»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
00:17 dukeleto left, dukeleto joined 00:23 wamba left 00:25 araujo left 00:26 araujo joined, zorgnax left
dalek ecza: 4508d7b | sorear++ | lib/ (2 files):
Start building RuntimeUnit objects
00:35
ecza: 04df192 | sorear++ | lib/ (2 files):
Serialized representation, not contstructors, for declarative prefixes
sorear that change alone brings STD.BOOT down from 477378 to 408958 (bytes of IL) 00:38
dalek ecza: 183a42a | sorear++ | lib/ (2 files):
Use a blob for the SubInfo constructor
01:21
01:54 donri left 01:58 kst joined 02:01 _jaldhar left, _jaldhar joined
sorear 294695 02:04
02:05 _jaldhar left, _jaldhar joined
dalek ecza: 75a1ba9 | sorear++ | lib/ (2 files):
Encode signatures as blobs
02:31
sorear 281777 (despite code to set up index->metaobject mapping)
02:35 huf left 02:51 yinyin joined 02:58 PZt left 03:03 lue left 03:08 am0c left 03:09 kfo joined 03:12 kfo_ left
dalek ecza: d39d163 | sorear++ | src/NieczaPass (2 files):
Avoid generating symbols for has methods in void context
03:17
03:25 am0c joined 03:27 PZt joined 03:32 agentzh joined 03:49 araujo left 03:53 dukeleto left, dukeleto joined 03:57 colomon left
dalek ecza: 583d03b | sorear++ | lib/ (2 files):
Eliminate top-level protosub fields, they do not pull their weight
04:10
ecza: 7febed1 | sorear++ | lib/CLRBackend.cs:
Coalesce identical constants
04:17 satyavvd joined 04:34 pernatiy_ left
dalek ecza: 481e8ca | sorear++ | lib/ (2 files):
Use a denser coding of strings, also coalesce them
04:45
04:51 _jaldhar left, _jaldhar joined 04:53 jaldhar_ joined 04:54 _jaldhar left 04:56 jaldhar_ left 04:57 jaldhar_ joined 05:08 _jaldhar_ joined, jaldhar_ left 05:09 fhelmberger joined 05:16 yinyin left 05:22 birdwindupbird joined 05:23 _jaldhar_ left 05:24 _jaldhar_ joined 05:29 koban joined 05:35 pamera joined 05:36 Eevee left 05:38 Eevee joined
moritz tadzik++ # gsoc blog post 05:43
05:50 wtw joined 05:53 jaldhar joined, _jaldhar_ left 05:57 _jaldhar joined 05:58 jaldhar left 06:03 pernatiy joined
dalek ecza: 280737c | sorear++ | lib/ (2 files):
Generate class member info as blobs
06:07
ecza: 77f924c | sorear++ | lib/ (2 files):
Move more of the sub2 logic into LoadSubInfo
06:08 _jaldhar left 06:10 amkrankruleuen left, _jaldhar joined 06:13 kaare_ joined 06:15 amkrankruleuen joined 06:19 _jaldhar left 06:21 _jaldhar joined 06:30 wamba joined 06:46 mj41 joined 06:51 amkrankruleuen left 06:52 amkrankruleuen joined 07:00 amkrankruleuen left 07:01 amkrankruleuen joined
sorear I wish zip -r wouldn't include vim swap files 07:07
nice, niecza --help goes from 5.1s to 2.5s after rebooting it with the current improvements 07:09
benabik sorear: That's why I like git-archive and generally include a Makefile recipe to include any build products I need.
sorear well, I'm making a zip of generated files mostly
sorear wonders if it still makes sense to use zip 07:10
benabik sorear: Ah. Less useful then, although that is the reason I use git-archive. :-D
07:22 Mowah joined 07:27 amkrankruleuen left 07:28 amkrankruleuen joined 07:35 SHODAN joined 07:36 timbunce joined, timbunce left
dalek ecza: f3e0436 | sorear++ | / (3 files):
Simplify handling of alternation metadata
07:41
07:41 amkrankruleuen left 07:44 amkrankruleuen joined 08:05 domidumont joined 08:16 huf joined 08:22 dakkar joined
dalek ecza: ea4bffe | sorear++ | lib/ (2 files):
First foray into a true constant pool
08:31
ecza: ad0859a | sorear++ | lib/ (2 files):
Extend pooling to CC[] and AltInfo
08:32 pamera left 08:33 cognominal left 08:40 amkrankruleuen left 08:46 amkrankruleuen joined 08:49 Jon_Doe joined, daniel-s joined
Jon_Doe perl6: say 3 08:53
p6eval pugs, rakudo 30ebf7, niecza v5-56-gf3e0436: OUTPUT«3␤»
Jon_Doe lolwut.
08:54 Jon_Doe left
tadzik pff 08:54
08:57 icwiener joined 09:11 nymacro joined
dalek ecza: 60b2044 | sorear++ | lib/ (2 files):
And now user variables get pooling too
09:18
ecza: 50b1205 | sorear++ | lib/ (2 files):
Finish converting all types of constants to the pool system
sorear out 09:25
STD.dll is under 2MB now \o/
09:37 PacoLinux left 09:41 tzhs joined 09:46 amkrankruleuen left 09:47 amkrankruleuen joined 09:49 amkrankruleuen left, amkrankruleuen joined 09:54 tzhs left 09:59 wamba left 10:02 noganex left 10:03 noganex joined 10:05 amkrankruleuen left 10:07 wamba joined 10:09 Moukeddar joined
Moukeddar hello Perl9 10:09
6*
10:10 Moukeddar left 10:11 amkrankruleuen joined 10:12 pernatiy left
jdhore1 Can I ask a really quick P5 question? 10:12
daniel-s jdhore1, well, there is #perl 10:14
but also
this place seems kinda quiet, so I don't think there's anyone to tell you off, even if you were doing something wrong by asking
jdhore1 Well...what's the best way to check if something isn't undef? (i'm having a brainfart)... if ($variable) {} ? 10:15
huf defined 10:16
if (defined $foo) { ... 10:17
10:17 amkrankruleuen left
jdhore1 huf, for Perl 5? 10:17
huf yes, isnt that what you were asking? 10:18
jdhore1 Yeah, I was just making sure
daniel-s in perl6, would it be if $variable.defined ?
cos now it's got a method
10:22 amkrankruleuen joined
jdhore1 Thanks guys 10:28
It's 6:30AM (nearly) and i haven't been to bead yet so i'm a bit tired
10:33 snearch joined 10:35 daniel-s left 10:37 amkrankruleuen left 10:48 amkrankruleuen joined 10:57 awoodland joined 11:02 Psyche^ joined 11:06 Patterner left, Psyche^ is now known as Patterner 11:08 agentzh left 11:20 MayDaniel joined 11:24 jfried joined 11:26 donri joined 11:31 pernatiy joined 11:39 MayDaniel left 11:41 tzhs joined, SHODAN left 11:42 donri left 11:43 pamera joined 11:45 mtk left, tzhs left 11:47 amkrankruleuen left 11:48 amkrankruleuen joined, yves_ left 11:52 wamba left 11:53 mtk joined 12:00 yves joined 12:15 snearch left, awoodland left 12:17 bacek left 12:19 colomon joined 12:22 bluescreen10 joined 12:23 satyavvd left 12:37 hanekomu joined 12:44 pernatiy left 12:47 SHODAN joined 12:53 amkrankruleuen left 12:54 Holy_Cow joined 12:55 Holy_Cow left, wamba joined 12:56 amkrankruleuen joined 13:02 donri joined 13:03 wamba left 13:05 daniel-s joined 13:08 amkrankruleuen left 13:13 amkrankruleuen joined 13:41 hercynium left 13:52 pernatiy joined
sorear daniel-s: the sub form still exists (and is substantially faster on niecza) 14:01
good * #perl6
14:05 PacoLinux joined, wknight8111 joined
sorear hello jdhore1. 14:19
14:20 satyavvd joined 14:22 domidumont left 14:23 hercynium joined, hercynium left, hercynium joined 14:25 wtw left 14:26 nymacro left 14:27 jevin left 14:30 jrockway left
dalek ecza: 69332cd | sorear++ | lib/CORE.setting:
Fix * smartmatch behavior
14:30
14:31 jrockway joined 14:32 jevin joined 14:33 alester joined 14:45 jerome left 14:49 jerome joined 14:51 jrockway left 14:52 jevin left 14:53 jrockway joined, jevin joined 14:54 am0c left 14:58 thou joined 15:00 koban left, SHODAN left 15:01 araujo joined 15:03 daniel-s left, domidumont joined 15:04 noganex_ joined, noganex left 15:07 am0c joined 15:09 huf left 15:10 thou left 15:11 orafu left, ymasory joined 15:12 orafu joined 15:24 wamba joined 15:28 rindolf joined
rindolf Hi all. 15:28
colomon o/ 15:29
15:29 cognominal_ joined 15:32 Sarten-X joined 15:46 kjeldahl joined 15:51 kensanata joined
jnthn evening, #perl6 16:01
dalek ecs: 74b208c | larry++ | S06-routines.pod:
don't include positional names in API

As suggested by masak++ and jnthn++, remove the names of positionals from the public API.
In its place, substitute the more useful ability to treat positionals as if they were named by the use of *numeric* keys. That is, move the ability to name positionals out of the binder and into the coercion from hash to capture. This answers the three objections to named positionals:
   * The positions of the positionals are already part of the public API.
   * This has no run-time overhead (unless you force this coercion to happen
at run time, using prefix:<|> for instance).
   * People will use this because many external formats do not support captures.
16:02
16:03 thou joined
jnthn \o/ 16:04
(walks by the neva)++
16:05 ymasory left
rindolf Hi jnthn 16:05
16:05 satyavvd left
jnthn o/ rindolf 16:05
rindolf jnthn: what's up? 16:08
jnthn rindolf: Not much...just getting back to dayjob after a vacation in Russia. :) 16:10
Well, vacation and YAPC. :)
rindolf jnthn: ah.
jnthn: YAPC?
jnthn YAPC::Russia
16:21 jfried left 16:28 birdwindupbird left 16:32 cdarroch joined, cdarroch left, cdarroch joined 16:34 satyavvd joined
colomon woah, that's a biggish change, no? 16:37
74b208c3d7dc6bc3fb913be3376492e2d72733f6, I mean
if nothing else, I foresee a good number of spectests needing to be fixed 16:38
16:46 satyavvd left
TimToady yes, it was up there on the Richter scale 16:48
but at this point in Perl 6's life, it's always better to fix something sooner rather than later 16:49
colomon Still, TimToady++ -- both for seeing it was the right thing to do and being willing to take the pain of doing it.
16:49 wamba left
colomon Though I'm doubting you will handle fixing all the spectests yourself. ;) 16:52
16:57 mj41 left
TimToady has learned to delegate anything that I can cajole other people into doing as well as I would do it... 17:00
17:00 pernatiy left 17:01 rindolf left
colomon jnthn: how hard would it be to implement that change in current Rakudo? (Maybe as a branch?) A lot easier to fix spectests with a "working" compiler to test them on. 17:03
17:06 sftp left 17:07 dakkar left 17:08 wamba joined 17:09 hanekomu left 17:11 kensanata left 17:16 wamba left, snearch joined, wamba joined 17:36 am0c left
dalek kudo/nom: 8cdd3e9 | jnthn++ | src/ (3 files):
Remove named to positional paramter binding.
17:38
colomon jnthn++ 17:41
17:45 Eevee left 17:46 sftp joined 17:49 am0c joined
moritz oh hai 17:57
moritz feels he missed something big-ish :-)
colomon you can't refer to positional arguments by name anymore. (external to the function, that is) 17:58
(NYI ... er ... NY unimplemented? in Rakudo, natch.) 17:59
18:03 Eevee joined, Chillance joined
dalek ast: 87efb8a | moritz++ | S32-array/ (10 files):
[S32-array] unfossile tests. Most of that is calling positional arguments by name
18:10
18:17 awoodland joined, arnsholt left, arnsholt_ joined 18:18 fhelmberger left 18:19 mberends joined
tadzik neh, a test passes on gsocmess, fails on Rakudo 18:19
I'm one spectest from gsoc-podparser passing the spectests :) 18:20
mberends tadzik: nice :)
tadzik mberends: have you seen the latest blag toast? 18:21
rakudo: say '=' ~~ /\w/
mberends tadzik: yes, where you almost finished weeks 1 and 2 during week 0 ;)
p6eval rakudo 30ebf7: OUTPUT«␤»
tadzik mberends: I was close to being a hipster cat: memegenerator.net/instance/7912124 ;) 18:22
18:22 ymasory joined 18:23 PacoLinux left
tadzik I still have some dunnos though, turns out that $=POD variable, scheduled for some deep august I think, needs to be available now, or some substitute of it 18:23
18:23 PacoLinux joined
mberends tadzik: that site looks like potentially too much information... 18:23
tadzik :P 18:26
that's the weird region of the internet 18:27
18:27 orafu left 18:35 kaare_ left, wamba left
dalek ast: c5fa27b | moritz++ | S32- (13 files):
[S32] remove more tests for now-gone calling-positionals-by-named tests
18:39
ast: 19614fb | moritz++ | S32- (14 files):
[S32] unfossile more tests
sorear colomon: 'deimplemented' would be good 18:40
moritz fears his internet will expire in 5 minutes :/
sorear that's about how soon I'm leaving for my first final 18:41
moritz sorear: I wish you the best of luck
tadzik sorear: what's that one from?
mberends good luck sorear 18:42
18:42 wamba joined
sorear colomon: or maybe 'disimplemented' 18:43
dalek ast: 85d2267 | moritz++ | S32-str/ (9 files):
[S32] more unfossiling
sorear thanks all 18:44
colomon sorear: so NYD either way? :)
sorear perl6: sub foo($x) { say $x.perl }; foo(5 => "bar")
p6eval rakudo 30ebf7, niecza v5-61-g69332cd: OUTPUT«5 => "bar"␤»
..pugs: OUTPUT«(5 => "bar")␤»
sorear TimToady: is foo("0" => "foo") magic now? 18:45
18:46 saaki left
TimToady I suspect only 0 => is, if the "" overrides the naminess 18:46
sorear TimToady: you just broke the old system for recognizing magic pairs (S06:713)
TimToady only specced string dwim for the run-time coercion
sorear TimToady: is \d+ '=>' <term> considered an "autoquoting" pair now? 18:48
moritz oh wow, the hotel wifi had a countdown from 24h... and then it was at 0, and disconnected
dalek ecs: 0140a6f | larry++ | S06-routines.pod:
fix 0 => $arg naminess
moritz ... and then I could re-ologin with the exact same login data :-)
tadzik :) 18:49
sorear TimToady: clarification needed: is it legal to mix types, like foo($x, 0 => $y) ?
moritz colomon: there are a *lot*
colomon: ... of tring spectest with :base etc.
s/tring/trig/
colomon moritz: oh!
let me take care of those.
tadzik yay, the spectests will run faster!
TimToady sorear: that would be a positional collision, so should complain
moritz oh wait, maybe :base is always named 18:50
TimToady (I suspect)
moritz ... or should be named?
dunno
sorear TimToady: in the old system, named arguments bump positionals, like foo($y,$x)
colomon moritz: $base is positional, but ... making it named might make sense.
tadzik rakudo: say " ==> and <==" ~~ / \h* <!before '=' \w> \N+ \n /
TimToady could argue it the other way too...
p6eval rakudo 30ebf7: OUTPUT«␤»
tadzik ...oh?
18:50 mj41 joined
moritz tadzik: no \n in the string 18:51
tadzik oh sure
rakudo: say " ==> and <==\n" ~~ / \h* <!before '=' \w> \N+ \n /
p6eval rakudo 30ebf7: OUTPUT« ==> and <==␤␤»
tadzik right
hrm, exactly the same piece of Pod works in gsocmess repo, fails when I moved the grammar into rakudo and ran the specetsts
bleh, spec-tests
sorear TimToady: when would the syntax foo(1 => ..., 0 => ...) be useful in source code? 18:52
moritz nqp-rx: say(" ==> and <==\n" ~~ / \h* <!before '=' \w> \N+ \n/);
TimToady sorear: for when a company establishes a stupid guideline that requires it :)
moritz nqprx: say(" ==> and <==\n" ~~ / \h* <!before '=' \w> \N+ \n/);
p6eval nqprx: OUTPUT« ==> and <==␤␤»
TimToady or for when you really want to say one of the named arguments first 18:53
for endweight reasons, say
tadzik phenny: "sztuka dla sztuki"?
phenny tadzik: "art for art's sake" (pl to en, translate.google.com)
colomon moritz: if we no longer have named positionals, is there any point in those tests anymore? those tests are essentially testing that the name is correct, if the non-named tests work then obviously explicit positional tests will work (or the compiler is broken).
tadzik googletranslate++
sorear +1 to remove 18:54
TimToady: ok
TimToady or for a foolish consistency when there are lots of other named args
sorear TimToady: does Perl 6 require left to right evaluation?
i.e. can foo(1 => bar(), 0 => baz()) be converted into foo(baz(), bar()); ? 18:55
TimToady I don't think arg separator comma enforces any order
though S, might :) 18:57
sorear does that extend fo foo() + bar() \equiv infix:<+>(foo(), bar()) \equiv bar() R+ foo() ?
S still seems pretty handwavey to me. Do not grok. 18:58
oops
I really have to go
TimToady o/
colomon oh noez, the Soperatives are on to sorear++ 19:00
19:00 uniejo joined
TimToady but in general it seems best to me not to spec order of evaluation except for operators that explicitly promise such 19:00
19:01 uniejo left
TimToady that, and the implicit ordering implied by laziness, but that doesn't mean the "shell" of the lazy list is unevaluated 19:02
colomon TimToady: doesn't that run the risk of making Perl 6 programs evaluate in different ways depending on the compiler you're running it on?
TimToady that's always the case, if you rely on unspecced semantics :) 19:03
colomon
.oO( the agony of side effects )
TimToady so write in an FP style and you won't have this problem so much :P
or make sure of your sequence points if you're relying on non-idempotence 19:05
maybe we can warn someday about side effects that are forced share a void context 19:07
*sink
19:08 mberends left
colomon Do you use S, to make sure of sequence points? 19:08
TimToady waves hands in colomon's general direction 19:09
19:09 mberends joined
colomon :p 19:09
TimToady but that's one of the (possibly related) imagined uses of S
19:23 mkramer1 joined 19:24 [Coke] left 19:25 [Coke] joined 19:28 mkramer joined 19:31 mkramer1 left 19:33 [Coke] left 19:35 [Coke] joined
pmichaud quick note: commit 0140a6f (0 => $arg naminess) seems to violate the notion of "avoid special cases" 19:40
and the example of doit :123<now> looks like it has the potential to be especially confusing 19:41
(perhaps make it clear that :123<now> is not a Pair
I'll go back and scan the irclogs to see what prompted the update in the first place... gotta run for now though -- bbl 19:43
19:48 Mowah left 19:52 bluescreen10 left 19:56 jevin left 19:57 mberends left, bluescreen10 joined
tadzik if somebody didn't break spectest recently, gsoc-podparser would pass the spectests 19:58
19:58 mberends joined
tadzik (was that valid English?) 19:58
19:59 _jaldhar left
tylercurtis tadzik: I think "hadn't broken" instead of "didn't break" would be more correct. 20:01
tadzik that's the same conditional as in "if I had a lot of money I would buy a car", right? 20:02
That's the only example I remember
I always learned through "that feels right" instead of knowing the rules, but that results in situation like "hrm, that sounds wrong"
nopaste.snit.ch/47164 -- the first one looks like a gsoc-podparser win 20:04
20:06 huf joined 20:07 snearch left 20:08 silug left 20:09 saaki joined
dalek ast: 98727c8 | tadzik++ | S32- (3 files):
Fix plans
20:17
TimToady pmichaud: the original discussion was at irclog.perlgeek.de/perl6/2011-05-21#i_3781961 20:20
20:22 thou left
jnthn colomon: I don't plan to do anything about it in master, however if anybody wants to then the patch I did in nom should make it relatively obvious what to do. 20:23
colomon how soon is r-nom going to land? ;)
jnthn colomon: However the patch can't be applied to master as the binder in nom has changed too much due to the parameter/signature unification.
colomon jnthn & masak: ooo, nice position paper there. :) 20:25
jnthn colomon: I'd love to land it for the June release, but it may take a little longer than that.
colomon: June has lots of red days.
colomon: Which will be conducive to hacking on nom. :)
colomon oh, for what's it's worth, I meant "land" in the rakudo-nom is hackable sense, not in the "replaces beta" sense. 20:26
jnthn oh!
Then sooner.
:)
TimToady wonders in what sense it's not hackable now... 20:27
colomon TimToady: in my case it was the "cannot be built on OS X" sense, but that may have been fixed by now
tadzik I think 'hackable' is like 'it builds, and mere mortals can work towards fixing the tests' 20:28
colomon tadzik++
or fixing its code.
jnthn TimToady: What tadzik said.
I need to put back methods and probably attributes before people can really dig in more :) 20:29
colomon that heady period last year where we had a zillion ng commits each day, because it was easy to hack on and make useful progress very quickly.
jnthn Right, that's the point I'm aiming at. 20:30
tadzik fun fun fun fun
colomon fun fun fun fun AND we can help lighten the load on jnthn++
jnthn :) 20:31
Hmm...huh. It builds on my laptop and not on my desktop... :/ 20:32
jnthn wonders whoz up with that
tadzik: Does current nom mostly build for you (e.g. up until the setting compilation fail)? 20:33
frettled jnthn: it's obvious that it requires personal warmth and tenderness
20:33 pamera left
jnthn frettled: Actually I can't use laptops while placed on my lap :) 20:34
frettled: My eyesight isn't good enough to see the screen properly without doing my back n.
*in
Tene phenny: ask masak whether he ever released his slides from yapc::ru
phenny Tene: I'll pass that on when masak is around.
Tene phthanks
moritz didn't he show a download link in here before the talk? 20:35
frettled jnthn: A not-so-classic case of PICNIC, then 20:36
…though I have a weird image of a severely contorted jnthn, trying to peer closely at a laptop screen in his lap 20:37
Tene moritz: I wasn't watching IRC then 20:38
moritz jnthn: nom build fails with "Illegal option --vmlibs" 20:39
Tene moritz: he did; found it
thanks
phenny: tell masak nm; moritz mentioned that you already posted a link
phenny Tene: I'll pass that on when masak is around.
moritz jnthn: ...during the step home/moritz/p6/rakudo/parrot_install/bin/nqp --target=pir --output=src/gen/perl6-metamodel.pir --encoding=utf8 \ --vmlibs=perl6_ops src/gen/Metamodel.pm
does that mean my nqp is too old? 20:40
seems like
(non-versioned dependencies)-- 20:41
tadzik jnthn: mostly build, yes 20:43
jnthn: depends on how "current" is your current
jnthn: I'm now running the spectests on gsoc-podparser, and I'm not going to make my cpu cry even more
it's running make spectest for like 5th time today already 20:44
moritz Tene: jnthn asked about nom :-)
tadzik Tene: he did, I have them
(the slides)
oh, you already know that
jnthn moritz: Yes, means nqp is too old. 20:46
moritz: Feel free to implement the versioning.
Tene moritz: eh? what about jnthn?
moritz Tene: sorry, meant to talk to tadzik
jnthn: huh, "/perl6 --setting=NULL --target=pir --output=src/gen/CORE.setting.pir src/gen/CORE.setting" spits out PIR to STDOUT 20:47
tadzik moritz: what did you mean? 20:48
ha, it's official!
*drums drums drums*
gsoc-podparser passes more spectests than master \o/
moritz tadzik++ 20:49
moritz doesn't think coeherent anymore and should really sleep now
good night
colomon o/ 20:50
20:51 ymasory left
tadzik o/ 20:52
20:55 thou joined
jnthn moritz: Yeah, debugging output. 20:57
moritz: I'll clear it up in a bit.
21:10 hercynium left
dalek kudo/nom: e11dd26 | jnthn++ | src/Perl6/SymbolTable.pm:
Make sure we can install symbols that originate from other compilation units in lexicals.
21:17
kudo/nom: 58b566b | jnthn++ | src/Perl6/Grammar.pm:
Allow defining of previously stubbed packages. Also detect illegal redeclarations and whine.
kudo/nom: 7ab3578 | jnthn++ | src/Perl6/SymbolTable.pm:
Also need to allow composing things from another SC - it shouldn't ever really happen in real code but it does in the bootstrap.
21:20 ymasory joined
felher Hey folks. Is there a more explicit way of testing if $some_object is an instance of class X than smart-matching $some_object against X? 21:26
tadzik rakudo: my $a = "foo"; say $a.isa(Str)
p6eval rakudo 30ebf7: OUTPUT«1␤»
tadzik rakudo: my $a = "foo"; say $a.isa(Num)
p6eval rakudo 30ebf7: OUTPUT«0␤»
tadzik there we go
felher tadzik: Great. Thnx!
tadzik you're welcome :) 21:27
21:27 wknight8111 left 21:32 icwiener left 21:35 mj41 left
PerlJam felher: smart matching is not "isa" anyway 21:46
felher PerlJam: Whats the difference between $object ~~ X, where X is a classname and $object.isa(X)? 21:48
dalek kudo/nom: 45cfbca | jnthn++ | src/Perl6/Metamodel/MultipleInheritance.pm:
Avoid tripping up over containers.
kudo/nom: 403525b | jnthn++ | / (7 files):
Add empty non-stub versions of various classes. We now generate PIR for the setting for the first time (e.g. this resolved the incomplete stubs errors). Sadly, said PIR doesn't actually compile just yet.
jnthn felher: Probably none in terms of results. In master, the .isa is probably faster. In nom, the ~~ will probably be faster. 21:53
tadzik jnthn: why so? 21:54
PerlJam felher: $object ~~ X is the same as $object.does(X), which is not quite the same semantics as $object.isa(X) for all X :) 21:56
jnthn tadzik: ~~ will hit the type check cache which handles the more general notion of acceptance. 21:59
22:00 pernatiy joined
felher PerlJam: Ah, k, thnx. I guess i have to read a bit more about that. :) 22:02
jnthn: what is the 'nom'-branch? Is it kind of a 'experimental' branch? 22:04
jnthn felher: It's "next generation" of Rakudo, essentially. New meta-model, much more support for gradual typing, preparing the ground for many future optimizations. 22:06
felher jnthn: Sounds good. :) ty. 22:08
dalek kudo/nom: 5ee5d8e | jnthn++ | src/ (2 files):
A little more sketching out of how we'll handle lazy magicals.
22:09
kudo/nom: 0a91e5b | jnthn++ | src/Perl6/Actions.pm:
Need to load 6model guts. With this, we make it through the build for the first time.
tadzik yay! 22:11
jnthn: I can't make it through the build 22:12
jnthn tadzik: Aww. Failure mode?
tadzik nopaste.snit.ch/47165
jnthn Dang, that's what I get on my desktop too 22:13
Wonder what's different on my laptop...
tadzik what about a fresh clone?
jnthn What's on line 1343, ooc?
colomon perl Configure --gen-nqp ? 22:14
jnthn ah, repr_bind_attr_str
I fixed that in nqp yesterday
tadzik: What colomon++ said. 22:15
wait, though...I did that...
colomon I was just asking if that was the way to build nom. :)
jnthn colomon: Ah 22:16
colomon: Yes, plus --gen-parrot if you need a Parrot also.
colomon so like perl Configure --gen-parrot --gen-nqp ?
jnthn Yes
22:16 alester left
colomon hits ctrl-C and types that in. 22:17
22:18 pernatiy left, ymasory left
jnthn colomon: fwiw, I suspect the OSX bug isn't resolved. 22:20
colomon thought I'd try, anyway.
tadzik jnthn: I think I have nqp installed, but will check
jnthn: yes it is
oh, but it's not todays 22:21
22:22 silug joined
jnthn hmm...something is odd. 22:23
colomon oh, great, have to deal with the PCRE bug too. sigh.
jnthn oh hmm
colomon just need to dig out the command-line argument for it. 22:24
tadzik --without-pcre
jnthn tadzik: oh wtf...somehow --gen-nqp isn't actually getting the latest nqp 22:25
fixed. 22:26
dalek kudo/nom: 83a3b57 | jnthn++ | build/gen_nqp.pl:
Fix --gen-nqp.
22:30
kudo/nom: e3e62e5 | jnthn++ | .gitignore:
Update .gitignore.
22:38 zorgnax joined
zorgnax Is there a Perl6 infix calculator example anywhere? 22:39
22:39 Chillance left, ymasory joined
tadzik well, the Perl 6 executable is one 22:42
but I'm not aware of any 22:43
zorgnax Maybe Ill make one 22:44
I dont see any clear example of one
tadzik you may want to look around for a Compiler-calculator from some pmichaud's talk 22:46
22:48 lumi_ left
jnthn tadzik: After the --gen-nqp fix, the build seems to be going better on my other machine now :) 22:49
tadzik jnthn: I'm not using --gen-nqp, and updating nqp helped
jnthn++ # it compiles! 22:51
so, where's the LHF? :P
jnthn tadzik: Add empty Code/Block/Routine/Sub/Method/Signature/Parameter in setting; see the patch where I added Mu/Any/Cool 22:52
tadzik how about all of what you listed? :) 22:53
colomon jnthn: woah. I get a seg fault now when I build rakudo-nom 22:54
sorear good * #perl6
tadzik hello sorear
how did the final go, what was it?
colomon Users/colomon/tools/nom/rakudo/parrot_install/bin/nqp --target=pir --output=src/gen/perl6-grammar.pir --encoding=utf8 src/Perl6/Grammar.pm
make: *** [Perl6/Grammar.pbc] Segmentation fault
jnthn colomon: I'm seeing that occasionally too :/ 22:55
colomon jnthn: twice in a row for me, suggested cure?
jnthn colomon: Sadly, none - it needs some time with a debugger, I suspect.
colomon: Feel free to try and get a gdb backtrace if you are able to reproduce it reliably.
colomon trying valgrind now. :) 22:56
sorear felher: $object ~~ X is actually closer to $object.^does(X) 22:57
tadzik: well. "english".
jnthn: has the ~~ changes landed in rakudo/nom yet? 22:58
colomon changes?
tadzik sorear: isn't english your native language?
sorear 14:53 < jnthn> felher: Probably none in terms of results. In master, the .isa is probably faster. In nom, the ~~ will probably be faster.
14:59 < jnthn> tadzik: ~~ will hit the type check cache which handles the more general notion of acceptance.
tadzik jnthn: wklej.org/id/534389/ how is that? 22:59
sorear tadzik: in the US school system, "english" also includes things like analyzing texts and learning how to write papers 23:00
23:00 zorgnax left
sorear tadzik: do they distinguish that from "polski" where you live? 23:00
tadzik that's like Polish in Poland. But here, that usually ends in high school
sorear: yeah, it's quite equivalent
so, no, they don't distinguish it. But you don't study Polski anymore when studying on the University. Unless you are actually on Polonistyka 23:01
jnthn sorear: I don't even have operators yet. :) 23:02
tadzik phenny: "Polonistyka"?
phenny tadzik: "Polish Studies" (pl to en, translate.google.com)
tadzik yeah, whatever
jnthn tadzik: nearly; you_are_here should come last.
colomon jnthn: valgrinding now for real. :)
tadzik jnthn: alright 23:03
jnthn: so once attributes and methods land, we could start filling those in? 23:04
felher sorear: yeah, thnx. But i sure have to read perlcabal.org/syn/S12.html to understand what .^does means and what those meta-objects are.
jnthn tadzik: Yeah. Those are next on the nommap.
tadzik supercool
felher But not today. good night, folks.
colomon o/ 23:05
tadzik g'night felher
dalek kudo/nom: 4c68206 | tadzik++ | / (8 files):
Add some new (empty) classes to CORE.setting
23:07
jnthn tadzik: Thanks. :) 23:08
23:08 thou left
tadzik thanks too :) 23:09
23:09 mtk left 23:10 Woody2143 left
tadzik should I also add everything else, like Rat, Complex, IO etc? 23:10
not sure if there's really a point in that :) 23:11
jnthn tadzik: Yeah, not that much point yet.
dalek kudo/nom: 0f60ad2 | jnthn++ | src/Perl6/ModuleLoader.pm:
Make setting's use Perl6::Metamodel function.
23:12
kudo/nom: 3eae147 | jnthn++ | src/Perl6/Actions.pm:
Fix package compilation up a bit.
kudo/nom: dff1ae8 | jnthn++ | src/Perl6/Actions.pm:
Toss incorrect CHECK-time. Also, get us closer to actual being able to run something.
23:12 lumi_ joined 23:13 Woody2143 joined
jnthn Oops, gotta teach in the morning 23:14
sleep &
23:17 _jaldhar joined
tadzik g'night 23:17
colomon arrrrgh. I think it worked fine under valgrind. 23:18
23:18 mtk joined 23:21 daniel-s joined 23:22 wamba left
sorear pmichaud: ping 23:25
daniel-s: hi 23:26
daniel-s hey
daniel-s is cold
sorear hugs
PerlJam resists the urge to shout "hug pile!" 23:27
daniel-s lol
google says it's 12C outside
tadzik is the game "grenade" known to you? It was quite popular back in the scouting dys
daniel-s ...google exaggerates
tadzik daniel-s: that's so nerdy. You could open the window 23:28
sorear Where I live, it's never even 1 C outside.
:/
daniel-s tadzik, and look outside? perish the thought! I might get a ray of sunlight on my skin or even breath fresh air 23:29
tadzik ;)
sorear just get a cheap thermometer
the kind with a bimetallic coil and a needle need no batteries and are pretty much indestructable
tadzik heh, I've worn a long pants for the first time in a week today, and I was boiling
PerlJam It's 31C where I am ... and that's cool for this time of the year. 23:30
23:31 awoodland left
colomon 22C here, and that's about normal, though it's disturbingly humid 23:31
23:34 ymasory left
colomon got a core dump, anyway... 23:34
tadzik I like it when I use valgrind to catch a segfault, and valgrind survives the segfault 23:35
sorear 62 outside, but I live in backwardland where the degrees are smaller 23:36
colomon backwardland++
PerlJam sorear: Just check the thing that displays temperature in celcius and you're good to go ;) 23:37
colomon I asked google to translate to C for me. :)
gist.github.com/987868 is my backtrace (nom crash) 23:38
sorear PerlJam: the bimetallic coil thermometer outside is only marked in F.
also, I think it's at least as old as I am.
23:39 yegor joined 23:41 whiteknight joined 23:46 yegor left
sorear phenny: tell pmichaud How hard do you intend to fight 1 => $foo ? I'd rather not be implementing it if it will go away soon. 23:48
phenny sorear: I'll pass that on when pmichaud is around.
TimToady surely it's how hard pmichaud++ fights relative to how hard /me fights... :) 23:49
PerlJam What's wrong with 1 => $foo ? 23:54
sorear PerlJam: irclog.perlgeek.de/perl6/2011-05-23#i_3791367 23:55