' ~~ /:s (\w+) (\w+) (\w+)/).list; .say for @a;
camelia rakudo 8a0859: OUTPUT«｢foo｣␤␤｢bar｣␤␤｢baz｣␤␤»
pmichaud r: my @a = ('foo bar baz' ~~ /:s (\w+) (\w+) (\w+)/).list; .Str.say for @a;
camelia rakudo 8a0859: OUTPUT«foo␤bar␤baz␤»
pmichaud r: my @a = ('foo bar baz' ~~ /:s (\w+) (\w+) (\w+)/).list; say @().join('|); # curious 13:32
camelia rakudo 8a0859: OUTPUT«===SORRY!===␤Unable to parse expression in single quotes; couldn't find final "'"␤at /tmp/RDHIC1_giW:1␤------> )/).list; say @().join('|); # curious⏏<EOL>␤ expecting any of:␤ method arguments␤ argument list␤ …
pmichaud r: my @a = ('foo bar baz' ~~ /:s (\w+) (\w+) (\w+)/).list; say @().join('|'); # curious 13:33
camelia rakudo 8a0859: OUTPUT«␤»
FROGGS colomon: there is a (similar ?) bugreport about backtracking, but this one here is worth reporting
pmichaud I'm not convinced there's a bug.
FROGGS r: .Str.say for ('foo.bar.baz qix' ~~ /[(\w+)+ % '.'] (\w+)/).list; # no bug?
camelia rakudo 8a0859: OUTPUT«foo␤bar␤baz␤ba␤z␤»
pmichaud looking.
PerlJam um ... what's the bug? 13:34
(or supposed bug)
colomon PerlJam: baz appears to get matched twice
PerlJam oh, I see
colomon > 'foo.bar.baz qix' ~~ /[(\w+)+ % '.'] (\w+)/
｢foo.bar.baz｣
0 => ｢foo｣
0 => ｢bar｣
0 => ｢baz｣
0 => ｢ba｣
1 => ｢z｣
FROGGS the result should be: foo bar ba z
baz shouldnt be in $0 13:35
colomon well, ba should be
FROGGS wait, I think it is reported already
ba is, and should be, yes
13:36 RandalSchwartz joined, RandalSchwartz left, RandalSchwartz joined, sanug left 13:37 skids joined
pmichaud okay, rakudobug. 13:38
FROGGS this bug? rt.perl.org/rt3/Ticket/Display.html?id=116895 13:39
13:39 bluescreen10 joined
pmichaud there's something else wrong here... rakudo *is* using a pure cursor model. 13:40
there's no notion of "take out a capture"
the problem has to be that a cursor isn't being thrown away properly.
r: say "abcde" ~~ / (a | b | bc | cde)+ »/ 13:41
camelia rakudo 8a0859: OUTPUT«｢abcde｣␤ 0 => ｢a｣␤ 0 => ｢bc｣␤ 0 => ｢b｣␤ 0 => ｢cde｣␤␤»
pmichaud r: my regex foo { a | b | bc | cde }; say "abcde" ~~ / <foo=&foo>+ » / 13:42
camelia rakudo 8a0859: OUTPUT«｢abcde｣␤ foo => ｢a｣␤ foo => ｢bc｣␤ foo => ｢b｣␤ foo => ｢cde｣␤␤» 13:43
pmichaud it's a problem from backtracking into a captured cursor 13:44
13:44 FROGGS[mobile] joined, FROGGS left 13:45 kaleem left
pmichaud ummm... hrm. 13:53
13:55 btyler joined, btyler left
timotimo hmm. tell me again what the difference between Array and List is, in practice? 13:55
because i think i'm having a little brainfart
13:55 konundra left
pmichaud an Array is a List of scalar containers 13:55
timotimo lists can be lazy, arrays can't? 13:56
13:56 konundra joined
PerlJam an array is a container, a list is not :) 13:56
Teratogen I always thought a list was something like (1,2,3)
pmichaud Teratogen: that's a Parcel
PerlJam: no, that's not really true.
arrays can be lazy: 13:57
PerlJam It should be. That way I don't have to change my mental model ;)
pmichaud r: my @a = 1..*; say @a[5];
camelia rakudo 8a0859: OUTPUT«6␤»
pmichaud the code handling subrule backtracking looks really... odd to me. 13:58
timotimo so, a list could contain non-scalar containers? or what does the first answer mean?
pmichaud yes, a list can contain non-scalar containers.
no, wait. 13:59
let me rephrase.
timotimo a hash in an array will always be inside a scalar container, but a hash in a list may be in a normal hash container, like you would get if you wrote my %foo;?
pmichaud no
a list can contain anything, containers or bare values
timotimo hm. but what does that mean in practice? 14:00
pmichaud an array places all of its elements into scalar containers
in practice, it means that the elements of a list can flatten, while the elements of an array do not flatten.
PerlJam Isn't there some nice documentation that elucidates the differences between Array, List, and Parcel ?
(If not, there should be)
pmichaud PerlJam: doc.perl6.org/language/containers perhaps 14:01
timotimo if you mean "the elements of", does that mean something like @a[1], @a[2] will always be a list of at most two elements if @a is an array, but if @a is a list, it can be more elements?
in preparing for my talk which is closing in on me at a dramatic pace, i'm realising how little i actually know about perl6. 14:02
pmichaud timotimo: no, that's not what it means
PerlJam: doc.perl6.org/language/containers#S...sty+Things # more directly 14:03
timotimo mhm
14:04 crab2313 joined
pmichaud my @a = <apple banana cherry>; my $b := (1, 2, @a, 5).list; 14:04
how many elements does $b have?
timotimo my experience would suggest it'd have 6 14:05
pmichaud that'd be correct.
what is $b[3] ?
timotimo banana
pmichaud and if you write $b[3] = 'berry'; will it work? 14:06
timotimo i think so, it would probably get the scalar container that was inside @a before, right?
pmichaud correct -- it would change @a[1]
so, that's a List. Here's an array.
oh, before I leave the List... if you write $b[0] = 'date' it will fail 14:07
because $b[0] isn't a container.
and thus you cannot assign to it.
so essentially, $b is a List consisting of 1, 2, @a[0], @a[1], @a[2], and 5.
here's the array case: 14:08
my @a = <apple banana cherry>; my @c = (1, 2, @a, 5);
what's @c[3] ? ("banana") 14:09
14:09 denisboyun_ left
timotimo now i could do @c[0] = "hello" and it will work, because "@c =" will create a scalar container for it, too 14:09
pmichaud correct.
and @c[3] is no longer the same container as @a[1]
timotimo because there was no :=
and = does copy-assign for arrays 14:10
pmichaud oh, it's not a binding issue.
my @a = <apple banana cherry>; my $b = (1, 2, @a, 5).list; # would have the same effect
I probably didn't need the bind := above.
timotimo oh, ok
14:10 crab2313 left
pmichaud A List is just a list of things. 14:10
An Array is a List where each element has been given its own scalar container. 14:11
timotimo if i had wanted $b to contain copies of @a's containers, i'd've gone @a[] instead?
pmichaud I think @a[] still gives back the containers.
14:11 prevost joined
pmichaud my @a = <apple banana cherry>; @a[].[1] = 'berry'; say @a; 14:12
r: my @a = <apple banana cherry>; @a[].[1] = 'berry'; say @a;
camelia rakudo 8a0859: OUTPUT«apple berry cherry␤»
timotimo oh
pmichaud yeah, .[] doesn't decontainerize.
timotimo is .clone for that?
pmichaud clone would do it, and I'm thinking that .values should do it also.
14:12 xilo left
pmichaud (it probably doesn't do it at present... that's one of the List/S07 issues I still need to work out) 14:12
timotimo right, S07 is still "wip" 14:13
PerlJam timotimo: copies would have been $b = [ 1, 2, @a, 5 ]; I think.
timotimo so it's kind of hard to not accidentally have two @vars share their content's containers
pmichaud timotimo: oh, I think it's pretty easy, in fact. 14:14
timotimo: I mean, shared containers is actually pretty rare.
bbiab
timotimo r: my @a = <apple banana cherry>; my $b = [ 1, 2, @a, 5 ]; say $b; $b[3] = "foo"; say $b;
camelia rakudo 8a0859: OUTPUT«1 2 apple banana cherry 5␤1 2 apple foo cherry 5␤»
timotimo r: my @a = <apple banana cherry>; my $b = [ 1, 2, @a, 5 ]; say @a; $b[3] = "foo"; say $a; 14:15
camelia rakudo 8a0859: OUTPUT«===SORRY!===␤Variable '$a' is not declared. Did you mean '@a'?␤at /tmp/ieRa9M06MS:1␤------> , @a, 5 ]; say @a; $b[3] = "foo"; say $a⏏;␤ expecting any of:␤ postfix␤»
timotimo r: my @a = <apple banana cherry>; my $b = [ 1, 2, @a, 5 ]; say @a; $b[3] = "foo"; say @a;
camelia rakudo 8a0859: OUTPUT«apple banana cherry␤apple banana cherry␤»
timotimo uh huh
i'll get back to cleaning
14:17 btyler joined
pmichaud yeah, the nqp code for handling backtracking into captured subrules isn't correct. 14:19
14:21 fhelmberger_ left 14:24 chris joined, chris is now known as Guest55566, Guest55566 left 14:25 dalecooper left 14:28 kaleem joined 14:31 xilo joined 14:39 census joined 14:40 Teratogen left, Teratogen joined 14:41 FROGGS joined 14:46 ajr joined, ajr_ left, ajr is now known as Guest58707, Guest58707 is now known as ajr_ 14:52 atroxaper joined 14:53 kaare_ joined
dalek : ffbff34 | (Tobias Leich)++ | / (3 files):
allow anonymous subs
 14:58
pmichaud 09:04 <bbkr> pmichaud: HTTP::Easy or JSON::RPC does not work in star RC (after request is sent to server it does not return response). I'm debugging it right now 15:09
09:12 <bbkr> can you please run following code on Star RC: use HTTP::Easy::PSGI; my $app = sub (%env) { print %env.perl; return [ 204, [ ], [ ] ]; }; my $psgi = HTTP::Easy::PSGI.new( host => "", port => 8080); $psgi.app( $app ); $psgi.run;
bbkr: did that used to work in previous versions of Star?
I just tried it in 2013.02 and 2012.12 and it seems to fail in those versions also.
15:10 birdwindupbird left, daniel-s left
bbkr checking 2013.02... 15:10
pmichaud I only want to block releases on regressions.
if a problem isn't a regression, it doesn't have to block the release.
15:10 Pleiades` joined, pmurias joined 15:12 daniel-s joined
pmurias pmichaud: I'm supposed to contact you about what I'm supposed to do during the community bondind period, is just hacking on the javascript backend ok? 15:12
pmichaud pmurias: since you're already familiar with the community, I suspect that getting started on the backend is okay. 15:14
I'll re-review all of the guidelines tomorrow-ish to make sure we're not overlooking something important. 15:15
bbkr pmichaud: 2013.02 is also broken. I know it is not regression but IMO in this case if you do not want to block release broken modules should be removed from *. 15:17
pmichaud are you saying that those modules are essentially useless? when were they added to R*, and were they working then? 15:18
15:22 btyler left
bbkr They were working at some point. I remember I was using PSGI server in 2012.11 or 2013.01 star releases 15:22
pmichaud I can try 2013.01.
Teratogen pmichaud, you slept three hours? 15:23
pmichaud Teratogen: less, actually :)
Teratogen ok
pmichaud had to get up and get kids to school 15:24
Teratogen ah
your spawn
pmichaud actually, not even that. :)
Teratogen so, you reproduced?
pmichaud officially, they're my stepkids
Teratogen ah
you should have a kid of your own though
don't you want to pass on your superior genes?
15:24 btyler joined
pmichaud getting a little late for that for me. :) 15:25
Teratogen baw, men can sire children up into their eighties
pmichaud Teratogen: not on their own, they can't. :)
Teratogen just need to find a young willing woman
young and willing
=)
wait is this all being recorded?
pmichaud and I'd have to get the approval of the one I have now. :)
moritz Teratogen: please simply stop being off-topic 15:26
Teratogen: even if it's not logged, it's still annoying
Teratogen ok =(
15:26 kaleem left
moritz Teratogen: and this is your last warning. 15:26
15:32 daxim left
bbkr pmichaud: I don't have star 2013.01 installed, I've tested pure 2012.12 with HTTP::Easy installed also in 2012 (those versions I had already built) and it works fine. 15:34
pmichaud gist.github.com/pmichaud/5671235 # here's what I get with 2012.12 15:35
bbkr this was fixed in github.com/supernovus/perl6-http-easy/issues/6 15:36
"bbkr closed the issue 4 months ago". so 2013.01 might be last working * release with HTTP::Easy. 15:38
pmichaud I don't follow that. 15:39
15:40 denisboyun_ joined
pmichaud Here's the question we're trying to answer: Are there any R* users that are currently using HTTP::Easy or JSON::RPC from a previous R* release that would be unable to do so in the new R* release? 15:42
bbkr yes. regression happened not on 2013.05 RC candidate but probably in 2013.02. but if someone is using 2013.01 and updates he will end up with broken HTTP::Easy based servers 15:43
pmichaud do we have an example of an HTTP::Easy program that works in 2013.01 but not in 2013.05rc1 ? 15:44
or, perhaps alternatively, is this a regression that you expect will be able to be fixed in the next 48 hours?
bbkr do you have 2013.01 compiled? 15:45
pmichaud yes.
bbkr what is the result trying the same one-liner as you just tried on 2012.12 ? 15:46
pmichaud gist.github.com/pmichaud/5671320 # result with R* 2013.01
15:48 RandalSchwartz left
bbkr oh, so binary string bug was still there. to summarize 2012.12 and 2013.01 were broken due to binary string bug and 2013.02 and 2013.05RC are broken due to infinite wait bug. so my bad - there was no working * release with HTTP::Easy since 2012.12 15:49
moritz then we should please release R* asap, and later investigate what to do with HTTP::Easy 15:50
pmichaud afaik, I have no other blockers on 2013.05rc1
which means it can become the 2013.05 release.
moritz I had an intermittent test failure with lwp-simple, but on the next 5 runs it didn't show up anymore 15:51
pmichaud I can put a note in the announcement that acknowledges a known problem with HTTP::easy
bbkr I'm not familiar with PSGI protocol, I cannot fix HTTP::Easy within 48 hours
LWP::Simple contains live test and it fails very often
pmichaud yesterday LWP::Simple was failing for me when feather was thrashing. 15:52
bbkr I mean the server that LWP tries to connect to
pmichaud I suspect that's because it's trying to connect to perl6.org (which is hosted on feather or something like that)
15:53 daniel-s left
pmichaud might be better to connect to somewhere like example.com or something like that. 15:53
bbkr github.com/supernovus/perl6-http-easy/issues/7 - this stalling issue looks familiar 15:54
pmichaud bbkr: thanks for the outstanding and quick reports on the rc candidate, btw 15:55
15:58 daniel-s joined
bbkr :) 15:59
16:00 FROGGS[mobile] left, FROGGS[mobile] joined
timotimo suddenly a clicking noise emerges from my head, as I understand why "foo".IO ~~ :e works 16:00
PerlJam
.oO( timotimo is some form of insect?!? )
 16:01
pmichaud I'm afk for a while.
16:02 FROGGS[mobile] left 16:03 mrlo left 16:05 kst` joined, Ultali joined, Ultali is now known as Ulti, btyler left 16:06 kst left 16:07 atroxaper left, mrlo joined
bbkr pmichaud: I've found also that HTTP::Easy in github.com/supernovus/perl6-http-e...4889d16b7d requires PSGI helper and github.com/supernovus/perl6-psgi is not in ecosystem 16:10
16:10 domidumont1 joined 16:13 domidumont left
bbkr oh, nvm, my panda list was outdated 16:13
16:14 adu joined 16:16 ajr_ left 16:22 arlinius left 16:25 dmol joined 16:26 djanatyn left 16:28 djanatyn joined 16:30 user_bowtie_ joined 16:31 user_bowtie_ is now known as bowtie_
bbkr pmichaud: I've nailed the issue with HTTP::Easy. it contains hack here: github.com/supernovus/perl6-http-e...pm6#L58-61 after removing those lines it responds and JSON::RPC works also 16:31
16:34 colomon_ joined, colomon left, colomon_ is now known as colomon 16:36 ggoebel left 16:40 raiph joined 16:45 konundra left 16:46 ggoebel joined 16:48 kaleem joined
bbkr pmichaud: I cannot track which IO::Socket::INET bug was this hack for. IO:: had major changes lately and looks like hack is now getting in our way instead of helping. I'm building RC2 - bbkr.org/~bbkr/rakudo-star-2013.05-rc2.tar.gz 16:50
16:51 adu left 16:52 adu joined 16:53 dakkar left 16:56 adu left
flussence bbkr: IIRc that one happened a few months ago along with the other parrot IO fallout 16:56
16:59 Chillance joined
bbkr pmichaud: my RC2 works (tested on OS X 10.7), HTTP::Easy responds, JSON::RPC also does client-server communication without problems 17:11
got to go now 17:12
17:12 bbkr left
lizmat bbkr: thank you! 17:17
17:18 pecastro joined 17:20 kaleem left 17:25 risou is now known as risou_awy, risou_awy is now known as risou 17:32 ajr joined, ajr is now known as Guest42580 17:33 Guest42580 is now known as ajr_, ajr_ left 17:34 ajr_ joined, arlinius joined
[Coke] Note that there have been failures on star related modules tests on feather for me for about a week now. 17:35
(so this a) needs a test and b) we need to bump versions included by R* to grab the test)
17:39 xinming left
dalek ecs: 617f456 | larry++ | S32-setting-library/Numeric.pod:
be clearer that lsb/msb look for 1 bits
 17:40
17:42 konundra joined 17:44 rindolf left, risou is now known as risou_awy
colomon TimToady++ 17:46
hmmm, bitwise negation is.... +^? 17:50
TimToady yes, which is what you'd want to search for the first 0 bit :) 17:55
lsb and msb are common in the culture, but maybe they should be dehuffmanized 17:56
colomon TimToady: very nice, yes, I was looking at the wrong formula. ;) 17:58
TimToady I think libtommath has an lsb finder in there somewhere
dalek rl6-roast-data: 3ce716d | coke++ | / (4 files):
today (automated commit)
 17:59
[Coke] niecza has been dirty for 316 days. pugs has been clean for 12 days. rakudo has been clean for 1 day.
colomon method lsb() { self +& -self } appears to work to get the least 1 bit
but its value, not its position 18:00
[Coke] so, one of those recent IO changes avoids the error
TimToady well, but it's unlikely to be fast for huge numbers, if there's a low-level routine in C that can do it by examining the words from smallest to largest 18:01
[Coke] wondering if it's the $.path vs $!path
colomon grumbles at TimToady's annoying correctness
TimToady well, the point is to have an optimizable target, not necessarily to optimize it right off the bat 18:03
would rather have the functionality first, so I can write RC code with it :)
colomon TimToady: no, now you've got me coding in C#
nwc10 what's the leaderboard on RC these days? 18:04
pmichaud bbkr: did you patch the upstream source or just the file directly in the R* tarball?
TimToady Tcl, Python, C, J, PicoLisp, Racket, Perl 6, D 18:05
pmichaud (looks like bbkr left.) :-(
TimToady we were briefly ahead of Racket last night :)
18:10 FOAD left 18:11 daniel-s left
nwc10 but they keep fighting back? 18:13
colomon n: say 16.ln2
camelia niecza v24-51-g009f999: OUTPUT«Unhandled exception: Unable to resolve method ln2 in type Int␤ at /tmp/QWZwoij5T6 line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4331 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4332 (module-CORE @ 582) ␤ at /home/p6…
nwc10 or, I guess, more accurately and diplomatically - people are also actively contributing Racket solutions?
colomon n: say 16.log2
camelia niecza v24-51-g009f999: OUTPUT«Unhandled exception: Unable to resolve method log2 in type Int␤ at /tmp/fq7vPmU9Mu line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4331 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4332 (module-CORE @ 582) ␤ at /home/p…
colomon n: say 16.log(2) 18:14
camelia niecza v24-51-g009f999: OUTPUT«4␤»
18:16 daniel-s joined 18:17 bowtie left 18:19 FOAD joined 18:21 daniel-s left 18:23 daniel-s joined 18:26 snearch joined 18:27 rindolf joined
colomon TimToady: is there some obvious non-iterative way to calculate msb? 18:35
18:36 bowtie_ left
TimToady well, if you know how many words are in the internal storage of, you generally only have to look in the top word 18:40
for arbitrary ints
Teratogen how would you get the msb in C? 18:41
just right shift until ...
TimToady for sized ints, you can do like the C solution in rosettacode.org/wiki/Find_first_and..._integer#C
Teratogen no wait
[Coke] Are we allowing msb functions on, say, Int? or just int ? 18:42
TimToady why would we restrict it? 18:43
[Coke] Why would we expose the internals of our non machine classes? ;)
TimToady of course, msb on a negative Int is +Inf 18:44
18:44 snuffles joined
TimToady you can do 2's complement in arbitrary precision as long as you have a way of representing an infinite number of 1 bits :) 18:45
so we can maintain that abstraction without saying how it's actually done internally 18:46
(which is just as likely to be 1's complement, but we don't care as long as the API is 2's complement) 18:47
which is specced, btw, at S02:702 18:48
but I probably need more clarification of msb semantics there... 18:50
18:50 kshannon left 18:55 domidumont1 left 18:56 FieldsaBB joined, kshannon joined, konundra left 18:59 Cugel joined
[Coke] mob 19:02
Teratogen wait for msb, why not just convert the int to binary, turn the binary into a string of 1s and 0s 19:07
and then look at the first character?
man, sometimes I am so smart I scare myself 19:08
no, wait 19:10
that can't work, it's too simple and clear 19:11
I must be doing something wrong
timotimo also, it's wasteful IMO. you should be able to do without the string conversion step by using bitwise operations 19:12
dalek ecs: 949c62a | larry++ | S32-setting-library/Numeric.pod:
explain what lsb and msb do for 0 and negatives
 19:13
19:14 kshannon left
TimToady Teratogen: yes, the point of having them as built-ins is to do something efficient rather than something elegant 19:15
19:15 Cugel left
PerlJam TimToady: beware the tyranny of the "or" ;) 19:16
TimToady in this case, I suspect tyranny is a necessary evil
and if the elegant happens to coincide with the efficient, it's usually due to the tyranny of the optimizer 19:18
19:19 gudahtt left 19:24 PacoAir left 19:25 PacoAir joined 19:33 snearch_ joined 19:35 snearch left 19:37 pecastro left
dalek kudo-star-daily: bc6b486 | coke++ | log/ (5 files):
today (automated commit)
 19:40
19:41 adu joined 19:43 kshannon joined
[Coke] panda is running (and passing) one more test. 19:45
19:47 konundra joined
PerlJam I typed "perl6" at an ubuntu command prompt on a machine that didn't have rakudo compiled and it said "The program 'perl6' is currently not installed. You can install it by typing: sudo apt-get install rakudo" 19:48
while not amazing, that's still quite neat :)
19:48 d4l3k_ joined
Teratogen that is pretty neat 19:48
19:52 cxreg2 joined, skids_ joined 19:53 dmol1 joined 19:54 dalek left, cxreg left, skids left, integral left 19:55 dmol left, lestrrat left, d4l3k_ is now known as dalek, lestrrat joined 19:57 REPLeffect left 19:59 daniel-s left, integral joined, integral left, integral joined 20:00 daniel-s joined 20:02 spider-mario joined 20:04 cognominal joined 20:07 REPLeffect joined, [particle]1 is now known as [particle] 20:11 rindolf left 20:21 snuffles left 20:22 PacoAir left 20:23 PacoAir joined, raiph left 20:24 nbleniad joined 20:25 btyler joined
timotimo i'm happy than i can do CALLER::<$=pod> 20:27
20:27 axesar joined
timotimo what's the idea behind not setting the $.level of "=item" to 1? 20:29
20:29 nbleniad left 20:31 konundra left 20:32 axesar left 20:33 domidumont joined 20:35 bruges_ left, bbkr joined
dalek osystem: ec8ff52 | Dagur++ | META.list:
added GD::Raw
 20:36
20:36 bruges joined 20:38 dmol1 left 20:40 pecastro joined
bbkr pmichaud: just the file. I've also contacted with Supernovus through github.com/rakudo/star/issues/22 so he can patch the source module 20:42
[Coke] Regarding this failure - is there a test now? 20:47
20:48 japhb_ joined 20:49 denisboyun_ left 20:51 domidumont left 20:52 kaare_ left 20:55 kshannon left 21:01 kshannon joined, Khisanth left 21:12 skids_ left 21:15 Khisanth joined 21:16 ohi3 is now known as ponbiki
[Coke] bbkr? 21:16
labster yay, I fixed the heisenbug! I have no idea how! 21:19
lizmat it's the only way! 21:22
masak "yay, I fixed the mandelbug" -- "so you believe..." 21:28
sorear mandelbugs? are those named after benoit mandelbrot?
masak "A mandelbug (named after Benoît Mandelbrot's fractal) is a bug whose causes are so complex it defies repair, or makes its behavior appear chaotic or even non-deterministic." en.wikipedia.org/wiki/Heisenbug 21:29
21:30 PacoAir left
[Coke] "Now you're just making stuff up" 21:31
however, the schrödinbug is a real thing. I find those all the time. :| 21:32
21:32 adu left
labster Or they find you 21:32
huf bad old code is always full of mandelbugs.
21:32 Boohbah joined
huf or when the problem itself was never defined... :D 21:32
Teratogen I am a big fan of cargo cult programming myself 21:33
21:33 PacoAir joined
labster Ubuntu's rakudo package is six months old. We should probably replace that soon. 21:35
[Coke] "we" ? 21:37
that's not under our control, is it?
21:37 xilo left
[Coke] wonders what the ports stance is on something like rakudo with multiple possible dependency chains. 21:38
21:39 tgt left 21:40 splitcells joined, prevost left
labster Well, someone has to submit .debs 21:42
21:43 GlitchMr joined
[Coke] ah. last I remember hearing, ubuntu is only going to pull things every so often anyway, so there's no point in updating them frequently. 21:43
... that said, we could probably automate the generate of whatever they need.
labster We should probably offer multiple packages. I wonder if we could have a rakudo-jvm and a rakudo-parrot with different config files, both of which depend on rakudo-core or something. 21:45
[Coke] eh. if we're bundling it based on a release, why try to separate out the core? 21:46
the only person we really help is the person who installs both. 21:47
(that said, if it's easy, sure.)
21:49 tgt joined 21:50 GlitchMr left, pmurias left 21:52 splitcells left
pmichaud AFAICT, someone else squatted on the rakudo packaging for ubuntu/debian 21:52
we can certainly start building our own, however.
fsvo "we"
I'd like to automate .deb and .rpm generation; I've just not had the tuits to tackle it myself yet. 21:53
21:53 xilo joined
pmichaud I do expect the build systems are going to want some attention in the next months though; the jvm build system was "make it work" more than anything else, and we'll soon be adding a javascript backend as well. 21:53
21:55 ajr_ left 21:56 proller joined, spider-mario left 21:57 splitcells joined
bbkr [Coke]: yes? 21:57
sorry, I was afk
21:59 irdan joined
bbkr and there is no test for HTTP::Easy as it requires forking to perform one 21:59
[Coke] ok. but without a test, it will break again. 22:00
is there a ticket somewhere saying we need a test? 22:01
flussence I wrote a thing for that, but it's kinda impossible to use effectively without an actual fork() function or built in IPC: github.com/flussence/Test-ClientServer
22:03 bluescreen10 left, pmurias joined
pmichaud one could potentially test using shell() or run() 22:03
bbkr [Coke], I posted issue. Supernovus is module author, I'm not sure if he plans to add tests without fork available. and mocking IO::Socket::INET is pointless as most of recent bugs came from there and underlying parrot
22:04 lustlife left
pmichaud I've also been wondering if Rakudo (Star) should fork the module repos that it includes, to provide some ability to vary from upstream versions of the modules. 22:04
bbkr shell() or run() detach process?
pmichaud I'm pretty sure shell() can background a process.
but clearly we have some way of testing client/server stuff in the spec repo
er, roast repo 22:05
so I'm not sure that fork() is a prerequisite for writing a test.
bbkr I don't recall any method that guarantees clean exit, ie. without leaving detached process that carshed or something 22:06
crashed*
pmichaud I know that IO-Socket-INET.t uses qqx{} to run a separate process for testing its client server stuff. 22:07
so clearly it's _possible_ in current Rakudo.
22:08 daniel-s left
bbkr I'll take a look at it and also on flussences module. JSON::RPC also does not have complete code coverage in tests so that may be double-useful. thanks for pointing me in the right direction. 22:09
22:09 daniel-s joined
pmichaud I don't think I want to make a Star release that has uncommitted patches in it. 22:11
(Perhaps that will be a necessary step, I'm just not sure I'm ready to make it this month)
bbkr Supernovus already responded in HTTP::Easy issue and I think he will remove this hack that breaks things very soon 22:13
another idea is to fork only those Star modules that break things in newest versions. and removing forks once pull requests with fixes are applied in original repos. managing all module forks is pointless. 22:18
pmichaud well, eventually there needs to be a way for module authors to identify "this is the stable version to include"
whether that's through release tarballs (a la CPAN) or specialized tags or ... whatever, we need to have some ability to easily sync modules bundled with Rakudo with non-head versions of the repos 22:19
bbkr maybe fat tags? git external points to fixed commit so tools are already there, only procedures are needed 22:20
pmichaud and it shouldn't be solely up to the downstream packagers to decide where the valid "safe" commit points are.
s/decide/figure out/ 22:21
lizmat pmichaud: hopefully we can discuss this in Austin and work this out, including other packaging related issues? 22:23
22:23 chrisdev joined
pmichaud lizmat: yes, I'd very much like that. I suspect we'll end up with short, medium, and long-term plans here. 22:23
lizmat cool: we've just about packed here, getting a few hours of sleep before we need to get on the road 22:24
pmichaud note that currently I'm not expecting to arrive in Austin until Saturday evening, and my trip may be cut short on the back end
lizmat in ~24 hours from now, we should be in Austin
that's clear: jnthn and masak are leaving on Sun...
we just wanted some extra days in Austin before needing to get back for the French Perl Workshop :-) 22:25
masak no, we're leaving on Sat. 22:26
pmichaud I should probably update the wiki
lizmat ah, ok
*phew*
bbkr there are 176 packages in ecosystem and about 30 in Star, it would be nice to refresh what Star bundle contains, many new modules appeared since Star list was "composed" 22:27
pmichaud throughout the trip I'll have to be checking in with Paula with some frequency, and may have to head back to Dallas if things get dicey
bbkr: I don't want to add a lot more to Star until we have a better process for managing modules, though.
as it is, the 30 modules we have now have held up releasing star for several months :-/
adding more will just make that problem worse. 22:28
(okay, not all 30 modules held up the release, but the more dependencies we add the more likely there are to be problems)
bbkr also included modules needs better tests to detect regression earlier. most of star/ecosystem modules are very undertested, they sometimes contain only module loading code as the only test. 22:29
22:32 skids_ joined
bbkr good night perl6 22:33
22:35 chrisdev left 22:38 chrisdev joined, splitcells left 22:48 silug left, silug joined 22:53 woolfy left 22:54 salv0 left 22:59 chrisdev left 23:00 thou joined 23:03 jerome__ left 23:04 chrisdev joined 23:05 census left 23:07 grondilu joined
grondilu n: say .perl given my $ = <foo bar>, <0 1>; 23:08
camelia niecza v24-51-g009f999: OUTPUT«($("foo", "bar"), (val("0"), val("1")))␤»
grondilu rn: say .perl given my @ = <foo bar>, <0 1>;
camelia niecza v24-51-g009f999: OUTPUT«["foo", "bar", val("0"), val("1")].list␤»
..rakudo 8a0859: OUTPUT«["foo", "bar", "0", "1"]␤»
grondilu rn: say .perl given my $ = <foo bar>, <0 1>; 23:09
camelia rakudo 8a0859: OUTPUT«$($("foo", "bar"), ("0", "1"))␤»
..niecza v24-51-g009f999: OUTPUT«($("foo", "bar"), (val("0"), val("1")))␤»
grondilu rn: say .elems given my $ = <foo bar>, <0 1>; 23:10
camelia niecza v24-51-g009f999: OUTPUT«2␤»
..rakudo 8a0859: OUTPUT«3␤»
grondilu rn: say .elems given my $ = <foo bar>, ^2;
camelia niecza v24-51-g009f999: OUTPUT«2␤»
..rakudo 8a0859: OUTPUT«3␤»
grondilu rn: say .elems given my $ = <foo bar>, <fooagain baragain>;
camelia niecza v24-51-g009f999: OUTPUT«2␤»
..rakudo 8a0859: OUTPUT«3␤»
grondilu rn: say .elems given my $ = (<foo bar>, <fooagain baragain>); 23:11
camelia rakudo 8a0859: OUTPUT«4␤»
..niecza v24-51-g009f999: OUTPUT«2␤»
grondilu wtf
23:14 proller left 23:16 jerome__ joined, snearch_ left 23:18 risou_awy is now known as risou, adu joined 23:25 salv0 joined 23:35 jerome__ left 23:42 snoopy left 23:48 jerome__ joined 23:53 Chillance left

Powered by Cro and the Raku® Programming Language.

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

Thank you!