»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
colomon pretty sure "powerfulest" is not a word in English 00:08
vendethiel CAN BE
:P
colomon …. but then carlin++ has already answered your question quite well. 00:09
m: say (1..*).grep([+] *.comb == 7)[^20]
camelia rakudo-moar 866817: OUTPUT«Cannot call 'Numeric'; none of these signatures match:␤:(Mu:U \v: *%_)␤ in sub infix:<+> at src/gen/m-CORE.setting:4181␤ in sub at src/gen/m-CORE.setting:17420␤ in block at /tmp/RU2TZxAUFp:1␤␤»
colomon m: say (1..*).grep(([+] *.comb) == 7)[^20] 00:10
camelia rakudo-moar 866817: OUTPUT«Cannot call 'Numeric'; none of these signatures match:␤:(Mu:U \v: *%_)␤ in sub infix:<+> at src/gen/m-CORE.setting:4181␤ in sub at src/gen/m-CORE.setting:17420␤ in block at /tmp/yZmDxp6RM9:1␤␤»
colomon m: say (1..*).grep(-> $i { ([+] $i.comb) == 7 })[^20] 00:10
camelia rakudo-moar 866817: OUTPUT«7 16 25 34 43 52 61 70 106 115 124 133 142 151 160 205 214 223 232 241␤»
colomon realizes that's not actually the sequence operator he used in there. 00:11
vendethiel oh, I actually already explained [OP], nice. Missing [\OP], though. I'll check what it does (array ?) later 00:19
timotimo [\| 00:23
m: [\~] <a b c d e f>
camelia ( no output ) 00:24
timotimo m: say [\~] <a b c d e f>
camelia rakudo-moar 866817: OUTPUT«a ab abc abcd abcde abcdef␤»
timotimo m: say [~] <a b c d e f>
camelia rakudo-moar 866817: OUTPUT«abcdef␤»
hoverboard ou 05:20
lizmat_ #76456 # just checking 05:42
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=76456
jiffies 。 07:22
jiffies hello? 07:37
nwc10 FROGGS_: unfortunately 5c191ee846461727a14f84cb1df75c0d35396c80 breaks NQP tests. Specifically t/nqp/19-file-ops.t (which no longer parses) and t/qregex/01-qregex.t 08:18
nwc10 FROGGS_: unfortunately it also breaks the Rakudo build 08:50
/home/nicholas/Sandpit/moar-san/bin/nqp-m --target=mbc --output=blib/Perl6/Grammar.moarvm --encoding=utf8 \ src/Perl6/Grammar.nqp
Unrecognized regex modifier :my at line 903, near " $*POD_ALL"
[backtrace.] 08:51
itz_ I have perl6-m built OK as of 8668171 on linux 08:56
nwc10 itz_: using which nqp? 09:01
FROGGS[mobile] ewww 09:02
I'm moving it to a branch then...
nwc10 that seems best, if you don't have time to fix it "Real soon" 09:03
dalek p/charrange: e1b6829 | (Tobias Leich)++ | src/QRegex/P6Regex/Actions.nqp:
use ignorecase in charrange on moar
 09:05
p: 3b7496c | (Tobias Leich)++ | src/ (2 files):
back to using enumcharlist instead of charrange on moar
 09:07
itz_ This is nqp version 2014.07-1-g6d4ea16 built on MoarVM version 2014.07
weird where does the nqp sha come from? 09:09
FROGGS_ 2014.07-1 is the first patch after the release 09:11
this is before the charrange stuff happened
itz_: your's is about 24 hours old: github.com/perl6/nqp/commits/master 09:12
itz_ ah I assumed wrongly rakudobrew used the latest nqp
masak antenoon, #perl6 09:37
* vendethiel thinks it's kinda weird that only :: hasn't received sigil invariance
vendethiel o/, masak 09:38
masak vendethiel: I think (a) :: has nothing to do with sigil invariance, and (b) the thing you're actually complaining about stems from a confusion about the function of the sigil, and its relation to the rest of the identifier.
vendethiel masak: it's probably that, for me, only the actual variable name (package-less) is the variable name, yet 09:39
vendethiel (it's true that it's unrelated to sigil invariance, actually, it's just that I always expect it to be `A::$B`) 09:40
masak m: module Foo { our $pi = 3 }; say $Foo::pi
camelia rakudo-moar 866817: OUTPUT«3␤»
masak vendethiel: thinking about it, I don't see any real reason it *couldn't* be Foo::$pi 09:41
but it would be a rather big break from Perl 5, for no really good reason.
also, it "feels" wrong, because the sigil has a function that doesn't rhyme well with being put in the middle like that. 09:42
the sigil is a noun marker.
vendethiel I admit I tend to struggle a bit with the module system. Hopefully it'll go better after I read my tutorial ... oh :-P
masak: it feels "right" to me, because what we're marking is the variable itself, not the package
masak yes, I know.
that's subordinate to marking the whole term up-front, though. 09:43
FROGGS_ nwc10: I was able to build nqp/rakudo on every backend right now... 09:53
nqp tests also pass 09:55
sorry for the noise
FROGGS_ and I've found one issue with the charrange branch: 10:21
nqp-m -e 'say( "2" ~~ / <[2..3]> | b / )'
Substring length (-4) cannot be negative
an | alternation breaks it 10:22
zengargoyle i'm thinking Foo::$pi makes more sense now than $Foo::pi does, esp. when written Foo::<$pi> 10:25
FROGGS_ we are moving the sigil the the last element in the compiler to look up a symbol in fact 10:27
but I know that we're checking for things starting with a certain sigil in the grammar
like in [[&foo]]
that would work with [[&Foo::foo]] but not with [[Foo::&foo]] 10:28
about the charrange... seems like setting the subtype breaks it :/ (I used it for ignorecase, but seems I can't do so) 10:29
zengargoyle doesn't know what [[&foo]] means. :) so many specs skimmed... 10:31
vendethiel zengargoyle: reduce with the foo sub
nwc10 FROGGS_: I left it running and I can confirm NQP, Rakudo and spectest PASS
FROGGS_ nwc10: thank you! 10:32
zengargoyle so subs vs operator takes an extra pair of []?
masak m: sub foo($a, $b) { $a + $b }; say 2 [&foo] 3 10:52
camelia rakudo-moar 866817: OUTPUT«===SORRY!===␤Error while compiling op ifnull: The 'ifnull' op expects two children␤»
masak m: sub foo($a, $b) { $a + $b }; say [[&foo]] 2, 3
camelia rakudo-moar 866817: OUTPUT«5␤»
masak submits LTA rakudobug
colomon Hmmm, three new failing modules this morning: YAML, XXX, and Test-Corpus. All fail in test stage. 11:21
BTW, I don't know if anyone else has picked up on this -- we have 50-60 more passing modules than we had a year ago. 11:22
colomon looking at YAML, it's Could not find t::Bridge in any of: /home/colomon/tools/yaml-pm6/lib, /home/colomon/tools/yaml-pm6/t where Bridge.pm is in t/ 11:30
hmmm…. is . no longer included in @INC?
FROGGS_ colomon: it never was 11:36
bbl
colomon FROGGS_: well, clearly something related has changed, because the module found Bridge.pm yesterday and doesn't today, and the module itself has not been changed in six months. 11:38
psch good day, #perl6 o/ 11:59
moritz \o psch 12:02
dalek osystem: 5a94e38 | Carlin++ | META.list:
add Printing::Jdf
 12:41
psch m: say <<a\ b>>.perl; #`[[ RT #76120, but: ]] say "a\ b".perl # should backslash-space throw UnrecognizedSequence here? 13:23
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=76120
camelia rakudo-moar 866817: OUTPUT«("a", "b")␤"a b"␤»
psch or does unspacing work in double quoted strings? 13:24
n: say "a\ b"
camelia niecza v24-109-g48a8de3: OUTPUT«a b␤»
masak I'm of two minds.
(a) it doesn't mean anything, so it should be disallowed (to catch spurious uses which are probably indicative of something wrong) 13:25
(b) it's harmless, so we might as well silently interpret it as a literal space
I think (a) is what qq strings usually do.
m: say "a\qb"
camelia rakudo-moar 866817: OUTPUT«===SORRY!=== Error while compiling /tmp/St_MNH0Prn␤Couldn't find terminator "␤at /tmp/St_MNH0Prn:1␤------> say "a\qb"⏏<EOL>␤ expecting any of:␤ argument list␤ prefix or term␤ pref…»
masak huh. 13:26
psch huh indeed 13:34
m: say q|a\qq |
camelia rakudo-moar 866817: OUTPUT«===SORRY!=== Error while compiling /tmp/4cOyPkFW4n␤Couldn't find terminator |␤at /tmp/4cOyPkFW4n:1␤------> say q|a\qq |⏏<EOL>␤ expecting any of:␤ argument list␤ prefix or term␤ …»
masak I kind of expected it to go "there's no \q, silly!" 13:40
psch m: say "\qqq" 13:41
camelia rakudo-moar 866817: OUTPUT«===SORRY!=== Error while compiling /tmp/aPbiqbXAkI␤Unrecognized backslash sequence: '\q'␤at /tmp/aPbiqbXAkI:1␤------> say "\qq⏏q"␤ expecting any of:␤ argument list␤ prefix or term␤ …»
psch i suspect Grammar.nqp:4204 plays a part there
masak looks 13:42
yes, probably. 13:44
masak but I had it as "q strings always let through things they don't recognize, qq strings are more strict" 13:44
I don't understand why Rakudo behaves as it does with the last three evals.
psch m: say 'ab\qq"\x63"' 13:59
camelia rakudo-moar 866817: OUTPUT«abc␤»
psch i think that means that broken cases of \qq have LTA errors? 14:00
masak waitwait... 14:18
that's a *feature*?
is this in S02 somewhere?
psch "Intepolating into a single-quoted string" 14:20
masak looks
masak huh. 14:21
and it's from (at least) 2008.
ok.
anyway, I don't think it should affect qq strings in any way.
FROGGS_ yay, I think my charrange patch works on moar.... (needed also NFA tweaks) 14:21
ohh, RT is back \o/ 14:22
psch right. so the question remains if '\ ' is meaningful in any way 14:24
masak did you use single quotes there deliberately? 14:25
psch yes, but wrongfully too :)
masak because I don't see backslash-space being problematic in single quotes at all :)
psch you're right of course, i meant "\ "
and it is meaningful right now as well, but it means " " 14:26
i can see the case in the original ticket as a bit more involved though, i.e. <<foo\ bar>>
that *could* mean <<foobar>>, if unspace works into extrapolation, but that doesn't happen currently 14:27
psch say <<\x61 \o142>> # this case works at least 14:27
m: say <<\x61 \o142>> # this case works at least 14:28
camelia rakudo-moar 866817: OUTPUT«a b␤»
FROGGS_ yay, m-spectest PASS 14:41
FROGGS_ I think RT #76734 was fixed two years ago... 14:45
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=76734
FROGGS_ lots and lots of tickets can be closed I think :o)
carlin lots of bugs will have just gone away as various bits were re-written 14:46
FROGGS_ nah... the ticket I pasted was fixed by me, I had that problem in my first module 14:48
though, it is weird that I have not spotted that ticket within these two years
ahh, RT #76734 already is marked testneeded 14:51
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=76734
FROGGS_ m: $_ = 'oh 2!'; s[\d+] += 40; .say 14:52
camelia rakudo-moar 866817: OUTPUT«===SORRY!=== Error while compiling /tmp/9_M0hRnKqF␤Malformed assignment operator␤at /tmp/9_M0hRnKqF:1␤------> $_ = 'oh 2!'; s[\d+] +=⏏ 40; .say␤»
masak FROGGS_: the test for RT #76734 is still being skipped. 14:55
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=76734
FROGGS_ huh 14:57
masak so I think it needs to be un-skiped before the ticket is closed. 14:58
FROGGS_ masak: it fails with: 15:01
Cannot modify an immutable A 15:02
in block at t/spec/S03-operators/assign.rakudo.moar:886
so that overloaded infix:<=> is not taken into account at all 15:03
dalek kudo-star-daily: 7bfe90d | coke++ | log/ (14 files):
today (automated commit)
 15:09
rl6-roast-data: 65a50f2 | coke++ | / (5 files):
today (automated commit)
[Coke] ... it's been nine days since pugs was run. *sniff* 15:13
carlin long live rakudo 15:14
[Coke] moar is aborting -new- tests.
FROGGS_ and I killed the star daily :/
[Coke] S32-list/uniq.t aborted 26 test(s)
integration/99problems-11-to-20.rakudo.moar aborted 14 test(s)
[Coke] Unrecognized regex modifier :my at line 903, near " $*POD_ALL" 15:15
?
FROGGS_ yeah
nqp HEAD is faulty at that time, but is back in a good shape 15:16
me is getting lazy when it gets all dark
masak nwc10: have you seen the way jQuery is handling their Big Major Version Upgrade? jQuery 2.0 was released back in April 2013, and most of the world still seems to be on jQuery 1.x. 15:46
nwc10: on CDNs, jquery-latest.js tends to point to the latest 1.x version. (and I think they tried it the other way for a while, and there was chaos/outcry) 15:47
Ven yay hell breaking loose :)
psch they're only serving 1.11.1 from now on as jquery-latest.js -> blog.jquery.com/2014/07/03/dont-use...latest-js/ 15:49
or rather "from then on" :)
masak nwc10: *but* they are keeping feature parity in all their releases.
so it's not quite the same reason to switch as with Perl or as with Python. 15:50
the reason to switch in jQuery's case is mainly to get a smaller jQuery.
Ven masak: not, it's to fight against crime :P 15:53
masak "[jQuery 2.0] is smaller, faster, and can be used in JavaScript environments where the code needed for old-IE compatibility often causes problems of its own." --blog.jquery.com/2013/04/18/jquery-2-0-released/ 15:53
Ven I think Perl 6 is maturing fast enough that I'm ready to tell people to use it. 15:54
Actually, let me fix that -- I think perl 6 is maturing fast enough that **I** am ready to use it :P
jQuery has no "mature enough". They're waiting on an external factor -- quite different, imho 15:55
masak this comment has all the rings of the Perl 5/Perl 6 rift: blog.jquery.com/2013/04/18/jquery-2...ent-541881
Ven: well, Python is waiting on an "external factor", too. namely, all of their ecosystem being *willing* to switch over. 15:56
timotimo Ven: it'd probably be nice if you pulled in your people in batches so that instead of getting a thousand newcomers at once, we can make sure to learn and improve between waves :P 15:56
Ven Disagreed. That's something they can influence (like we can)
timotimo: well, I might not have evangelized thousands ! But I definitely got over 20 people interested, and 10 doing something with it right away. 15:57
timotimo :) 15:58
raiph <masak> I don't see any real reason it *couldn't* be Foo::$pi ... <FROGGS> ... not with [[Foo::&foo]] 16:06
I personally find Foo::$pi much preferable even if it doesn't work in some cases 16:08
nwc10 masak: this seems somewhat like Apache, where 1.3x stuck around for years and years, while 2 was also out 16:21
and it does seem that there's a good pull-factor for jquery 16:22
(I don't acutally know what the end user pull case for Apache 2 was)
japhb nwc10: Wasn't it the improved worker model? My memory is pretty fuzzy. 16:24
dalek blets: 1e1bbe6 | (Herbert Breunung)++ | docs/appendix-b-grouped.txt:
fix utf
 16:28
timotimo .u ⟪ 16:47
yoleaux U+27EA MATHEMATICAL LEFT DOUBLE ANGLE BRACKET [Ps] (⟪)
timotimo i don't think that's the right character actually
timotimo m: say (1, 2, 3) ⟪+⟪ ( 5, 5, 5, 5, 5, 5 ) 16:48
camelia rakudo-moar 866817: OUTPUT«===SORRY!=== Error while compiling /tmp/Nok9gcPCIP␤Two terms in a row␤at /tmp/Nok9gcPCIP:1␤------> say (1, 2, 3) ⏏⟪+⟪ ( 5, 5, 5, 5, 5, 5 )␤ expecting any of:␤ postfix␤ infix stopper␤…»
timotimo m: say (1, 2, 3) «+« ( 5, 5, 5, 5, 5, 5 )
camelia rakudo-moar 866817: OUTPUT«6 7 8 6 7 8␤»
Ven Trying to sway a pythonista to perl 6. This is gonna be a hard one :D 16:49
timotimo don't forget to stay honest :)
carlin it's a tough sell since python has `import antigravity` 16:51
timotimo yeah, i was thinking the other day we ought to have something like "import this" 16:51
pmurias why does the nqp-moar and nqp-jvm use a different MAIN signature? 17:04
Ven timotimo: Are you implying I'd need to lie ;-) ?
timotimo well ... dunno?
Ven belieeeve :P 17:05
masak raiI think many with you would prefer Foo::$pi. my point -- enhanced by FROGGS' use case -- is that having it that way is a way to misunderstand sigils and exactly why they are useful. cf string interpolation for another such use case. 17:06
raiph*
Ven (not in me, I mean. In the ability of P6 to swing a pythonista gangsta)
timotimo :) 17:08
maybe i'll spend some time writing python benchmark equivalents to the perl6bench benchmarks we have?
at least the microbenchmarks should be doable
japhb timotimo: And ... that would be why I changed 'perls' to 'compilers' in perl6bench. ;-) 17:10
timotimo ayup 17:11
you think we have any chance against cpython performance wise?
definitely no chance against pypy, that's for sure
FROGGS_ still would be good to know where we are 17:12
timotimo of course
pmurias has the performance improved so much recently?
timotimo our performance, you mean?
pmurias yes 17:13
FROGGS_ this year happened a lot
timotimo jnthn.net/perl6/bench/2014-07-15.html - look at this
last year's parrot vs last month's moar and a moar from a few days ago
japhb timotimo: IIRC CPython got faster than perl5 at some point, and certainly PyPy is really quite fast, but I'm with FROGGS, I'd love to know exactly where we stand.
vendethiel timotimo: it's really amazing :-) 17:21
carlin while_concat_native graph is nice 17:26
does rakudo really beat perl5 in mandelbrot? that's impressive
FROGGS_ well, it looks like... 17:27
japhb carlin: When using Math::Complex, yes. I suspect that if done using plain nums and shuffling the real/imaginary parts around by hand, perl5 would win for now. 17:29
*When perl5 is using Math::Complex
carlin come a long way since 2009 when colomon++'s mandelbrot scripts would eithe take hours to finish or segfault 17:37
retupmoca $ perl6 -e 'sleep(9999);' # 0% CPU usage 17:39
$ perl6 -e 'Promise.in(2);sleep(9999);' # 100% cpu usage (forever)
^ that's a little annoying
retupmoca (on perl6-m) 17:39
ChoHag How can I replace or in some way augment the code which sets an attribute's value? 17:43
ChoHag So that if a consumer of a class calls $class.attribute() = "value", I can have a piece of code which gets executed. 17:44
Prior to or post the actual changing of the value. 17:45
timotimo needs Proxy 17:46
ChoHag Hmm I tried that before and couldn't get it to do ... something.
I can't remember nwo.
FROGGS_ ChoHag: see github.com/perl6/roast/blob/master...rs/proxy.t 17:47
ChoHag Thanks. 17:48
I shall decipher that when I'm not melting into the sofa.
FROGGS_ m: class Foo { has $!bar; method bar() is rw { Proxy.new( FETCH => sub ($self) { $!bar }, STORE => sub ($self, $value) { $!bar = $value**2 } ) } }; my $f = Foo.new; $f.bar = 42; say $f.bar # ChoHag 17:51
camelia rakudo-moar 866817: OUTPUT«1764␤»
ChoHag That too.
ChoHag It's a hot day. I'm a Pratchett troll. 17:51
FROGGS_ ChoHag: though you must know that FETCH/STORE might be called more often then you would expect
ChoHag OK I'll bear that in mind. 17:52
timotimo aye, our code gen and stuff assume fetch is extremely cheap, so it'll do that quite often in many cases 17:58
m: my $foo = Proxy.new( FETCH => -> { say "fetched!"; 10 } ); say "hi" if $foo > 5; 17:59
camelia rakudo-moar 866817: OUTPUT«Too many positional parameters passed; got 1 but expected 0␤ in block at /tmp/fhKp1xetbT:1␤␤»
timotimo m: my $foo = Proxy.new( FETCH => -> $_ { say "fetched!"; 10 } ); say "hi" if $foo > 5; 17:59
camelia rakudo-moar 866817: OUTPUT«fetched!␤hi␤»
timotimo m: my $foo = Proxy.new( FETCH => -> $_ { say "fetched!"; 10 } ); sub test(Int $arg) { say "called" }; test($foo);
camelia rakudo-moar 866817: OUTPUT«fetched!␤called␤»
timotimo huh. 18:00
carlin m: Proxy.new( FETCH => -> $_ { return 1 } ); 18:03
camelia rakudo-moar 866817: OUTPUT«(signal )»
timotimo carlin: that tries to return from outside the pointy block
that's not a good idea :)
carlin yeah, I got that impression from the segfault :p
timotimo i wonder where we have to catch that to give the proper "trying to return from something that you can't return from" or something error 18:04
dalek rl6-bench: 4b8e696 | (Geoffrey Broadwell)++ | README:
README Really Quick Start section, copyright bump, and minor cleanups
 18:07
rl6-bench: a7e9d83 | (Geoffrey Broadwell)++ | README (2 files):
Rename README to README.md so GitHub will render it
dalek rl6-bench: 1e57fe1 | (Geoffrey Broadwell)++ | README.md:
D'oh! Forgot a
 18:20
japhb Is there an easy way in current NQP to specify an array containing the values of a range? I want to be able to say something not much more complex than 'my @a := 1 .. 10;' 18:34
I recognize this is more likely an op than syntax, but if so, what op?
timotimo not that i know, no :(
japhb Gah. 18:35
japhb imagines an nqp::rangelist($min, $max)
An easy way to do '... for 1 .. 10' or 'for 1 .. 10 { ... }' would be nice too, where this is turned into a fast loop, rather than actually instantiating the entire list. 18:37
vendethiel read "nqp::evangelist" -- a bit tired, uh 18:39
japhb vendethiel: To what do you refer? 18:44
Ven japhb: to "nqp::rangelist"
japhb Oh, heh. :-) 18:45
dalek ast: b76ede5 | (Tobias Leich)++ | S03-metaops/cross.t:
RT #112602 test for 1..* X* 1..*
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=112602
Ven it ... it works :O ? yay~~
japhb FROGGS_: That should probably test that both sides of the X iterate properly over at least 2-3 iterations each. 18:47
(Or maybe that's another test)
FROGGS_ japhb: um, what?
japhb FROGGS_: Oh, the @a was misleading me, I read 1..* X* @a. 18:49
Though ... why mention @a at all if you're not going to test it?
dalek ast: 79a52c8 | (Tobias Leich)++ | S03-metaops/cross.t:
fix left-over from copy&pasting, japhb++
Ven what's the difference between `1..* X* 1..*` and `1 X* 1..*` , though :P ? 18:51
FROGGS_ well... :P 18:52
japhb Ven, functionally ... none, really.
Ven "it looks cool", ha
FROGGS_ but neither side is treated eagerly, so it matters for the test
flussence perl6: say 'ab\qq"\x63"'; 19:07
camelia rakudo-{parrot,jvm,moar} 866817, niecza v24-109-g48a8de3: OUTPUT«abc␤»
flussence this... scares me for some reason.
psch perl6: say 'ab\qx[echo "c"]' # ... 19:08
camelia niecza v24-109-g48a8de3: OUTPUT«Unhandled exception: rungather may not be used in safe mode␤ at /home/p6eval/niecza/lib/CORE.setting line 0 (rungather @ 0) ␤ at /tmp/tmpfile line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4595 (ANON @ 3) ␤ at /home/p…»
..rakudo-moar 866817: OUTPUT«qx, qqx is disallowed in restricted setting␤ in sub restricted at src/RESTRICTED.setting:1␤ in sub QX at src/RESTRICTED.setting:9␤ in block at /tmp/tmpfile:1␤␤»
..rakudo-jvm 866817: OUTPUT«abc␤␤»
..rakudo-parrot 866817: OUTPUT«qx, qqx is disallowed in restricted setting␤ in sub restricted at src/RESTRICTED.setting:2␤ in sub QX at src/RESTRICTED.setting:9␤ in block at /tmp/tmpfile:1␤␤»
dalek rl6-bench: 0923c8d | (Geoffrey Broadwell)++ | microbenchmarks.pl:
Add NQP version of split_string_constant

Also switch Perl 5 and Perl 6 variants to while loops to match NQP, and add regex tag to split_string_regex.
psch well, it's spec 19:09
dalek ast: 42c8533 | (Tobias Leich)++ | S03-metaops/zip.t:
RT #116036 test for <a b> Z=> []
 19:25
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=116036
dalek ast: db3b9ec | (David Warring [email@hidden.address] | integration/advent2009-day24.t:
completing advent 2009 day 24
 19:30
masak dwarring++ 19:31
dalek p/charrange: 0376eaf | (Tobias Leich)++ | src/vm/moar/QAST/QASTRegexCompilerMAST.nqp:
we use the first positional as subtype
 19:50
p/charrange: 7557ea1 | (Tobias Leich)++ | src/vm/parrot/ (3 files):
port EDGE_CHARRANGE to parrot backend
timotimo FROGGS_: what do your changes to charrange do? improve performance of case insensitive charranges? 19:56
FROGGS_ timotimo: yes, case insensitive ranges are also just a bounds checking now 19:57
and, we do not build the enumcharlist at all now
I wanna bench it, because it might make sense to not use charrange but enumcharlist in case the range is very small 19:58
m: say "\x[10001]" ~~ /<[\x10000..\xEFFFF]>/ # was not implemented on moar at all btw 20:01
camelia rakudo-moar 866817: OUTPUT«(signal )»
dalek p/charrange: efb67a8 | (Tobias Leich)++ | src/vm/jvm/runtime/org/perl6/nqp/ (2 files):
port EDGE_CHARRANGE to jvm backend
 20:06
dalek p/charrange: 2631218 | (Tobias Leich)++ | src/QRegex/ (2 files):
use charrange instead of enumcharlist

We have charrange now on all backends, also for case insensitive matches and also have EDGE_CHARRANGE support in the NFA. That mean we do not need to create an enumcharlist compatible string at all, which gets very expensive when the range is big enough.		 20:08
ast: 1c84a0b | (David Warring [email@hidden.address] | integration/advent2009-day06.t:
Fix/unfudge incorrect test.
FROGGS_ eww, I'm recompiling my perl6-m but I've got a running `p6bench setup` which uses it >.< 20:09
ChoHag Can I silence the warning: useless use of 'is rw' on $!a 20:11
The weather is now reasonable, and for some reason the Proxy object seems to work like it didn't last time I failed to use it.
FROGGS_ no, does not seem like as if you could silence it 20:12
ChoHag I thought not.
FROGGS_ m: class Foo { has $!a is rw } # also, that file/line is not helpful at all :/
camelia rakudo-moar 866817: OUTPUT«useless use of 'is rw' on $!a in any at src/Perl6/World.nqp:1673␤␤»
ChoHag I probably want a $. variable anyway with a wrapper rather then writing my own accessor function out of whole cloth. 20:13
But since I'm squeezing in some coding in while the weather and a toddler allow, I'm not applying much in the way of actual thought.
FROGGS_ ChoHag: you can make it a $. and provide the accessor method, that should work out
ChoHag Can you remove methods from a class at runtime? 20:14
FROGGS_ maybe using the MOP
ChoHag The MethodContainer.nqp suggests not.
FROGGS_ but I'm not sure
ChoHag Definitely appreciating that most of perl 6 is (not quite) written in perl 6. 20:15
FROGGS_ but btw, I don't think you need the 'is rw' at all, you can assign to it from within the class
yeah, otherwise I would have problems fixing bugs :o) 20:16
ChoHag I'm writing something which sort of fits in the space of library or api. 20:22
So I need to provide options I don't know about.
masak yay, host07.perl6.com:8080/report/masak looks much better today \o/ 20:26
something still wrong with Perl6::Literate, though. 20:28
not really sure how to find out what that is...
timotimo masak: well done! what broke druid? 20:30
i recall fixing druid a while back :)
masak timotimo: EVAL, bless(*), .exists, and either smartmatching or grammars changing semantics. 20:32
timotimo: yes, I saw you had fixed a lot of things. thank you for that.
Ven so, state of the art web frameworks in perl 6 are bailador and that other thingie ? 20:35
Ven tries to remember the other framework he's seen
cognominal sometimes I feel the Perl 6 is an adventure game where information nuggets can be on irc, on the web (can be on weird places like host07.perl6.com:8080). With tadzik, probably even geocaching is involved. 20:45
or maybe is is a mystic things where knowldeged is imparted from gurus. 20:46
timotimo hah :)
dalek ast: 70902a0 | (David Warring [email@hidden.address] | integration/advent2009-day07.t:
added a couple more tests
 20:47
lizmat stops working on the migration of Foo.moarvm -> Foo.pm.moarvm 20:48
Need to dig deeper into ModuleLoader.nqp to be able to fix the last failing spectests
sightseeing again& 20:49
cognominal with synopsess pointed to unimplemented things to increase the confusion. Admittedly today, you need some persistance to find them.
masak cognominal: it is our job to make the adventure game simple to play. 21:31
cognominal: re host07.perl6.com:8080, it's fairly OK that it's not more prominent, since it's a fairly internal role it's playing. documentation, tutorials, FAQs, blogs should be prominent. 21:33
zengargo1le i wish the Apocalypses and Exegeses were updated/re-implemented. the specs can be a *long* read, really repetitive in places, sorta too detailed in places. 21:41
masak I've sometimes toyed with the idea of going back and re-writing Apocalypses.
I actually re-implemented some Exigeses long ago.
vendethiel has read all specs, but did not read any of the exigeses/apocalypses 21:42
masak oh, and I've sometimes wanted to ghostwrite A09 :)
vendethiel A09 ? 21:43
masak S09 references A09, which never existed.
the thought of writing it in the style of TimToady, passing it off as an actual Apocalypse, always appealed to me.
zengargo1le sometimes i'd just like the meat of the specs without the detailed analysis/reasoning. 21:44
vendethiel I admit that sometimes the NYI puts you off :D. 21:47
"oh it can do that ? amazing" *try it* "well, apparently it can't yet" :P 21:48
(admittedly, I've had a LOOOOOT of "oh it can do that ? YES IT CAN §§")
zengargo1le that's why i haven't gone back and re-read the Apocalypses and Exegeses, too afraid they'll be so wrong by now... i just have the hazy memories of reading them some 10 years ago. :P 21:50
zengargo1le as a 'this will tell you 90% of the stuff you need to know' sort of way. 21:51
masak vendethiel: you should have been here in 2008 :P 21:52
vendethiel masak: mmh ? nothing worked :P ?
or stuff starting working feature by feature ?
masak vendethiel: "I can check if a file exists? ... no, apparently not. but I need to, because I'm implementing a wiki engine."
vendethiel when I hear about it; it's like nothing worked since 2008 :P 21:53
masak vendethiel: ended up doing `try { open } CATCH { $exists = 0 }`
:)
vendethiel I like just having gotten here.
masak 0, not False, because *Bool didn't exist*.
:)
vendethiel I can tell people "yeah, I use perl 6 in prod, only 3 months after starting looking into it"
vendethiel I really believe this has a different sound to people 21:54
dalek blets: 44c4b69 | (Herbert Breunung)++ | docs/ (3 files):
basics, intro spacing semicolon
vendethiel masak: why didn't Bool exist :o) ? 21:58
masak NYI. 21:59
vendethiel oh ...
but, if I understand it well, before 2008, most efforts when towards pugs, right ? 22:02
masak Pugs went to sleep in, hm, late 2006? Rakudo started picking up speed sometime in late 2007.
there was a year there when I felt like nothing much happened. 22:03
vendethiel So that's why. makes sense...
So much happened this year, it's really impressive ...
dalek blets: afedbff | (Herbert Breunung)++ | docs/tablet-2-basic-syntax.txt:
spaces section 30% more crispy
masak 'night, #perl6
dalek p/charrange: 88abbfe | (Tobias Leich)++ | tools/build/MOAR_REVISION:
bump moar revision for charrange
 22:37
Ven I really need to explain contexts, uh .. 22:44
m: open 'nofilehere'; CATCH { when X::IO { .say } }
camelia rakudo-moar 866817: OUTPUT«open is disallowed in restricted setting␤ in sub restricted at src/RESTRICTED.setting:1␤ in sub open at src/RESTRICTED.setting:5␤ in block at /tmp/WxMarWbDy0:1␤␤»
Ven mh, right. 22:45
Ven Where should PHASERS be ? After flow control constructs ? Before operators ? After them ? Before class ? After exceptions ? 22:50
Ven (the hardest while making a tutorial is to make everything fit at the right place !!) 22:51
(probably after exceptions, to explain keep and undo) 22:53
FROGGS after exception, yeah 22:54
Ven m: contend {} 22:57
camelia rakudo-moar 866817: OUTPUT«===SORRY!=== Error while compiling /tmp/d8YuO5SwVU␤Undeclared routine:␤ contend used at line 1␤␤»
Ven m: say FINAL { 55 } 23:01
camelia rakudo-moar 866817: OUTPUT«===SORRY!=== Error while compiling /tmp/erC84ZEpZ6␤Undeclared name:␤ FINAL used at line 1␤␤»
Ven oh, FINAL is NYI ? 23:02
Ven erm, is there an explanation on "leaving a block succesfully" ? I feel like I've been told already, but can't find it in the specs 23:09
Ven hmm, pre/post aren't really loop-related 23:12
colomon Ven: I would take "leaving a block succesfully" as meaning you are leaving the block, either from the falling off the end or a return or ... 23:15
Ven colomon: yeah, but that means with no exception
m: try { return; } 23:17
camelia ( no output )
Ven m: try { return; CONTROL {} }
camelia ( no output )
Ven m: try { return; CONTROL { default { say 1 } } } # segfaults here
camelia ( no output )
Ven maybe it was fixed in the last week. Or it's a repl fault
oh, control gets triggered by warn (but not by die) ? 23:19
Ven doesn't really understand fail vs warn vs die, and the s04 part about that doesn't have any example on throwing exceptions 23:22
Ven is COMPOSE NYI ? can't seem to find much in the roast either 23:24
Well, that's all for tonight I guess is.gd/HEHfje 23:27
g

Powered by Cro and the Raku® Programming Language.

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

Thank you!