»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
00:02 raiph joined
lizmat perlpilot might know 00:02
hope he'll be able to fix...
catching some sleep now
jnthn Same
'night
lizmat so gnight, #perl6!
00:04 colomon left, stevan_ left 00:05 FROGGS left 00:12 tgt left 00:17 FROGGS joined 00:26 dayangkun joined 00:29 zamolxes left 00:33 zamolxes joined, rurban1 joined 00:35 Suhosin left, colomon joined 00:36 rurban2 joined, rurban2 left
[Coke] ugh. niecza's bootstrap requires both versions of Order enum to work. 00:37
00:37 rurban1 left
[Coke] ... but this hack might work. 00:37
colomon errr…. huh? 00:38
what are you trying to do?
[Coke] track the spec change on Order:: 00:40
I was just fudging a ton of tests, but realized I might be able to fix the root.
TimToady you can have dups in a definition
(on the number mappings)
or you can maybe declare constants 00:41
colomon [Coke]: it may mean you need to fix the bootstrap.
TimToady constant Increase of Order = Order::Less; or so
colomon or rather, the bootstrap may need to be fixed. I've done it once.
[Coke] colomon: well, I'll see if I can get it working, then you can bootstrap, and then we can pull out the deprecated versions. 00:46
I'll leave that to you, though.
colomon okay
might not be able to get it done tonight, I've got to run to a Christmas music rehearsal as soon as my wife gets off work. 00:47
I just moved the Advent calendar TOC to where it belongs. 00:50
00:51 ivanshmakov left
lue r: say 3 cmp 5; 00:52
00:53 stevan_ joined
camelia rakudo-jvm 262e60: OUTPUT«(timeout)» 00:53
..rakudo-parrot 262e60: OUTPUT«Less␤»
[Coke] eh. if the non-bootstrapped version works, there's no rush to update the bootstrap at all.
lue p: say +(3 cmp 5;)
camelia rakudo-parrot 262e60: OUTPUT«-1␤»
[Coke] colomon: doesn't matter, my simple fix broke more than it fixed. 00:56
colomon want me to take a look at it from scratch? 00:58
[Coke] it'd unbust a lot of niecza failurs if you get it working. 00:59
-my enum Order (:Increase(-1), :Same(0), :Decrease(1));
+my enum Order (:Less(-1), :Same(0), :More(1), :Increase(-1), :Decrease(1));
that fails. ;)
colomon are Increase and Decrease still in the standard, or was that just your attempt at backward compat? 01:01
TimToady that's just backcompat 01:02
n: my enum Order (:Less(-1), :Same(0), :More(1), :Increase(-1), :Decrease(1)); say Order(-1) 01:04
camelia niecza v24-98-g473bd20: OUTPUT«Less␤»
TimToady how does it fail?
n: my enum Order (:Less(-1), :Same(0), :More(1), :Increase(-1), :Decrease(1)); say 3 cmp 4
camelia niecza v24-98-g473bd20: OUTPUT«Increase␤»
TimToady ah, I see
01:05 zakharyas left
[Coke] n: my enum Order (:Increase(-1), :Same(0), :Decrease(1), :Less(-1), :More(1)); say 3 cmp 4 01:09
camelia niecza v24-98-g473bd20: OUTPUT«Increase␤»
colomon seemed to work for me on the first try? I didn't do anything fancy.
what spectests should I look to? 01:10
t/spec/S03-operators/cmp.t 01:11
spectesting 01:12
01:13 raiph left, mstwu00 joined
colomon hasn't tried bootstrapping, but seems to have everything working. he is suspicious this is too easy. 01:23
01:25 arcterus joined 01:26 tipdbmp left, tipdbmp joined 01:34 FROGGS left
colomon [Coke]: I show make spectest failing tons of tests. Like. more than have Order in them. Is this normal? 01:36
01:37 rurban1 joined
dalek ecza: 74c1bda | (Solomon Foster)++ | lib/CORE.setting:
Bring Order up to current spec.
 01:41
colomon not sure if that's really right or not, I'm trying a spectest without it on my Linux box.
01:42 rurban1 left, raiph joined, jnap left 01:56 jnap joined, pmurias left
TimToady lots of failures, but none of them seem related to Order 02:07
well, lots of warnings, anyway 02:08
timotimo jnthn++ fulfilled a quest i had on questhub and i received all the energy points for it :P
02:09 araujo left 02:10 araujo joined, mstwu00 left
TimToady looks like it failed about 19 files 02:12
seems about right
timotimo :o 02:25
i'd pay 71 euros for a ride to paris 02:26
but it'd only take 3 hours
meh. 02:27
TimToady masak: you might want to reply to the 2nd comment on Day 2, which seems to think that type objects are the only "class objects" in Perl 6 02:53
02:57 TimK1 joined, TimK1 left 03:02 tmetro joined
dalek ecza: c4668c6 | (Solomon Foster)++ | t/spectest.data:
Remove keybag and keyset tests.
 03:11
03:14 Vlavv` left 03:17 hypolin joined, hypolin left
dalek ecza: e07311c | (Solomon Foster)++ | lib/CORE.setting:
KeySet => SetHash, KeyBag => BagHash.
 03:22
ecza: 692b4e4 | (Solomon Foster)++ | t/spectest.data:
Turn on S02-types/sethash.t.
03:29 Vlavv` joined
lue rn: say ⟅1, 2, 3, 2, 2, 1⟆.WHAT # :P 03:29
camelia rakudo-parrot 262e60: OUTPUT«===SORRY!=== Error while compiling /tmp/pw5eX61ia8␤Two terms in a row␤at /tmp/pw5eX61ia8:1␤------> say ⏏⟅1, 2, 3, 2, 2, 1⟆.WHAT # :P␤ expecting any of:␤ argument list␤ prefix or term…»
..niecza v24-102-g74c1bda: OUTPUT«(timeout)[auto-compiling setting]␤»
..rakudo-jvm 262e60: OUTPUT«===SORRY!=== Error while compiling /tmp/uiqtKOp_me␤Two terms in a row␤at /tmp/uiqtKOp_me:1␤------> say ⏏⟅1, 2, 3, 2, 2, 1⟆.WHAT # :P␤ expecting any of:␤ argument list␤ prefix or term␤…»
03:31 ssutch left 03:32 ssutch joined 03:42 TimK1 joined
TimK1 Does "use ModuleName <foo>" work on the latest Rakudo release? Same question regarding "need ModuleName; ModuleName::foo"? Neither one seems to be cooperating with me at present. 03:44
03:45 jnap left
raydiak TimK1: I haven't tried the former personally, but afaik the latter should work as long as foo is scoped with "our"...eg "our sub foo { ... }" 04:03
04:07 V_S_C joined
colomon r: say BagHash.new("a", "b", "b").pickpairs.perl 04:08
camelia rakudo-parrot 262e60, rakudo-jvm 262e60: OUTPUT«("a" => 1,).list␤» 04:09
V_S_C b'bye! 04:10
04:10 V_S_C left 04:16 jnap joined 04:21 jnap left 04:28 preflex left, preflex_ joined, ChanServ sets mode: +v preflex_ 04:29 preflex_ is now known as preflex
diakopter EVARTHINGZ A CURSOR 04:37
TimK1 raydiak: Thanks for the response. The sub in question was modified with "is export"… I assumed that implied "our," because it's exported by default; but maybe I should do some more research. 04:38
raiph www.youtube.com/watch?v=a-RAltgH8tw...mp;t=11m1s # "when I was a fundamentalist functional programmer ... you're using mutable state, you should not do that, use Haskell or a real language ... now ... I write code for a living, I kinda like imperative programming -- you can get stuff done!" 04:44
05:18 rurban1 joined 05:23 zakharyas joined
TimK1 So… The two modules I've been playing with each have subs declared "sub foo is export," which seems to be the standard form. But this means they've exported subs that are limited to lexical scope. Sounds like the considered best practice *ought* to be "our sub foo is export" so that a user can always "need Module; Module::foo," rather than cluttering up his namespace. 05:26
lue TimK1: have you tried "sub foo is export" with "use Module" ? 05:33
05:34 SamuraiJack_ joined
PerlJam TimK1: unless "namespace cluttering" is considered a feature :) 05:39
05:43 MikeFair left
PerlJam TimK1: also, the user can "use" a module into a namespace of his choosing, thus negating the "clutter" aspect if desired. 05:43
05:43 MikeFair joined
TimK1 lue: That works fine, but I was looking for a way to do the equivalent of P5: "use Module (); Module::foo." 05:46
PerlJam: I should try that. Might be a solution to my dilemma. 05:47
TimK1 Turning in now. Late here. 05:48
05:57 ssutch left
moritz \o 06:01
06:05 rurban1 left 06:13 rurban1 joined 06:16 zakharyas left 06:17 raydiak left 06:18 raydiak joined, rurban1 left
raydiak oh cool, my cursor keys work again...remind me not to get too adventurous with .tmux.conf :P 06:19
06:34 darutoko joined 06:41 TimK1 left, TimK1 joined
moritz raydiak: don't get adventurous with .tmux.conf :-) 06:47
06:48 RolentoLCH joined
raydiak moritz: now you tell me...next time remind me a couple hours ago :) 06:49
raiph www.youtube.com/watch?v=TS1lpKBMkgg # for anyone interested in scala: make sure you know about this talk 06:53
06:55 dakkar joined 07:00 kaleem joined 07:19 jnap joined 07:23 jnap left 07:26 denisboyun joined 07:29 Rotwang joined, Rotwang left 07:43 cooper left 07:44 arcterus left 07:46 dwoldrich joined 07:51 rindolf joined 07:53 cooper joined, cooper left, cooper joined
raydiak how can I find out what the underlying representation is of a particular class or object? 07:53
07:56 denisboyun left 08:00 ssutch joined 08:16 FROGGS joined
moritz raydiak: I don't think that's introspectable yet 08:18
nwc10 jnthn: current nqp-indy-whatsit breaks Rakudo setting compuilation after Stage ast: 08:21
java.lang.NullPointerException in name (gen/jvm/stage2/QASTNode.nqp:375) in (src/Perl6/Optimizer.nqp:856)
oh, naughty terminal. 08:22
08:29 dmol joined
raydiak well...where can I read about the specific separate representations and what they do? is Array a P6Array? or does such a thing even exist? maybe it's a P6Opaque like a lot of other things, but then...where does it store the values? and why do I sometimes see Opaque capitalized and othertimes not? is there some big swath of a synopsis I skipped over somewhere? 08:30
moritz raydiak: the spec is pretty light on representations, since most of it is implementation detail 08:32
raydiak: Array's repr is P6Opaque, as far as I can tell
raydiak moritz: I suppose it's a great thing that I don't really have to care, at all, which is how I'll spin it for the advent post :) 08:34
moritz there aren't many core types that have a repr different from P6Opaque
mostly the native types, and the 'compact' array types like Blob/Buf/utf8 08:35
if you do a git grep 'is repr' in rakudo, you'll see what I mean 08:36
raydiak ah, I see...so it has to do with what's going on inside the VM, which most P6 programmers shouldn't have to worry about 08:37
moritz yes
there's a notable exception though
nativecall
raydiak and that is what I would use if I wanted to call out to C libs or somesuch? 08:38
moritz yes
for example class Point is repr('CStruct') {
has num64 $.x; has num64 $.y;
}
that will correspond to a C struct with two 'double' members 08:39
and you can pass that directly to an 'is native' function that expects such a struct, with no copying of the contents going on
see github.com/jnthn/zavolaj (scroll down to 'Structs' in the readme, for example)
raydiak it all sounds very...fast 08:40
moritz well, that's the goal
the whole point behind reprs is that sometimes, for a particular use case (optimization, or integration with other programs) you want to store your objects in a different format than usual 08:41
but that shouldn't change the object's API
and since the spec is mostly about API, there's not much about reprs there 08:42
raydiak ahhh...well that explains part of my initial confusion, then...I was wondering about details which are intentionally tucked away out of sight
08:44 [particle]1 joined
raydiak love the concept, though...being able to work with native types as if they were (mostly) normal P6 data 08:44
08:46 [particle] left
raydiak thanks moritz++, for de-confusing me yet again 08:47
FROGGS do we need the method new in class Orthotope here rosettacode.org/wiki/K-d_tree#Perl_6 ?
moritz well, you can remove it, and change the calls to .new to use named params 08:48
FROGGS k, that is what I thought
08:49 xenoterracide joined, [particle] joined 08:50 [particle]1 left
jnthn nwc10: You need the opts branch of Rakudo also. 08:52
p: say Array.REPR
camelia rakudo-parrot 262e60: OUTPUT«P6opaque␤» 08:53
moritz ok, I was wrong about the introspection part :-)
jnthn++
nwc10 aha
moritz r: say Int.REPR
camelia rakudo-parrot 262e60, rakudo-jvm 262e60: OUTPUT«P6opaque␤»
moritz p: say uint32.REPR
raydiak thanks jnthn++ :)
camelia rakudo-parrot 262e60: OUTPUT«P6int␤»
moritz p: say utf8.REPR
camelia rakudo-parrot 262e60: OUTPUT«VMArray␤»
JimmyZ p: say Mu.REPR 08:54
camelia rakudo-parrot 262e60: OUTPUT«P6opaque␤»
raydiak p: say $*OUT.REPR 09:03
camelia rakudo-parrot 262e60: OUTPUT«P6opaque␤»
09:04 preflex left
raydiak p: say Hash.REPR 09:05
camelia rakudo-parrot 262e60: OUTPUT«P6opaque␤»
nwc10 jnthn: works on "my" machine.
raydiak so, when S12 says "Other possibilities are P6hash, P5hash, P5array, PyDict, Cstruct, etc.", those are all strictly theoretical at this point... 09:07
09:08 preflex joined, ChanServ sets mode: +v preflex
jnthn Well, CStruct exists :) 09:08
Note that Hash is a P6opaque that inside of it contains a lower-level VMHash
Most of the built-in types are P6opaque as you're meant to be able to mix in to them. 09:09
raydiak but it's opaque! I'm not supposed to know what's inside of it! :)
jnthn Right, and you typically don't :P
raydiak that is interesting to know, though...something tells me that I may end up poking at internals (much) further in the future 09:11
what's in an array? is there a VMArray?
FROGGS yes, there is
jnthn Well, it's a bit more involved than that due to laziness. 09:12
Tehre *is* one for the reified bit, but then there's also an unreified bit too
raydiak oh, I didn't think of that
FROGGS hmmm, I thought this will be generated every hour? feather.perl6.nl:3000/projects.json
moritz: is it possible that a job isnt running since the last feather reboot? 09:13
JimmyZ wonders what indy branch is about 09:15
FROGGS invoke dynamic 09:16
jnthn: gah!! :P
JimmyZ JVM
moritz FROGGS: I thought it's dynamically generated
FROGGS: ask tadzik :-)
FROGGS tadzik: PING!! :D 09:17
tadzik :o
FROGGS tadzik: how is feather.perl6.nl:3000/projects.json generated?
tadzik hourly
FROGGS aha!
a cronjob?
tadzik yep
FROGGS aha!
tadzik I wonder if it's running :>
FROGGS don't seem like 09:18
tadzik 0 * * * * perl /home/tjs/modules/updatelist.pl
FROGGS I miss a module I added like 12 hours ago
umm 09:19
I am a prat
moritz does it appear on modules.perl6.org? 09:20
if not, something might be wrong with the META.info
FROGGS moritz: I borken the META.info
borked*
moritz we really need a validator/linter for that
FROGGS I hope it is valid now...
DrEeevil mmmh, bugs ... 09:22
FROGGS moritz: no, it is not on modules.perl6.org either, but I guess it will show up
DrEeevil I presume it's not desired that nqp only installs one backend after building multiple?
FROGGS no, `make install` should install all configured/built backends 09:23
DrEeevil riiight
--backends=parrot,jvm only installs nqp-p
using 2013.11 release tarball
moritz DrEeevil: uhm. Did it build perl6-j ? 09:26
DrEeevil moritz: yes, it built lots of java stuff
moritz: just trying to isolate --backend=java to see what gets installed then
moritz DrEeevil: that's very weird, because it uses the installed nqp-j to build perl6-j
DrEeevil ah, f'd up path 09:27
doesn't honour DESTDIR
this will be fun to debug!
moritz it runs $(PERL) tools/build/install-jvm-runner.pl "$(DESTDIR)" $(PREFIX) 09:28
(nqp's j-install, that is)
DrEeevil right. I wonder why that diverges from nqp-p 09:29
(and I bet moarvm will be even more packaging fun :) )
moritz because nqp-p has the paths baked in
09:39 xinming left 09:40 denisboyun joined 09:41 RolentoLCH left 09:42 xinming joined 09:44 ivanshmakov joined
raydiak okay, have a couple paragraphs started for my advent post, should finish it off tomorrow; then work like a madman to get Pray ready in time for a later post...but for now, bed :) 09:52
g'night, #perl6 \o
moritz good night
09:53 denisboyun left
DrEeevil Welcome to MoarVM! 09:53
Updating submodules .................................... FAIL
git error: fatal: Not a git repository (or any of the parent directories): .git
oh come on
09:53 dwoldrich left
FROGGS :( 09:53
DrEeevil f'ing github craperony
their tarballs are broken, known issue for over a year now 09:54
DrEeevil just cooks new tarballs then
10:03 kivutar joined
moritz any volunteers for the Saturday advent slot? 10:03
10:04 segomos left
moritz takes a looka tht the usual suspects, and correlates with the number of claimed slots 10:04
10:04 dmol left, segomos joined
moritz PerlJam! 10:05
frettled Well volunteered!
10:06 dmol joined
moritz frettled! 10:07
10:07 fhelmberger joined
moritz also well volunteered! 10:07
masak TimToady: perl6advent.wordpress.com/2013/12/0...mment-3335
moritz arnsholt: want to write something for the advent calendar, maybe?
masak r: class is { method foo { say "can call a class 'is'!" } }; is.new.foo 10:08
camelia rakudo-parrot 262e60, rakudo-jvm 262e60: OUTPUT«can call a class 'is'!␤»
masak \o/
r: class is is Array { method foo { say "can call a class 'is'!" } }; is.new.foo
camelia rakudo-parrot 262e60, rakudo-jvm 262e60: OUTPUT«can call a class 'is'!␤» 10:09
masak wow, frettled and PerlJam are stepping forward? that's awfully nice of them.
moritz I am stepping them forward :-) 10:10
masak yes. the "volunteer" as a v.tr. :) 10:11
10:12 hummeleB1 joined 10:14 preflex_ joined, ChanServ sets mode: +v preflex_, preflex left 10:15 preflex_ is now known as preflex 10:16 SamuraiJack_ left 10:18 SevenWolf left 10:24 raiph left
FROGGS tadzik / moritz: JSON::Pretty is now in projects.json! so it is all right :o) 10:33
moritz \o/
but... is it pretty? :-)
tadzik (:
FROGGS it is :o)
moritz is it web scale? :-)
frettled masak: aw, shucks ;) 10:34
Regretfully, there's not enough time for that :(
hoelzro morning #perl6! 10:37
10:37 filly joined
FROGGS ho hoelzro 10:38
tadzik ain't nobody got time for that! 10:39
10:40 xenoterracide left
hoelzro [...] is a non-capturing group, right? 10:40
tadzik yes
hoelzro so if I have [ ... <postfixish> ], does $<postfixish> exist?
tadzik I think so 10:41
moritz yes
hoelzro mkay 10:42
tadzik you can do it, it's all up to you mmmkaay
hoelzro is trying to understand why postfixish doesn't exist in the AST dump for 'foo.bar'
masak hoelzro: yes, it exists. 10:44
hoelzro: grouping with [] is invisible to capture groups. 10:45
frettled: aww. that's leaving a lot of volunteering to PerlJam, you know :P
hoelzro I see
frettled masak: yeah, I know, but PerlJam is more than up to the task. I'm sure he can jam two calendar entries in there! 10:51
masak :) 10:54
10:55 kivutar left 10:56 kst` joined, kst left, Alina-malina left 10:57 Alina-malina joined
hoelzro that's his jam, after all 10:59
10:59 stevan_ left 11:03 vaclavstrachon joined 11:05 dayangkun left
hoelzro grr 11:16
how do I iterate over the key parts of $/ in NQP?
nqp::iterator($/) doesn't cut the mustard
11:22 jnap joined
FROGGS nqp-p: my $/ := "abc" ~~ / $<word>=\w+ /; my $i := nqp::iterator($/.hash); say(nqp::iterkey_s(nqp::shift($i))) 11:24
camelia nqp-parrot: OUTPUT«word␤»
FROGGS nqp-p: my $/ := "abc" ~~ / $<word>=\w+ /; my $i := nqp::iterator($/.hash); say(nqp::iterval(nqp::shift($i)))
camelia nqp-parrot: OUTPUT«abc␤»
FROGGS nqp-p: my $/ := "abc" ~~ / [$<word>=\w]+ /; my $i := nqp::iterator($/.hash); say(nqp::iterval(nqp::shift($i)))
camelia nqp-parrot: OUTPUT«3␤»
hoelzro ah, .hash
thanks FROGGS
FROGGS np
$/ keeps a .list and a .hash 11:25
11:27 jnap left 11:28 denisboyun joined 11:36 denisboyun left 11:37 denisboyun joined 11:52 dayangkun joined 11:54 atroxaper joined 11:59 denisboyun left 12:00 dayangkun left 12:08 kivutar joined 12:10 Khisanth left
timotimo ohai 12:15
12:18 PZt left 12:21 Khisanth joined 12:22 dayangkun joined
FROGGS hi timotimo 12:22
12:26 d^_^b joined
timotimo on indy-args i get NPMCA in get_bool inside QAST :( 12:26
FROGGS like that? 12:31
<nwc10> java.lang.NullPointerException in name (gen/jvm/stage2/QASTNode.nqp:375) in (src/Perl6/Optimizer.nqp:856)
nwc10 if so, check out the branch opts 12:32
timotimo opts is a rakudo branch. i cannot build nqp in the first place 12:33
nwc10 oh, righto
no, not seeing that
timotimo weird. i may need to realclean or something
FROGGS or remove install/ 12:34
timotimo then i'd have to rebuild parrot and moarvm, too :| 12:36
seems like that helped? weird. 12:37
no, it didn't
nwc10 isn't using --gen-whatever
moritz tries too 12:38
nqp built here (manualy checking out the branch, configuring with --prefix=../install --backends=jvm) 12:39
timotimo weird! 12:43
12:45 PZt joined 12:51 p5eval joined
timotimo bisecting this i have to skip lots, because parrot didn't get the callstatic op mapped until very late 12:58
lizmat good *, #perl6! 12:59
timotimo good good, good!
colomon \o 13:00
lizmat preparing for commute to patch -p1 13:02
colomon patch -p1 is a thing and not just a unix command? 13:04
timotimo it's a perl hackathon in paris 13:05
starts tomorrow i think
masak patchathon!
colomon \o/ 13:06
moritz hackapatch
timotimo patchopatch
masak the silly. it goes up to 11. 13:07
moritz r: say ~<patchathon hackapatch patchopatch>>>.chars 13:08
camelia rakudo-parrot 262e60, rakudo-jvm 262e60: OUTPUT«10 10 11␤»
timotimo the "don't make throw-away slurpies" commit broke the parrot build for me
hoelzro ahoy lizmat 13:09
timotimo hoelzro: maybe today perl6 will enter linguist!
what do you think? :)
hoelzro I hope so! 13:10
masak "enter linguist"? that sounds elaborate.
timotimo measures times for perl6 stage compile 13:11
sweet! almost down to 100 seconds! 13:15
stage optimize is also a whole bunch faster 13:16
FROGGS ??
really?
timotimo 100 seconds for stage parse* 13:17
why do i keep writing "stage compile"?
hoelzro timotimo++ # nice!
timotimo that's not my work, that's all jnthn
FROGGS timotimo: which branch?
timotimo indy-args with HEAD~1 reverted and opts
FROGGS umm, I'll wait :o)
masak so, indy-args^, then :) 13:18
timotimo no
er ... yes
masak oh, wait, HEAD~1 reverted. never mind.
timotimo indy-args^ reverted
masak right.
timotimo i'm willing to bet that making that commit work will give another decent speedboost to get my timing below 100s 13:19
it's at 101s now 13:20
13:22 denisboyun joined
masak .oO( performance 101 ) 13:22
13:24 jnap joined 13:29 jnap left
lizmat patch.pm/p1 13:30
commute commencing&
13:31 lizmat left
TimK1 Has anyone built MoarVM on a Mac? 13:32
13:32 TimK1 left 13:33 TimK1 joined
tadzik I think diakopter does that 13:34
moritz TimK1: yes, I'm pretty sure, because the README mentions it explictly
(Or `nmake` on Windows). Currently it is known to build on Windows with MSVC,
with `gcc` and `clang` on Linux & MacOS X. We're expanding this with time.
13:34 hummeleB1 left 13:35 woolfy left
TimK1 Oh yeah… Look at that. I've been having trouble… But before I go any further, I'm going to try again, following again exactly the steps in the README. 13:39
13:40 kaleem left 13:42 hummeleBop1 joined 13:45 filly left 13:47 hummeleBop1 left
PerlJam frettled: :-PPP 13:55
timotimo a single !nqp::isnull made nqp build again 13:56
now let's see how much faster it is. 13:57
13:57 denisboyun left
FROGGS push push push! 13:58
13:59 kaare__ joined 14:00 labster_ joined 14:01 labster_ is now known as labster 14:04 denisboyun joined
masak p: say Date.today.day-of-week 14:04
camelia rakudo-parrot 262e60: OUTPUT«4␤»
masak p: say Date.today.month 14:05
I'm happy we decided not to make those two zero-based.
FROGGS yeah 14:06
hoelzro what is '$#array' in Perl 6?
FROGGS p: say 1 14:10
masak hoelzro: @array.end
moritz p: $#array
FROGGS it is dead :o(
masak I think I killed camelia :/
hoelzro ah, thanks masak
FROGGS .oO( You bastards! ) 14:11
hoelzro heh 14:12
14:12 camelia left 14:13 camelia joined
moritz p: $#array 14:14
14:15 ChanServ sets mode: +v camelia
camelia rakudo-parrot 262e60: OUTPUT«===SORRY!=== Error while compiling /tmp/yNoXnF3fBy␤Unsupported use of $#variable; in Perl 6 please use @variable.end␤at /tmp/yNoXnF3fBy:1␤------> $#array⏏<EOL>␤» 14:15
14:16 PacoAir joined
timotimo i needed some !nqp::isnull($node.node)s, but now it should work 14:18
FROGGS timotimo: there is another get_bool that explodes
timotimo some more*
FROGGS this one I think: nqp/src/vm/parrot/QAST/Compiler.nqp:727
compiling right now 14:19
timotimo i just added all of them :P 14:20
it doesn't make stage parse faster, it seems.
14:20 PacoAir left
jnthn timotimo: Oh? That's...surprising. 14:21
It's less memory pressure, either way...
14:21 PacoAir joined 14:22 denisboyun left
timotimo there seems to be more null pmc accesses in get_bool even after guarding all the $node.node's 14:23
14:24 denis_boyun joined
timotimo maybe the accessor methods should be tweaked to never return null instead 14:24
14:24 jlaire left
timotimo perhaps it's something about nqp::create that differs from self.CREATE? 14:25
14:26 rurban1 joined
timotimo i don't get it :\ 14:28
jnthn self.CREATE just calls nqp::crete(...) 14:29
I've no idea where the nulls are coming from, given the fact that things work on JVM...
timotimo oh well 14:31
dalek ecza: 989dffc | (Solomon Foster)++ | / (2 files):
Changes needed to make baghash.t work.
 14:32
14:36 ajr joined, ajr is now known as Guest7409 14:37 Guest7409 is now known as ajr_
timotimo jnthn: anyway, good work on that branch :) 14:40
it's a delight to see you just churn out worthwhile optimisations while i'm failing to come up with new good ones >_< 14:41
jnthn I'm happy they're helping :) 14:46
Reducing setting compilation time is important for our overall throughput. 14:47
timotimo i agree
jnthn And the benefits help suer code too
*user
14:47 bluescreen10 joined
jnthn Wow...storm is building here. 14:48
14:49 kaare__ is now known as kaare_
FROGGS .oO( The storm is strong in him, master! ) 14:49
jnthn I'll probably be able to get a train home, but there's more trains cancelled than there are leaving.. 14:50
14:52 hummeleB1 joined
Woodi Hi :) 14:53
Someone wrote: Your base class breaks encapsulation by exposing implementation details to subclasses in the form of protected members.
jnthn I think it's the subclass that'd be doing the breaking, but the base class is kinda flaunting its model... :) 14:54
Woodi is that some flavour of OO ? when you instantiate child class into object there is no parent class anymore! All attributess are object own, seeing it is using parrent kidney or brain is interesting vision... 14:56
just do not understans :)
after instantiation thinking in parrent/child class is not valid to me... 14:57
except reflection but things like refactoring are not part of design... 14:58
jnthn Woodi: The runtime operational view is little comfort to the maint programmer :) 15:00
atroxaper Hello #perl6 ! 15:02
Woodi yes, that's two different jobs/times :) but where is a breaking encapsulation ?
hi atroxaper :)
atroxaper Woodi hi!
Does anybody know why i can't generate html from specs's pod? 16-io.pod for example. q{Two terms in a row} in line 6 15:05
tadzik oh, they're POD, from Perl 5, not Pod from Perl 6
atroxaper I ran perl6 /path-to-pod/S16-io.pod --doc=html
Ulti r: say -Inf..Inf 15:06
camelia rakudo-parrot 262e60, rakudo-jvm 262e60: OUTPUT«-Inf..Inf␤»
Ulti r: my @infinity = -Inf..Inf 15:07
camelia ( no output )
Ulti is that a timeout?
jnthn Woodi: If I make a field private in a class, I can refactor as I choose. If it's protected, it's part of the API that I need to maintain for subclasses. If the subclasses are out of my control, then I need to maintain that API. If the API is in terms of fields, not methods, I've little flexibility in doing so.
atroxaper tadzik: Hm... Are all of Perl6 specs written in Perl5 POD?
jnthn Woodi: I'm not sure I'd claim that accessing protected fields is breaking encapsulation. I guess I'll simply argue that they're not encapsulated at all. :) 15:08
15:09 xenoterracide joined
FROGGS Ulti: no 15:09
r: my @infinity = -Inf..Inf; say "no"
camelia rakudo-parrot 262e60, rakudo-jvm 262e60: OUTPUT«no␤»
jnthn atroxaper: I think except S26
atroxaper tadzik: I want to improve pod2html generator. And found only one way to get Pod object: qq{eval slurp($pod_file.path) ~ "\n\$=pod";}
FROGGS r: my @infinity = -Inf..Inf; say @infinity[^10] # bang? 15:10
camelia rakudo-parrot 262e60, rakudo-jvm 262e60: OUTPUT«-Inf -9223372036854775808 -9223372036854775807 -9223372036854775806 -9223372036854775805 -9223372036854775804 -9223372036854775803 -9223372036854775802 -9223372036854775801 -9223372036854775800␤»
FROGGS ohh
Ulti hmmm why does my 2013.09 R* hang
jnthn wtf...
FROGGS Ulti: are you doing that in the repl?
Ulti heh I like -Infinity +1
dalek ecza: 425def6 | (Solomon Foster)++ | lib/CORE.setting:
Copy changes to Bag as well.
Ulti FROGGS: yeah
FROGGS Ulti: the repl tries evaluate it I think 15:11
Ulti: try the same in a -e
Ulti yeah I guess its always in sink context because it wants to print it
dalek ast: dafcd34 | (Solomon Foster)++ | S02-types/bag (2 files):
Niecza refudge.
FROGGS Ulti: yeah 15:12
Woodi encapsulation in the first place is for things not being in class hierarchy, it is essence of OO. but hidding thing from subclasses is another form of hidding and maybe it is necesarry too...
FROGGS p: my %x = (:a(1), :b(2) :c(3)) 15:13
camelia ( no output )
FROGGS p: my %x = (:a(1), :b(2) :c(3)); say %x
camelia rakudo-parrot 262e60: OUTPUT«("a" => 1, "b" => 2).hash␤»
FROGGS std: my %x = (:a(1), :b(2) :c(3)); say %x
camelia std 3b262af: OUTPUT«ok 00:01 128m␤»
tadzik atroxaper: yeah, I don't think we have a different way
FROGGS hmmm
is c just an adverb to b there?
15:14 jnap joined
atroxaper tadzik: and S26 don't parse too :) Said =begin wihtout =end ... 15:14
tadzik oh, S26 should definitely parse 15:15
the HTML rendering we have is a proof that we, at least at some point, were able to parse it
if we're not anymore, I'd be seriously mad :)
due to the time necessary to parse it we don't have it as a test anywhere tohugh 15:16
FROGGS .oO( My name is hugh )
tadzik :D
We are Hugh
FROGGS :o)
jnthn
.oO( Are hugh serious? )
FROGGS jnthn: you have no idea! 15:17
tadzik you can't say Hugh without hug
15:17 btyler joined
Woodi that API protecting is probably keeping Liskov in place, subclasses have some restrictions on parrent attibutes.. 15:17
atroxaper tadzik, it we're not anymore. Really. 15:18
tadzik I'm mad >:(
15:19 TimK1 left, V_S_C joined
FROGGS tadzik: seems like you *are* getting active this year :o) 15:19
tadzik KHAAAAN
moritz you can always BISECT
tadzik BISEEEECT 15:20
atroxaper BISECT!)) 15:21
tadzik and the last rendering of S26 doesn't have a timestamp
atroxaper Maybe rendering is fine. I can't even eval pod file.
moritz rendering uses eval :/ 15:23
Woodi btw. I learned a little of pre-assembler knowledge and it's kind a shock that x86 cpus have same instructions since '70s... looks 'backward uncompatibility' is just human error... 15:27
geekosaur similar instructions. how they're written has changed a bit though
but yes, you can trace many of the intel opcodes back to the 4004 if you put your mind to it 15:28
Woodi but looks more portable then HLLs :)
V_S_C x64 succeeded rather than IA64 as it overlapped with x86 to greater extent! Someone commercial minded said the market does not always make the right choice & played along. 15:31
atroxaper tadzik: i think i found! 15:34
15:34 colomon left
atroxaper tadzik: github.com/perl6/specs/commit/b7db...3b9975e76e it's comment with bag 15:35
commit
tadzik: need to add q{=end defn} in 605 line 15:36
Locally it works. 15:37
15:39 ajr_ left 15:43 p5eval left 15:44 ajr joined, ajr is now known as Guest92278 15:46 Guest92278 is now known as ajr_
tadzik ooh! 15:47
so it's not my code's fault \o/
should be =para defn or something
atroxaper+_+ 15:48
atroxaper++ even
atroxaper tadzik: thank you!
15:48 rurban1 left
atroxaper tadzik: i think it should be add =end defn in 605 line as in other examples around 15:50
jnthn attempted decommute :) & 15:51
FROGGS jnthn: good lok!
ohh, that was germish 15:52
jnthn: good lök! # :o)
err
15:56 FROGGS left 15:59 Psyche^ joined 16:00 Psyche^_ left
TimToady rosettacode.org/wiki/Sokoban#Perl_6 <-- passed up J, now C is 34 away :) 16:04
16:04 rurban1 joined 16:09 vaclavstrachon left
V_S_C @TimToady J? 34? 16:15
nwc10 TimToady: I forgot, and I couldn't find it by digging - which is the page with the leaderboard? 16:16
TimToady rosettacode.org/wiki/Rosetta_Code/R...popularity 16:20
though you have to run one of the programs to get the current results
I think they updated it a couple days ago
oh hey, just thought of a big advantage to turning the postcircumfix:<[ ]> method into a function; you can easily make strings be subscriptable in a lexical scope without inflicting those semantics on anyone else 16:27
16:31 dakkar left 16:33 spider-mario joined
TimToady nwc10: btw, the p6 one of those runs considerably faster under niecza, if you're wanting to run it 16:36
16:36 ajr_ left, ajr joined, ajr is now known as Guest98547, Guest98547 is now known as ajr_ 16:38 V_S_C left 16:41 thou joined 16:42 jnap left
jnthn got home without much trouble 16:47
TimToady it occurs to me that the Sokoban solver would parallelize nicely with a channel instead of a queue
16:50 tipdbmp left 16:53 jnap joined
TimToady we actually put the encapsulation boundary *inside* the class, not outside of it 16:54
when you say $!foo, you're inside, and when you say $.foo, you're outside 16:55
that might rise to the level of an advent idea
16:58 atroxaper left
Ulti TimToady unfortunately Perl6 isn't on this plot sogrady-media.redmonk.com/sogrady/f...g-0912.png 17:01
17:02 tipdbmp joined
TimToady well, of course not, that's measuring "real" popularity, not the insanity of a few RC contributors :) 17:05
masak: added a comment to yours, perl6advent.wordpress.com/2013/12/0...mment-3339 17:07
17:08 zakharyas joined 17:10 colomon joined 17:22 kaleem joined 17:27 sjohnson joined 17:28 sjohnson left 17:29 sjohnson joined 17:30 sjohnson left, sjohnson joined
TimToady decommuting for six hours or so & 17:31
17:38 bluescreen100 joined 17:39 kaleem left 17:42 bluescreen10 left 17:46 beastd joined 17:51 FROGGS[mobile] joined
colomon [Coke]: ping? 17:53
hoelzro ok, nobody fix this: rt.perl.org/Public/Bug/Display.html?id=120704 17:55
I'm going to see if I can do it
diakopter HERE, HAVE SOME STOP ENERGY. 17:56
<WINK> 17:57
hoelzro if I can't manage after a few days, please feel free to step in ;) 17:58
18:02 stevan_ joined 18:07 sftp left, sftp joined 18:08 Rotwang joined 18:10 telex left
jnthn r: my $a = (say 1; $a) = (say 2; $a) 18:10
camelia rakudo-parrot 262e60, rakudo-jvm 262e60: OUTPUT«1␤2␤»
jnthn hoelzro: That needs to be 2 1 I guess :)
I wonder if we're not handling childorder correctly somewhere... 18:11
Or something like that...
diakopter erm
that used to work
18:12 telex joined
diakopter r: my $a = (BEGIN say 1; $a) = (BEGIN say 2; $a) 18:13
camelia rakudo-parrot 262e60, rakudo-jvm 262e60: OUTPUT«1␤2␤»
diakopter mwahaha
FROGGS[mobile] you are crazy :o) 18:14
diakopter hrm
hoelzro the odd thing is how commenting odd the line in push fixes things. 18:15
18:20 sftp left, ajr_ left 18:23 sftp joined 18:29 kivutar left 18:30 xenoterracide left 18:34 Rotwang left 18:41 darutoko left
[Coke] colomon: you rang? 18:44
./scroll end
colomon: is it better than: 18:45
colomon [Coke]: how many test failures are you getting in Niecza these days? I'm getting a number of them with make spectest, but some (most?) of them appear to be some spurious problem with the make spectest harness rather than actual niecza errors.
[Coke] github.com/coke/perl6-roast-data/b....out#L6065 ? 18:46
colomon argh, that's far more errors than I'm getting, and not the same ones. :\ 18:47
what are you using to run the tests?
[Coke] that is the run before you fixed order.
colomon ah, maybe that helped 18:48
dalek kudo-star-daily: 90c115e | coke++ | log/ (5 files):
today (automated commit)
 18:49
rl6-roast-data: 4e9765a | coke++ | / (3 files):
today (automated commit)
[Coke] uh, that last commit looks wonky. 18:50
colomon: gist.github.com/coke/7811061 18:52
that's from today's run.
all those were already failing before your fix.
colomon \o/
[Coke] can you see if the test file I skipped in t/spectest.data recently passes with your fix? (hard for me to commit from $dayjob) 18:53
colomon++
(down to 56 failures in niecza, from 468) 18:54
colomon I also made some tests that had never worked before in niecza work. ;)
[Coke]: yes, cmp.t works now. I'll turn it back on. 18:58
[Coke] colomon++ thanks 18:59
dalek ecza: 17d73e4 | (Solomon Foster)++ | t/spectest.data:
Turn S03-operators/cmp.t back on.
19:01 raiph joined
dalek ast: c8d9486 | coke++ | S03-metaops/ (2 files):
Partially Revert "partial niecza fudge"

This partially reverts commit 6312c187880ad4e8da1cde7df665f6f17ad9f501.		 19:05
[Coke] I think that's the last of the ones I fudged the other day.
(run the tests) this script: 19:06
dalek kudo/moar-support: 5c6e780 | jnthn++ | src/vm/moar/ops/perl6_ops.c:
Cheating p6stateinit.

Hopefully we can actually get to Hello world! without implementing state vars.
kudo/moar-support: 087641c | jnthn++ | src/vm/moar/ops/perl6_ops.c:
Implement p6staticouter.
[Coke] github.com/coke/perl6-roast-data/b.../niecza.sh
colomon^^ 19:11
19:12 fhelmberger_ joined
colomon [Coke]++ # keeping on top of things 19:12
19:13 fhelmberger left 19:14 jlaire joined, cognominal joined, cognominal__ left
FROGGS[mobile] jnthn++ 19:15
jnthn Those don't solve the nasty closure issue, though...
FROGGS[mobile] is the closure issue "my" gc bug? 19:16
jnthn No
FROGGS[mobile] damn
19:16 fhelmberger_ left
jnthn Ugh...hope the power holds with this storm... 19:17
FROGGS[mobile] :/ 19:18
no storm here anymore
jnthn I'm sure I had this problem when doing the JVM port... 19:21
jnthn looks thorugh the commit history
19:22 ajr joined 19:23 ajr is now known as Guest91500, Guest91500 is now known as ajr_
jnthn finds 4b6519aba in the NQP repo and thinks that may be it. :) 19:32
.oO( Clearly, I was smarter back in May... )
19:36 xinming left 19:37 xinming joined 20:05 rindolf left 20:11 FROGGS joined 20:13 spider-mario left 20:19 cooper left 20:25 peekxc joined
peekxc Hello All! I have a question related to Perl 6.0 (or Rakudo if you want) vs the current stable release, Perl 5.18; specifically in regards to the basics of opening filehandles. Is this the right channel and if so can anyone help me? 20:26
Juerd Yes.
diakopter japhb: fix the internet, please. kthxbai 20:27
Juerd I will probably not be able to help but you've come to the right place for Perl 6 questions.
japhb diakopter, Did I break it again? Dang.
diakopter japhb: my gmail says 502. make it stop.
20:28 SevenWolf joined
diakopter japhb: [I'm sure you love this sort of harassment.. :S] 20:28
"Please try again in 30 seconds. That’s all we know. " That's not a very strong statement of knowledge... 20:29
japhb It's my favorite. ;-)
peekxc Alright, so I've done a fair bit of research and I've looked at the documentation, I'm learning Perl now from the Randal Schwartz book "Learning Pearl." They go over many different ways to open and use filehandles, however most of what they do seems to be use the syntax for 6.0, which is the triple term setup (open $file_fh, '<', 'some_location';)
jnthn Apart from that's not the syntax for 6 at all :)
my $file_fh = open('some_location', :r); # how that'd look in 6 20:30
japhb "SORREEE, WE BROAK DA NETZ. WE IZ WOORKING ON ITZ."
20:30 cooper joined
peekxc reallreally? 20:30
hmmmmm
diakopter peekxc: which tutorial are you reading for 6.0?
peekxc "Learning Perl" by Randall Schwartz
in Chapter 5 he says, and I quote: 20:31
PerlJam peekxc: That book is for Perl 5
diakopter peekxc: that's not for Perl 6
SevenWolf Learning Perl is Perl 5 and you are in a Perl 6 Chat room
peekxc Ahh, I see
PerlJam There is no "Learning Perl 6" yet
peekxc I misread 5.6 for Perl 6
PerlJam peekxc: and if it's for 5.6, it's still horribly out of date :) 20:32
SevenWolf well basics are still valid
ingy arvo! o/
peekxc Am I correct that the latest "stable" release is 5.18.1 though?
diakopter peekxc: yes, but Perl 6 is on a completely different trajectory from Perl 5 20:33
PerlJam SevenWolf: sure, but a book written for a 13 year old version of Perl is likely to miss a few things :)
BinGOs v5.18.2 will be out in a week and a bit.
peekxc As I've come to understand
SevenWolf hey peekxc if your interest is in Perl 5 channel #perl may be more suited
20:33 kivutar joined
SevenWolf I agree PerlJam 20:33
peekxc Thank you Seven, but before I depart, let me ask you all a quick question
PerlJam waits for a crazy non sequitur 20:34
diakopter japhb: mail plz. ok, thanks, awesome.
peekxc If you were developing a program that was going to be used over a very very large database, cross-platform, and stability was key, would you use 5.18 or 6.0 ?
japhb diakopter, :-D
SevenWolf two languages depends on your preference
peekxc I know 6.0 cbumps up the efficiency quite a bit, but my concerns lie in stability
diakopter peekxc: version 5, absolutely
peekxc Thanks, dia, and Seven 20:35
japhb (Once in a while the sleeping deities awake, jiggle a fiber coupling, and then fall back to sleep.)
peekxc I'll be taking my leave then, thanks all for the help
btyler peekxc: the naming scheme is a little weird, but perl 5 and perl 6 are different languages. for something like that a recent version of perl 5 is a great choice
diakopter peekxc: 6.0 bumps the efficiency up a few thousand times in the opposite direction
peekxc What do you mean "in the opposite direction"? 20:36
diakopter it's not up
peekxc as in it's less efficient?
btyler perl 6 in its current form is still rather slower than perl 5 in most (all?) cases
PerlJam diakopter: still depends on what efficiency you're talking about. :)
diakopter I'm curious who told you it was more efficient (runtime performance, startup performance)
SevenWolf I enjoy Perl 5 but as soon as 6 is out there I may switch.. I will switch.. I think I will miss Sigil Variance
PerlJam Perl 6 is quite a bit more expressive than Perl 5.
peekxc wikipedia ;)
I see 20:37
SevenWolf I've also heard that as of right now Perl6 isn't as fast as Perl 5
jnthn Perl 6 as a language should be more optimizable in the long run. At the moment the implementations have much optimization to go.
peekxc Well this is all very useful info to me, as I'm the one needing to make the decision. Thank you.
japhb peekxc, Right now, Perl 6 compilers are generally slower than Perl 5, sometimes a lot slower. The *design* of the language is made to ramp up significantly in efficiency, but the *implementations* are still in the "get everything working" phase, instead of the "make it blazingly fast" phase.
PerlJam SevenWolf: btw, what do you mean by "out there"? Perl 6 is "out" already. Has been for a while. What are your criteria to switch? 20:40
SevenWolf hmm well I actually am more of a Perl 5 guy, Im very interested in Perl 6 20:41
and from what I've read, Perl 6 has third part compilers?
party*
20:41 peekxc left
SevenWolf Is Rakudo the official implementation? 20:42
FROGGS SevenWolf: not really third party, but there is not a single implementation as there is for Perl 5
perigrin The party is just larger so in between having punch they throw out compilers.
SevenWolf lol
FROGGS SevenWolf: all implementations are official which pass the test suite 20:43
japhb SevenWolf, Rakudo is the most active implementation in terms of development effort.
FROGGS yes, that makes rakudo the leading implementation
SevenWolf is implementation of threading complete with any of the p6 compilers?
japhb Complete, no. Working nicely? Well, I like the rakudo-jvm threading. 20:44
FROGGS SevenWolf: the only implementation that support threads so far is rakudo on the jvm
japhb Concurrency, really.
FROGGS not complete, since the the implementation (and spec) of threads just has started
perigrin you keep using these terms like "complete" and "official" ... hang out a while and you'll realize why they're not really applicable.
japhb (Scheduled tasks in a thread pool)
FROGGS SevenWolf: think more in terms of "useful" and "appropriate" 20:45
:o)
japhb SevenWolf, see jnthn++'s talks/slide decks on Concurrency in Perl 6.
FROGGS (... for the given task)
or the advent calender
SevenWolf I see, I'll check that out japhb 20:46
PerlJam perigrin, FROGGS: the transition from "complete/official" to "useful/appropriate" might make a good advent post ;)
20:46 kaare_ left
japhb (newest is best WRT those talks, by the way; there was a syntax change recently) 20:46
jnthn jnthn.net/papers/2013-npw-conc.pdf is the current best bet.
FROGGS SevenWolf: btw, if you want to dig into Perl 6 a perfect thing to do would be to port a Perl 5 module you are interested in to Perl 6
20:47 cooper left
FROGGS PerlJam: well volunteered! :P 20:47
SevenWolf I have Rakudo installed on linux, I most likely will.
PerlJam FROGGS: There are a few slots open still, so I may write one or two, but I haven't felt up to it yet and I'm not sure I will before xmas. 20:48
FROGGS I see
I plan to to start on my first post tomorrow 20:49
japhb SevenWolf, you say "I have Rakudo installed on linux". From experience, sometimes that means people used the system package manager to install it. If so, that's almost certainly seriously out of date if you're interested in concurrency and threading. 20:51
Best to build from git. (The procedure for which has also changed/improved lately.)
SevenWolf ah yes, that is the case somewhat.. except I have no package manager, I just sudo apt-get install ...
Im on kali 20:52
PerlJam apt is a package manager last time I checked :)
SevenWolf well
what i meant was no graphical
my mistake
FROGGS that doesnt matter :o)
it is just that the (debian guys)
it is just that the (debian guys)++ dont bundle it every month 20:53
SevenWolf I am aware of out of date repositories, honestly i rely on them far too much lol
20:54 cooper joined
FROGGS it is just dramatic what happens here within a month, and if you then see that the packaged one is a year old... well :o) 20:55
japhb Yeah, it's night and day. 20:57
SevenWolf Candidate: 0.1~2012.01-1 20:58
O_o im in trouble
lol
FROGGS *g* 20:59
you are, indeed :o)
japhb tries to imagine how different that actually is. 21:00
A lot, let's understate it that way. 21:01
SevenWolf lol
building from source now
21:01 zakharyas left, zakharyas1 joined
FROGGS SevenWolf++ 21:02
dalek p/indy-args: 31d3bce | jnthn++ | src/vm/moar/QAST/QASTCompilerMAST.nqp:
Fix lexical code-gen/dyn-comp interaction.
 21:07
jnthn bah, wrong branch, but still... 21:08
FROGGS you can cherry-pick it, no? 21:09
jnthn Or just merge the branch soon :P
FROGGS merge soon merge often :o) 21:10
jnthn Yeah. I heard that my last opt in it busts Parrot build, though...didn't get to look yet
21:10 cognominal left, cognominal joined
FROGGS a pro pos merge soon... do you have time to look at the labels branch these days? 21:11
btw, I want to squish the sized-arrays branch into a single commit so it is easier to review (for you)
jnthn +1 to squish to a single commit to review 21:15
Labels branch...what's current status?
FROGGS k, now I have to lean how to do that :o)
jnthn git rebase -i, hack the text file
Labels branch...what's current status?
FROGGS jnthn: there is a nqp branch that adds support for labels for all loop for nqp code
for parrot backend
21:17 dayangkun left 21:23 dayangkun joined 21:26 bbkr joined 21:27 jeffreykegler joined 21:28 dayangkun left 21:29 dayangkun joined
jeffreykegler Hi. I'm here to borrow ideas for Marpa (my parser project), again. 21:31
Currently, Marpa avoids the use of strings by making everything an integer, but it may add upper layers.
Is there a doc on Perl6's use/interpretation of Unicode? 21:32
PerlJam jeffreykegler: care to write an advent article about what you've borrowed from Perl6? :)
jeffreykegler Things like normalization, equality testing, etc.?
Perljam: my own blog has been neglected for a month now 21:33
Perljam: sorry 21:34
moritz S02 contains some bits about Unicode 21:35
jeffreykegler Marpa's upper layer is now Perl5, but exactly imitating its interpretation of Unicode does not seem like a good idea 21:36
In particular, Perl 5 recognizes integers well outside the official Unicode range, which is great if you want to use Unicode ... 21:37
... as another means of encoding large integers ...
21:38 daverccr joined
jeffreykegler ... but perhaps not so great in terms of standard conformance. 21:38
I found perlcabal.org/syn/S02.html#Unicode_Semantics 21:41
timotimo what did i miss today? 21:44
21:45 stevan_ left
raydiak jeffreykegler: the first couple paragraphs of S32 might also be useful to you 21:45
jnthn timotimo: A *huge* storm. :P 21:46
And I did a couple of small patches. 21:47
:)
timotimo sounds good :)
still no idea about the nulls?
jeffreykegler raydiak: github.com/perl6/specs/blob/master...ry/Str.pod -- the stuff about Str? 21:49
21:50 BenGoldberg joined
jeffreykegler Looking at stuff like this, I'm beginning to think of the EBCDIC vs. ASCII war as an age of peace and tranquility. :-) 21:52
raydiak jeffreykegler: yes, sorry - S32 Str is what I meant :) specifically the first couple paragraphs where it talks about different unicode levels 21:55
21:55 TimK1 joined, daverccr left
raydiak though you may also get some answers by looking at the functions there, too... 21:57
lue is reminded to blog about his provisional S15... 21:58
dalek kudo/sized-arrays: b227e62 | jnthn++ | src/core/ListIter.pm:
Allocate more sensibly in ListIter.
kudo/sized-arrays: 262e600 | (Elizabeth Mattijsen)++ | src/core/Mu.pm:
Properly DEPRECATE bless(*) calls
kudo/sized-arrays: 621739b | (Tobias Leich)++ | src/core/ (2 files):
Merge branch 'nom' of github.com:rakudo/rakudo into sized-arrays
lue FROGGS++ : for a second I had no clue who made that push :) . 22:01
FROGGS lue: I just try to rebase that branch (with no luk) 22:02
luck*
lue
.oO(luck -> luk -> luek -> lük)
jnthn Lök at that onion! 22:04
diakopter use the fours, luk 22:05
FROGGS jnthn: btw, that "good lok" was meant as a pun, because Lok eq train in german
jnthn Oh!
Yeah, same here. :) I'm not sure why I thought of onions...
FROGGS same here? my dict said something about onions too :o) 22:06
22:06 tgt joined
jnthn ooh 22:07
22:08 xenoterracide joined, zakharyas1 left
timotimo i'm pretty tired now, but maybe tomorrow i'll get knee-deep into debugging the jvm interop dispatcher generator stuff 22:10
jnthn :) 22:11
jnthn wonders if he'll sleep tonight with all this wind...
jeffreykegler Re Perl6 & Unicode: S02 refers to "NFG". Searching the web, I cannot find references to NFG outside of the Parrot/Perl 6 world. 22:13
PerlJam jeffreykegler: aye, that's because it originates with Perl 6. 22:14
well ... s/Perl 6/TimToady/ probably 22:15
lue jeffreykegler: It's basically NFC, with non-Unicode-standard negative numbers assigned to graphemes without precomposed characters. 22:16
timotimo www.amazon.de/review/R2X1LSQAG1SAPF...ag=mflp-21
er, ww 22:17
jeffreykegler lue: Thanks
lue you're welcome :)
PerlJam wonders if what lue said is the same thing as "leaves no uncomposed composing characters" 22:19
dalek kudo/moar-support: c462531 | jnthn++ | src/core/ (3 files):
Skip remaining few things that break setting load.

With this, we can now do "Hello, world" with Rakudo on MoarVM; though it spits out a couple of warnings for the moment too.
timotimo aaaaw yiss :3 22:20
m-BOOTSTRAP still gcbugs locally
FROGGS timotimo: --optimize=1 ?
FROGGS tries 22:21
lue PerlJam: perhaps, though it likely implies that Unicode has those precompositions, instead of Perl 6 having to make it up.
(Critical in that is that we'll never be able to look at an NFG string directly.)
22:24 zakharyas joined
jeffreykegler Thanks. What I think I was looking for was a relatively, simple pat choice among the choices in the Unicode standards ... 22:27
... which I could just borrow from Perl 6. Any complaints, I could then say,
... "Hey look, this is the way Larry does it."
... but it looks like I am not going to be able to get away with anything like that. 22:28
jnthn timotimo: Need indy-args branch fwiw 22:31
I'll get that merged soon but not tonight.
22:33 jlaire left 22:35 dmol left 22:37 jlaire joined 22:39 bluescreen100 left
raiph masak, jnthn, of interest to edument? exercism.io/ # "Exercises with crowd-sourced code reviews" 22:47
FROGGS[mobile] that storm is back and it is just crazy 22:49
22:50 FROGGS left 22:51 stevan_ joined
jnthn raiph: Hm, interseting idea... 22:52
FROGGS[mobile]: It comes and goes here :) 22:53
23:03 xenoterracide left 23:11 kivutar left 23:25 thou left 23:32 stevan_ left 23:45 woolfy joined 23:46 denis_boyun left 23:50 woolfy left 23:54 dmol joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!