»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! | YAPC::NA talk submissions due end of Month. Have you submitted your talk?
Set by toddr_ on 18 March 2013.
FROGGS ó.ò 00:00
I'd say make a gist but github is down somehow
hmmm, the website is fine... 00:01
timotimo: maybe you have to use nqp::getattr ? 00:02
lue FROGGS: why do you say github is down? (fwiw I got a "connection reset by peer" error from git earlier today, but it worked fine the second time)
timotimo i don't know what i'm doing
i'm just trying to make my constructor work ;_;
FROGGS lue: I cant push 00:03
timotimo: can you paste a diff
?
timotimo sprunge.us/hTCb - here's the/a code
FROGGS git pull
Connection closed by 207.97.227.239
fatal: The remote end hung up unexpectedly
timotimo: maybe it is because of the sigilless objects? 00:06
timotimo what exactly?
gist.github.com/timo/5226114#file-adt-pm6-L107 - this was the original code which worked no problem 00:07
FROGGS timotimo: forget what I said... it is the same with $Tree, $Branch, ...
timotimo the only change was how the new-foo methods are built
if you remove the "new-$name.lc()" between method and ($signature), you'll also get rid of the "useless has-foo something declaration something mumble" message 00:13
00:13 spider-mario left
FROGGS timotimo: I believe I'm too tired to be of any help 00:14
timotimo don't worry. you tried :)
dalek : c261fd3 | (Tobias Leich)++ | / (2 files):
removed Perl6::P5World
 00:15
: 3eefed5 | (Tobias Leich)++ | lib/Perl6/P5 (2 files):
added p5disect_longname to work around a missing token
FROGGS and I can push... so, gnight and goo luck
timotimo i guess i'll revert the change. well i don't really have to, i can just get it back from the gist
good night :)
FROGGS ya, a bit
timotimo although i think i want to keep this for later when i figure out how to make it work
01:44 Liz joined 01:45 Liz is now known as Guest1410 01:46 Chillance left 01:51 Guest1410 is now known as lizmat 01:57 fgomez joined
dalek osystem: 749b9da | timo++ | META.list:
add the amazing ADT module.
 02:08
colomon woah, what 02:09
diakopter timotimo++
timotimo ☺ 02:10
diakopter who's Tobias
timotimo i enjoy praise, so i do funny stuff
FROGGS is tobias
funny stuff like work
diakopter FROGGS: disect is spelled with two s
timotimo colomon: hopefully i didn't put your hopes up too far, it doesn't do very good compile-time stuffs yet 02:12
sorear timo is not FROGGS, FROGGS is FROGGS on github
istr timo is supernovus
diakopter O_O 02:13
sorear oh, you were replying to something earlier
02:13 xinming joined
timotimo does supernovus just mean "very new"? 02:15
diakopter male of supernova? 02:17
timotimo so i'm exploding?
sorear supernovus is somebody's handle 02:18
timotimo oh 02:19
no, that timo is me
"Timothy" supernovus is supernovus on github 02:23
sorear totten or such 02:38
timotimo indeed
comments about the ADT module?
02:44 FROGGS_ joined 02:48 FROGGS left
dalek rl6-roast-data: 91961dd | coke++ | / (3 files):
today (automated commit)
 02:48
02:48 dwoldrich_ joined 02:50 dwoldrich_ is now known as dwoldrich 03:01 Targen joined 03:08 orafu joined 03:10 Targen left 03:11 Targen joined 03:39 preflex_ joined 03:40 preflex_ is now known as preflex 03:43 rindolf joined 03:46 renormalist left 03:53 renormalist joined 04:31 thou left 04:54 dwoldrich left 04:58 thou joined 05:03 colomon left 05:08 quester joined
quester rnp: printf "%d" 05:17
p6eval pugs: OUTPUT«(timeout)» 05:18
..rakudo 221a95: OUTPUT«Null PMC access in get_integer()␤ in sub sprintf at src/gen/CORE.setting:2471␤ in sub printf at src/gen/CORE.setting:2476␤ in block at /tmp/HOqXzOn375:1␤␤»
..niecza v24-35-g5c06e28: OUTPUT«Unhandled exception: Not enough arguments for this format string␤ at /home/p6eval/niecza/lib/CORE.setting line 1370 (sprintf @ 4) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1371 (printf @ 5) ␤ at /tmp/csUgPBAymx line 1 (mainline @ 3) ␤ at /home/p6e…
05:20 toddr_ joined
quester rnp: printf "%s" # rakudo bug #62316, closed in Aug 2009 05:21
p6eval pugs: OUTPUT«(timeout)»
..rakudo 221a95: OUTPUT«Null PMC access in get_string()␤ in sub sprintf at src/gen/CORE.setting:2471␤ in sub printf at src/gen/CORE.setting:2476␤ in block at /tmp/Bu7zOuzh1Y:1␤␤»
..niecza v24-35-g5c06e28: OUTPUT«Unhandled exception: Not enough arguments for this format string␤ at /home/p6eval/niecza/lib/CORE.setting line 1370 (sprintf @ 4) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1371 (printf @ 5) ␤ at /tmp/3oBC6P0NjI line 1 (mainline @ 3) ␤ at /home/p6e…
quester submitted it as rakudo bug #117317 05:46
06:01 SamuraiJack joined
moritz if a bug report already existed, why not reopen that one? 06:20
06:24 dwoldrich joined
quester moritz: The old bug report, #62316, had been closed for two a half years. I would suspect the bug itself is likely new, even if it does have the same symptoms. 06:26
moritz quester: if you don't know it, it's pure speculation. And if you don't know what's going on, you have to go after the symptoms, not the causes 06:27
anyway, I can merge the two
quester Ah, I see your point... Okay, thank you. 06:28
06:49 am0c joined 07:06 quester left 07:27 kivutar joined 07:43 drbean left 07:44 drbean joined 08:00 cognominal left 08:04 cognominal joined 08:08 toddr_ left
FROGGS_ diakopter: thanks for pointing out the typo, my mind was already in bed at that time 08:23
08:23 FROGGS_ is now known as FROGGS
diakopter oh; was it just typo in the commit message? I didn't check 08:25
08:26 dwoldrich left
FROGGS diakopter: no, even within the rakudo repo it is spelled that way: github.com/rakudo/rakudo/blob/nom/...d.pm#L1906 08:27
I just copy-pasted that
diakopter ah
moritz ah yes, I've stumbled over that once or twice before 08:39
08:40 lizmat left
FROGGS moritz: should I fix that? 08:41
08:41 Liz joined 08:42 Liz is now known as Guest91638 08:43 Guest91638 is now known as lizmat
moritz FROGGS: yes 08:43
08:45 labster left
dalek : 4ae984e | (Tobias Leich)++ | / (5 files):
revive Perl6::P5World
 08:54
: 85bbba6 | (Tobias Leich)++ | t/v5/ (37 files):
pull in Perlito test files
FROGGS will push to nom after spectest 08:55
09:07 domidumont joined
dalek kudo/nom: b972ca3 | (Tobias Leich)++ | src/Perl6/ (3 files):
fixed spello as diakopter++ pointed out
 09:08
FROGGS sad there is no RT ticket for that :o) 09:09
dalek rlito: 0ff7ea2 | (Flavio S. Glock)++ | src (2 files):
Perlito backend - stub code for Perlito6::X64::Assembler, Perlito5::X64::Assembler
 09:15
09:16 spider-mario joined
dalek : 5bfeb4a | (Tobias Leich)++ | / (2 files):
fix infix:sym<,>
 09:21
FROGGS std: use v5; sub a ($x, $y) { 42 } 09:30
p6eval std 86b102f: OUTPUT«Potential difficulties:␤ $x is declared but not used at /tmp/feB3Cf_b8v line 1:␤------> use v5; sub a ($x⏏, $y) { 42 }␤ $y is declared but not used at /tmp/feB3Cf_b8v line 1:␤------> use v5; sub a ($x, $y⏏) { 42 }␤ok 00:01 52m␤…
FROGGS std: use v5; sub a ($x, $y) { $x + $y }
p6eval std 86b102f: OUTPUT«ok 00:01 53m␤»
FROGGS std: use v5; sub a ($x, $y) { $x + $y }; print a(2, 4) 09:31
p6eval std 86b102f: OUTPUT«ok 00:01 53m␤»
FROGGS I expected: Malformed prototype for main::a: $x,$y at -e line 1.
09:32 domidumont left 09:33 domidumont joined 09:45 domidumont1 joined 09:46 thou left 09:47 domidumont left
masak good forenoon, #perl6 09:55
tadzik hello masak 09:56
FROGGS hi masak
moritz \o masak, tadzik, FROGGS, * 09:59
lizmat good *, all! 10:00
10:01 domidumont1 left 10:02 domidumont joined
lizmat is in lyon recovering from a chartreuse night 10:02
moritz \o lizmat 10:05
sorear o/ all 10:07
what's a chartreuse night
tadzik it's the one when you go to beed sober and wake up drunk
moritz iirc chartreuse is some sort of wine 10:08
tadzik or that's what I remember from FPW
sorear imagines a chartreuse aurora over .nl that makes people who see it ill 10:09
masak that pretty much sums it up. :) 10:11
huf reusing charts... this is what the economy has come to?! 10:12
moritz doesn't think lyon is in .nl :-)
huf someone could be lyon here :)
(eh, not really0
masak the edit distance between 'lyon' and '.nl' isn't horrible... :P 10:13
*lol*, 5.2.0 ended up being called "Stuffed Parrot" :P 10:15
moritz larger than between tcl and t2 :-)
masak :P
moritz lol
masak and the quote is from Spamalot, "I am not dead yet"!
Util++ 10:16
lizmat fwiw: en.wikipedia.org/wiki/Chartreuse_(liqueur)
masak oh, dang. rt.perl.org/rt3/Ticket/Display.html?id=117317 makes me want to hurry up with my sprintf implementation.
lizmat Chartreuse Night is a tradition started by Philippe 'book' Bruhat, usually happening one night on YAPC::Europe 10:17
masak I got pretty far into the, hm, first 10% or so.
lizmat it involves several bottles of Chartreuse and other liquors, a place near a hotel, and N number of perl geeks 10:18
last night, N was 3 10:19
masak but... it's green!!! o.O
lizmat and the number of bottles was also 3
(albeit there were not all full to start with)
masak and not just mildly green, but kind of... radioactively so.
lizmat indeed.. "the green monster" it is also sometimes called
masak why does that sound like an understatement? :P
lizmat although personally, I like to reserve that for "Verveine du Velay" 10:20
masak "It is composed of distilled alcohol aged with 130 herbal extracts."
you people are crazy!
:P
what if one of those herbs are dangerous?
tadzik one?!??
that'd be softcore ;)
masak heh. 10:21
lizmat also note, the green Verveine is 55% :-)
sorear if it's under any selective pressure not to be eaten, it's probably poisonous. ;) 10:22
tadzik that's a "wow" even by my standards :)
lizmat makes mental note to take a bottle to the Polish Perl Workshop
tadzik oh-o 10:25
that's gonna be fun :)
(un)fun fact: drinking on streets is forbidden in Poland 10:28
masak tadzik knows this the hard way. ;)
tadzik yeah :) 10:29
huf is it enforced? 10:32
masak seems so.
huf we have this law too, and it's one of the ones the police overlook unless they have some other reason to bother you 10:33
tadzik yeah, it is
I ignored the 100PLN ticket the police gave me
huf like, if you're loud, homeless, nonwhite, whatever, they'll stop you from drinking, but otherwise go ahead :)
tadzik after 3 months I got a snail-mail letter which was a copy of the letter they sent to my bank
huf huh.
lizmat is checking out, offline until probably later tomorrow 10:34
10:34 lizmat left
huf come to think of it, being loud here doesnt mean this law will get enforced... :) 10:34
tadzik it ordered the bank to freeze my account temporarily, transfer the amount on the ticket + some additional fee to the police, then unfreeze my account again
(or keep it frozen if there's not enough mana^Wmoney there)
so yeah, they don't fsck around ;)
huf huh. shoudl've used a brown paper bag i suppose ;)
tadzik haha 10:35
maybe I should've talked them away from it, instead of continuing to work on my laptop with my friend and explaining the Chomsky hierarchy and why regexes can't parse html 10:36
sorear
.oO( regex which parses real-world html: .* )
moritz sorear: works only in Perl 6, in Perl 5 you need an /s :-)
10:37 domidumont left 11:03 eternaleye_ joined 11:04 preflex left 11:05 preflex joined 11:06 dmol joined 11:13 SamuraiJack left
masak today's mini-challenge: set up a Perl 6 program that responds to Ajax requests. 11:18
arnsholt Did anything ever happen to the project to revive Emmentaler? 11:23
masak I saw some activity in that region, led by (I think) sergot. but it's a while ago. 11:24
tadzik I think it's just cronjobs that aren't running 11:30
well, it's a bit basic right now too 11:31
but was mostly working at feather.perl6.nl/~sergot/modules/
masak sergot++
ecosystem testing is going to be doubly important once we're on the JVM, too.
simply to hold together the implementations and backends.
11:37 PacoAir joined, preflex left 11:39 preflex joined 11:41 SamuraiJack_ joined
arnsholt Yeah, I had a thought as I put off getting out of bed today 11:57
Essentially, I'd like to set up something largely similar to what I saw in use at Opera 11:58
tadzik which thing
?
arnsholt The Spartan infrastructure 11:59
tadzik hmm
arnsholt The code driving it is kinda awful (I've seen it =), but the actual functionality is pretty awesome
Of course, a volunteer-driven project like this needs something a bit different, since we can't just decide to set up a bunch of slaves to do testing all the time, but the idea is sound I think 12:00
My first plan was setting up some code that automatically fetches and builds Rakudo from git (and maybe NQP as well) 12:02
Then run the test suite and spectest suite and capture the TAP stream. Parse it and extract data, plus render it to HTML for easy inspection
Once that works, integrate with Panda to do the same for modules 12:03
12:15 _sri left 12:16 _sri joined 12:27 kivutar left 12:31 Chillance joined
dalek osystem: 5db057e | mrhdias++ | META.list:
Was added the perl6-GD module.
 12:38
osystem: 9626e4d | moritz++ | META.list:
Merge pull request #23 from mrhdias/master

Add the perl6-GD module.
12:46 am0c left
DrEeevil arnsholt: sounds like a great plan. I'm willing to spend some time and resources on helping out with it 12:46
arnsholt Cool! 12:47
I think the initial version should mostly be Perl 5 (and possibly some Makefiles. Makefiles are cool =) 12:48
DrEeevil yeah, sh+make+perl5 sounds good
keep it as simple as needed
12:49 SamuraiJack__ joined, SamuraiJack_ left
arnsholt Hmm. Actually, I should probably not use too much make (or sh) 12:57
The easier it is to run it on Windows, the better
13:01 crab2313 joined 13:02 SamuraiJack__ left 13:04 SamuraiJack joined
masak make runs on windows. 13:12
you just can't use all the nice GNU syntax.
FROGGS ya, well, you can have two Makefiles...
a dmake one and a gnu make one
arnsholt Well yeah, but I don't have a Windows machine, so I won't really know if it works properly on Windows or not =) 13:14
13:14 SamuraiJack_ joined
FROGGS well then somebody has to port your makefile to windows :o) 13:14
13:14 SamuraiJack_ left 13:15 SamuraiJack left
arnsholt True, true 13:15
13:15 SamuraiJack_ joined 13:25 crab2313 left
masak rn: my $i = 0; until $i = 100_000_000_000 { $i++ }; say now - INIT now 13:30
p6eval niecza v24-35-g5c06e28: OUTPUT«Unhandled exception: Cannot use value like Instant as a number␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/lib/CORE.setting line 293 (Any.Numeric @ 6) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /tmp/iFxciPYgYz line 1 (mainline @ 1…
..rakudo b972ca: OUTPUT«0.033560␤»
masak :) 13:31
moritz rn: my $dummy = do for 1..100_000_000_000 { }; say now - INIT now 13:35
p6eval niecza v24-35-g5c06e28: OUTPUT«(timeout)Potential difficulties:␤ $dummy is declared but not used at /tmp/kPIZm1Ea1h line 1:␤------> my ⏏$dummy = do for 1..100_000_000_000 { }; ␤␤»
..rakudo b972ca: OUTPUT«0.0366133␤»
dalek : 03f3283 | (Tobias Leich)++ | lib/Perl6/P5Grammar.pm:
comment out term "shift"
 13:37
: 7ecd591 | (Tobias Leich)++ | / (2 files):
added implicit slurpy @_ parameter to subs
13:38 SunilJoshi joined 13:54 pmurias joined
DrEeevil people still care about OS/2 and Windows? :) 13:59
FROGGS :o) 14:00
masak DrEeevil: you greatly underestimate the number of lost souls out there, and the work ahead. :) 14:02
DrEeevil masak: my last installation of the redmond gameloader died in 2007 due to hardware upgrade
I just don't see a reason to use inferior tools :) 14:03
masak indeed. and that's also not the point where I was arguing. :)
DrEeevil well, yeah, at my current workplace there's still some windows being used because lots of people write nonportable code 14:04
we almost had to aquire new hardware for a legacy app because of a customer
masak pesky customers. 14:05
DrEeevil well, it'd have needed more hardware than our current platform to serve less users too 14:06
which makes me very sad, because that is already inefficient code that we just throw more hardware at 14:07
14:08 skids joined
arnsholt <3 cpanm 14:09
pmurias FROGGS: Perl6::P5World is calling a non-existant dissect_longname method
shouldn't Perl6::P5Grammar be called Perl5::Grammar? 14:10
FROGGS pmurias: no, you need latest rakudo/nom 14:11
there was a typo I fixed
14:13 eternaleye_ is now known as eternaleye
pmurias FROGGS: there's no dissect_longname method in Perl6::World 14:16
moritz commit b972ca3987a945fcca68273a996889d18f50e036
Author: Tobias Leich [email@hidden.address]
Date: Sun Mar 24 10:07:47 2013 +0100
fixed spello as diakopter++ pointed out
that's the commit that renamed it from disect to dissect 14:17
pmurias ahh, I have cloned a fork :( 14:19
FROGGS not my fault \o/ 14:20
moritz forks a clone 14:21
14:23 Psyche^ joined 14:27 Patterner left, Psyche^ is now known as Patterner
masak forks a dongle 14:31
DrEeevil omg 14:33
DrEeevil adds extra drama
tadzik haha 14:36
I'm offended!
tadzik takes a picture
Patterner that won't work here. no girls^Wwomen^Wwopersons^Wwepersiblings 14:41
14:42 census joined
dalek : 450a0e0 | (Tobias Leich)++ | lib/Perl6/P5 (2 files):
implement sigil-change (@a, $a[0])
 14:44
14:45 pmurias left
DrEeevil tadzik: be careful, the last person that did that got fired ;) 14:48
14:52 pmurias joined 15:02 census left 15:03 colomon joined 15:13 kivutar joined 15:26 SunilJoshi left
moritz fires a forked dongle at masak 15:26
masak Patterner: I'm offended by your assertion that there aren't any females here... ;) 15:29
there's at least a handful. I won't call them out.
moritz now that ~/.perl6/lib is gone from @*INC, I can try to set up a production rakudo that's completely separate from the dev setup 15:31
GNAAAA 15:36
now I'm running into the "Method 'seek' not implemented for type Socket" parrot bug again
because at the time of rakudo's release, there was no new parrot release, so we're still stuck with the old version :( 15:37
dalek rlito: d7f0089 | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - add t5-x64/01_sanity.t
 15:43
rlito: 810d6f3 | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - t5-x64/01_sanity.t runs in perl and node.js
hoelzro ahoy #perl6 15:44
15:45 domidumont joined
masak hoelzro! \o/ 15:47
hoelzro almost done with the pygments highlighter! 15:48
masak yay
hoelzro it now highlights Grammar.pm perfectly =)
tadzik whoa
hoelzro just need some more tests 15:49
tadzik hoelzro++
hoelzro and make sure that some corner cases work 15:50
and handle filetype detection =/
on that note...is 'use strict' valid Perl 6?
tadzik if there's 'strict.pm' somewhere..
masak hoelzro++
hoelzro I wonder if I find 'use strict' in a file, if I should tell pygments to consider it as Perl 5 15:51
tadzik: of course =)
moritz n: use strict; say 42
p6eval niecza v24-35-g5c06e28: OUTPUT«42␤»
hoelzro damn
oh well
moritz hoelzro: rakudo doesn't support it yet, but there's supposed to be 'no strict' and 'use strict;' in Perl 6 too
hoelzro I see 15:52
15:52 xinming left
hoelzro well, I guess that's off the table 15:52
15:53 xinming joined 15:56 census joined
rindolf Hi hoelzro, moritz , masak , tadzik . What's up? 15:59
Isn't "use strict;" superfluous in Perl 6?
hoelzro hello rindolf
just talking about the pygments highlighter I'm working on =)
moritz rindolf: not on -e one-liners
rindolf moritz: ah.
hoelzro after this, it's on to the next thing =)
which I think will be improving Perl 6 packaging 16:00
dalek rlito: b6c2983 | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - more internal functions
 16:08
moritz meh. ilbot6 won't join channels :( 16:09
arnsholt Whoops
At least you're at the point where you can run your program and see what happens =) 16:10
16:12 ggoebel_ joined
masak for the record, I didn't expect the discussion with allison on #parrot to get heated. I'm sorry about that. 16:14
it was kind of a slippery slope into all-out disagreement and talking past each other.
arnsholt I didn't see it coming either, FWIW =/ 16:15
masak ironically -- in a "ha ha not funny" kind of way -- the structure of the conversation mimics the structure of the teams. it's the same inability to work together that manifests on both levels. 16:16
arnsholt That's a good point
moritz that's why a stable VM has such a big appeal
masak how the heck did we end up with a Parrot that does not have the best interests of Rakudo/Perl 6 at heart? :(
moritz it doesn't require "us" to talk to "them", because "their stuff" simply works 16:17
masak and whose *leaders* then complain at us for not having any users?
makes me want to cry.
moritz masak: former leader, fwiw
masak: I haven't heard such FUD from cotto
masak troo. 16:18
16:37 jeffreykegler joined, raiph joined
dalek rlito: 61a5f3c | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - add the registers
 16:43
16:46 fglock joined
fglock hi 16:46
I'm adding a jit to perlito 16:47
moritz does that mean you're also adding a native backend?
or is there one already?
jeffreykegler Question: Perl 6 is a family of languages, not just a language, I am told. Where is the most extensive write-up of that? 16:48
moritz jeffreykegler: that's news to me
jeffreykegler: there are several compilers which all implement slightly different subsets of Perl 6 16:49
jeffreykegler: but I wouldn't call that a family of languages
jeffreykegler moritz: larry told me this once, on this channel, some years ago
... and the same idea appears in the Perl tablets
moritz oh
I think what he meant is that there are several "slangs"
fglock moritz: yes, I'm working on Perlito5::X64::Assembler now
moritz that is, regexes are a different "language" than the mainline 16:50
fglock I started Perlito5::X64::Assembler, but I can just cross-compile perl5 to perl6
moritz and quotes are yet another language
so yes, in that sense, one can talk about a family of language
s
jeffreykegler: sorry, I misunderstood at first
jeffreykegler Yes, I think the slangs, with the main language, were what Larry called the "family of languages"
fglock I mean, Perlito6::X64::Assembler 16:51
sorry
moritz jeffreykegler: S02 has a bit on slangs
jeffreykegler: perlcabal.org/syn/S02.html#Slangs
cognominal rn: class A { has $!a; method new($a) { A.bless(A.CREATE(), a => $a) }; method gist { say $!a }}; say A.new(42) 16:52
p6eval rakudo b972ca, niecza v24-35-g5c06e28: OUTPUT«(Any)␤True␤»
cognominal hum
jeffreykegler mortiz: thanks. Anything more extensive? I'm interested in theory & ideas as much (or more) than implementation 16:53
cognominal moritz, what is wrong with my online? 16:55
moritz cognominal: only public attributes are initialized by default 16:56
cognominal: doc.perl6.org/language/classtut#Constructors
cognominal thx 16:57
moritz sorry abou the formatting being a mess
cognominal anyway being silent about passing a private attribute is probably wrong 16:59
moritz allowing a public constructor to initialize a PRIVATE attribute by default would be wrong 17:00
it would mean that private attributes can be written (but not read) from the outside by default
17:01 ilbot6 joined
cognominal yes, I understand the rational 17:01
moritz I'm afraid that one way or another, we have a FAQ at hand 17:04
cognominal I am kind of rusty in that OO stuff... 17:06
arnsholt moritz: What was the root of your socket problem, OOC? 17:07
moritz arnsholt: a for loop that wasn't sunk properly :( 17:08
arnsholt Right. Which makes it unexpected laziness, right?
moritz correct 17:09
arnsholt Oh, right. This is why a for loop at the end of a sub will behave unexpectedly, isn't it?
cognominal I try to reconstruct my original line of thought. I suppose I considered bless sort like a private method. 17:10
moritz arnsholt: in this case it's simply never executed :(
arnsholt Yeah
moritz is so >.< close to making for-loops eager again
arnsholt I'm somewhat embarassed to note that I spend most of my "Perl 6" time programming in C, and not Perl 6 =) 17:11
masak arnsholt: :) 17:14
moritz: that is my feeling too. for loops have always "felt" eager to me.
dalek rlito: 0e13f14 | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - add push/pop
arnsholt I do program a bit of NQP, though. It's almost Perl 6, but not quite
jeffreykegler Btw, here's the exchange I was referring to at 9:48am: irclog.perlgeek.de/perl6/2008-02-03#i_191262 I note in that exchange, that Larry does not use the phrase "family of languages"; he says Perl 6 will be "many languages" 17:19
arnsholt twiddles his thumbs and waits for the compiler to compile itself 17:20
Progress! 17:24
17:30 ilbot6 left
masak I do not want to continue the discussion on #parrot. but it strikes me that regardless of how many users Perl 6 has, Parrot could have had excellent HLL interop, admirable GC, and wonderful thread support. it doesn't. 17:41
and I don't see what the fact that it doesn't has to do with the success or failure of Perl 6. 17:42
arnsholt In happier news, I think my vmlist NQP branch is close to compiling itself again =) 17:44
masak yay 17:45
17:45 renormalist left
dalek rlito: 073e847 | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - more push/pop
 17:46
arnsholt Yeah, I'm pretty happy about that
For a while I was unsure how to go about the HLL config stuff, but after some Parrot source-diving, it was pretty straightforward
17:51 thou joined
cognominal Today, I am seeing parrot as a way to bootstrap rakudo. Thx to jnthn recent efforts, we are close to see other VM as backend. And even if in TimToady mind, Perl 6 is probably what Perl 5 ought to be, I don't see any interest thinking one in term of the other. some Perl 5 habits translate to Perl 6, that's nice. Some Perl 6 ideas have been implemented in Perl 5. that's cool. But thinking one in term of the other is just creating a bigger ghetto. 17:51
arnsholt Gah. Stuff I don't understand in the regex code
17:52 ilbot6 joined
arnsholt ilbot6 \o/ 17:53
moritz it even connects to mysql
arnsholt moritz++ # ilbot6
17:54 ilbot6 left
moritz it just doesn't log yet :( 17:54
arnsholt But progress!
I have a long-running NQP branch that hasn't been able to compile itself for several weeks. Progress is always good =)
moritz aye :-) 17:55
TODO: implement method ping in DBDish::mysql::Connection 17:58
17:59 ilbot6 joined
moritz strike previous TODO. New TODO: test DBDish::mysql::Connection.ping 18:00
arnsholt \o/
18:02 ilbot6 left
dalek rlito: 27f3c4b | (Flavio S. Glock)++ | src5/lib/Perlito5/X64/Assembler.pm:
Perlito5::X64::Assembler - add scale factors for operands
 18:03
Iish: 60438cf | moritz++ | lib/DBDish/mysql.pm6:
implement ping method for mysql backend
 18:08
18:12 bloonix joined
moritz r: say now 18:15
p6eval rakudo b972ca: OUTPUT«Instant:1364148985.722886␤»
moritz r: say now.Int
p6eval rakudo b972ca: OUTPUT«1364148992␤»
masak r: sub what-apout { say @_ }; what-apout now 18:16
p6eval rakudo b972ca: OUTPUT«Instant:1364149015.075067␤»
masak :P
moritz how do I get today's date, but from the perspective of UTC, not localtime? 18:17
18:19 ilbot6 joined, ilbot6 left
dalek : ff19c0c | (Tobias Leich)++ | / (3 files):
in the middle of allowing to have named iterators in for loops
 18:19
ruoso commute &/win 4 18:20
ruoso not looking when typing 18:21
18:25 ilbot6 joined, ilbot6 left
masak IRC: not looking when typing since 1988. ;) 18:33
dalek rlito: 54135fa | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - add some mov variants
 18:34
ast: db2d159 | coke++ | S0 (6 files):
pugs fudge
rl6-roast-data: 5fc0f95 | coke++ | / (3 files):
today (automated commit)
18:38 cognominal left
dalek rlito: d2b88ac | (Flavio S. Glock)++ | src5/lib/Perlito5/X64/Assembler.pm:
Perlito5::X64::Assembler - add repmovs variants
 18:40
18:43 ggoebel_ left 18:44 cognominal joined 18:45 cognominal left, cognominal joined
dalek rlito: c50d5a9 | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - add shld, shrd, rdtsc
 18:49
p/vmarray-list: 2de3300 | (Arne Skjærholt)++ | src/6model/reprs/VMArray.c:
Implement copy_to in VMArray.
 19:02
nqp/vmarray-list: ee613a6 | (Arne Skjærholt)++ | src/ (2 files):
nqp/vmarray-list: Remove dependencies on get_number VTABLE for nqp::list.
arnsholt *sigh* That does tend to happen, doesn't it?
19:03 dalek joined, ChanServ sets mode: +v dalek
dalek rlito: ab5f52f | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - add xchg
 19:05
19:07 skyheights joined 19:10 benabik joined
skyheights liked the discussion earlier on #parrot as both sides articulated valuable points. 19:10
19:15 dmol left 19:18 skyheights left
masak skyheights: guess so. it was hard to be a part of, though. I could sense messages missing their target in both directions. :/ 19:18
19:21 denisboyun joined
arnsholt Hmm. Suddenly I'm sad jnthn++ is away on holiday 19:23
hoelzro ok, all that's left is file type detection.
19:24 dmol joined
tadzik hoelzro: are you looking for 'use vSomething'? 19:24
hoelzro yup
tadzik vim comments?
hoelzro right now, I only look for shebang lines and 'use v6'
modelines did occur to me
tadzik std: package F { }; 19:25
p6eval std 86b102f: OUTPUT«ok 00:00 42m␤»
tadzik std: package F; 1;
p6eval std 86b102f: OUTPUT«===SORRY!===␤Semicolon form of package definition indicates a Perl 5 module; unfortunately,␤ STD doesn't know how to parse Perl 5 code yet at /tmp/oGdn1lIiEw line 1:␤------> package F⏏; 1;␤ expecting any of:␤ name␤ trait␤Parse…
tadzik well, these days Perl5 has the former form too
arnsholt IIRC top-level module, class and grammar declarators are pretty reliable Perl 6 indicators as well
hoelzro yeah, I thought of that too =/ 19:26
arnsholt: pretty much, yes
but with the advent of Moose and MooseX::Declare...
it makes things a little blurier
arnsholt Ah, right
hoelzro right now, I'm working on building a corpus of p5/p6 code to test against 19:27
tadzik 'module' and 'grammar' are still Perl6-only, I think
hoelzro: what if Perl 6 pygments somehow subclassed Perl 5 pygments? Could it work for both? 19:28
hoelzro yes, but grammar isn't in every Perl 6 file =/
yikes
I think that would cause more harm that good
well, for starters, I know it wouldn't work
19:28 ilbot6 joined
hoelzro the Perl 6 lexer needs to be of type ExtendedRegexLexer 19:28
because I have some special logic that can't be expressed in just regexes 19:29
moritz ilbot6 now logs into a database
tadzik whoa
moritz++
arnsholt hoelzro: Yeah, grammar won't catch everything, but it'll be a pretty reliable indicator 19:30
moritz I found a bug in the mysql DBDish driver along the way
grammar and 'use v6;' are both very reliable 19:31
hoelzro true 19:32
19:34 pmurias left 19:35 pmurias joined
masak moritz++ 19:40
moritz: was it the #parrot discussion that made you go work on ilbot6? :) 19:41
19:41 ilbot6 left
moritz masak: no, I worked on it before the #parrot discussion 19:41
and it was the reason I didn't want to waste my time on pointless discussions, because I had better to do :-)
dalek Iish: 3790014 | moritz++ | / (3 files):
fix quoting in mysql

the old mechanism did not consider undefined values NULL, and failed for empty strings and for stuff that looked vaguely like numbers, but were not well-formed numbers		 19:42
pmurias fglock: hi 19:43
fglock: what do you plan to use for garbadge collection?
hoelzro ugh. 19:45
tests tend not to have grammar.
and I'm not seeing a lot of use v6 there, either
tadzik hmm
'use Test'
hoelzro =(
tadzik in Perl 5 it's always (I think) Test::Simple, Test::More or so 19:46
hoelzro I might just favor Perl 5
tadzik that too
hoelzro if you want your tests/files probably highlighted, 'use v6'
modules will probably be fine
tadzik yes, I'd lean to that
hoelzro I'd rather incorrectly highlight 6 as 5 rather than 5 as 6 19:47
19:47 Targen left
masak a heuristic that would work quite well, I think, would be looking for any occurrence of '$obj.method' in the code. Perl 5 programs don't tend to have that. 19:48
pmurias hoelzro: one thing you should look "# vim: ft=perl6" 19:49
* look for
fglock pmurias: for perl6 I'll look for an existing dynamic loadable library, if possible
masak or many other Perl 6-specific idioms, such as 'map { ... }, @note_the_comma'
but those things would remain heuristics, of course.
fglock for perl5, it needs reference counting
masak bet there are good ones, though :)
19:50 ilbot6 joined
fglock pmurias: so it needs to be implemented closer to the emitter 19:50
pmurias so you plan to work on both Perlito and Perlito5? 19:51
19:51 ilbot6 left 19:52 Targen joined
moritz meh, the mysql backend doesn't like real-world data 19:53
fglock yes
lue wonders if there's a nice summary mode for this #parrot discussion on perlgeek 19:54
19:55 Guest89482 joined 19:58 colomon left
arnsholt I think you'll just have to read the log =) 19:59
19:59 jeffreykegler left, cognominal left
arnsholt moritz: Since you're working on DBIish, I'll warn you that I hope to make some changes to how strings are handled in Zavolaj at some point in the (hopefully) not too distant future 20:00
moritz arnsholt: the ownership stuff?
arnsholt Just have to get all my yaks in a row
The ownership stuff too
But also marshalling between CArrays and Bufs 20:01
So that string stuff that works like foo(char *data, size_t datalen) can decode the Str into a Buf and create a CArray from that 20:02
20:02 Guest89482 left
arnsholt Since the somewhat naive approach we have now breaks horribly with Unicode in cases like that 20:02
After that works I'll do the ownership stuff 20:03
20:06 cognominal joined
arnsholt But first I have to get this vmarray yak shaved, which might block on me talking to jnthn 20:07
[Coke] rakudo is still green. pugs will be green tomorrow - that leaves niecza with 18 failures. 20:10
hoelzro pmurias: I was thinking it, yeah 20:11
masak: so just /\$\w+\.\w+/? 20:12
moritz [$@%&]\w+\.\w+\( 20:14
20:16 SamuraiJack_ left
dalek rlito: 0be697d | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - add labels
 20:16
masak well, the paren is optional, but yes.
hoelzro twigils are also a good giveaway 20:20
moritz though note that $foo.bar() is also valid Perl 5 syntax 20:21
a concatenation of a variable and a method call
20:21 domidumont left
masak sub call, yes. 20:23
maybe better to search for the absence of '$foo->bar()', actually.
std: my $foo; $foo->bar() 20:24
p6eval std 86b102f: OUTPUT«===SORRY!===␤Unsupported use of -> as postfix; in Perl 6 please use either . to call a method, or whitespace to delimit a pointy block at /tmp/DHrv_gCKit line 1:␤------> my $foo; $foo->⏏bar()␤Parse failed␤FAILED 00:00 42m␤»…
arnsholt Good idea. -> is an excellent discriminator for Perl 5 20:25
lue
.oO(This parrot discussion looks interesting. Too bad it seems to have ended)
 20:28
dalek rlito: 7b6df97 | (Flavio S. Glock)++ | src5/lib/Perlito5/X64/Assembler.pm:
Perlito5::X64::Assembler - add conditions
arnsholt hoelzro: I guess what you want is a set of heuristics that are highly specific in both directions 20:31
20:32 kaare_ joined
FROGGS well, you could define rules which vote for either Perl 5 or 6 and then if it is Perl5 0.18 <=> 0.72 Perl6 you know what to choose 20:32
so just put these rules/guesses in a list 20:33
... even if they are just good guesses
masak yeah. basically a Bayesian filter. 20:34
www.paulgraham.com/spam.html
20:35 ilbot6 joined, ilbot6 left, kivutar left
arnsholt masak: I was thinking decision tree, but yeah 20:35
FROGGS: Actually, I should have thought of that 20:36
moritz meh. ilbot6 segfaults when inserting stuff into a postgresql database :(
arnsholt hoelzro: If you've collected a decent-ish corpus of Perl 5 and Perl 6, you could consider training a Naive Bayes model to discriminate
lue
.oO(If you go Bayesian, you might be able to set up some sort of temporary pygments-using webpage that goes "Upload a Perl 6 or Perl 5 file!" ==> Upload file ==> "We detected this is a Perl 6 (or Perl 5) file. Is this correct?" ==> answer yes/no, to improve the filter)
20:37 ilbot6 joined, ilbot6 left
arnsholt FROGGS++ # Reminding me of stuff I should've thought of =) 20:37
masak lue: yes, but that would be yak shaving.
since the goal is to syntax-highlight code. :) 20:38
timotimo if we have a corpus of the same problem solved in perl5 and perl6, we could build a machine translation system :D
arnsholt I'm not sure how well SMT would work for programming language translation =D 20:39
OTOH, you could consider compilers (terribly complicated) RBMT systems ^_^
timotimo :)
lue masak: But how would you improve (quickly|easily) the bayesian filter on github if it gets things wrong? :)
arnsholt moritz: Segfaults are no fun. Tell me if it's my faualt =) 20:40
moritz arnsholt: I have no idea 20:43
20:44 rindolf left 20:46 cognominal left
FROGGS gdb? 20:48
masak lue: manually, on a case-by-case basis, I suppose.
moritz jnthn++ # very nice syntax errors messages, with accurate line numbers 20:49
hoelzro I would prefer to avoid something that requires explicit teaching 20:50
I want the logic to be visible
arnsholt In that case, a decision tree might be the best way to go
20:50 kurahaupo joined
moritz or simpler 20:50
arnsholt A cascade of highly discriminating choices is essentially a decision tree, FWIW
moritz use v6;| grammar | module | # vim: ft=perl6 => Perl 6 20:51
otherwise: Perl 5
hoelzro that makes the most sense 20:52
and it's in line with S01
20:52 ilbot6 joined
FROGGS well, this might catch these Perl6ish modules from cpan that have these keywords in pod 20:52
moritz hoelzro++ # p6 hilighting
hoelzro FROGGS: nice idea 20:53
moritz one can check 'use v6;' on the start of a line
hoelzro I'm checking for that at the moment
moritz r: say ?42
p6eval rakudo b972ca: OUTPUT«True␤»
20:53 ilbot6 left
hoelzro the grammar/module thing would be nice as well 20:53
the trick about *that* is proper implementation =) 20:54
20:54 ilbot6 joined
moritz r: say ?0 20:55
20:55 ilbot6 left
p6eval rakudo b972ca: OUTPUT«False␤» 20:55
arnsholt hoelzro: I'd urge you to try some naive Bayes models though. They're pretty easy to implement, and efficient 20:56
Definitely one of the very best bang-for-buck machine learners
pmurias hoelzro: at the start of a line with a semicolon at the end of the line? 20:57
hoelzro pmurias: is that re: grammar/module?
pmurias hoelzro: yes 20:58
hoelzro but that wouldn't match 'grammar MyGrammar {', right?
pmurias so with a { at the end of the line 20:59
21:00 dmol left
FROGGS my grammar FooBar is WhateverGrammar does Something[Int] \n { 21:00
21:01 dmol joined
hoelzro ah 21:01
dalek rlito: 84ad906 | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - add "jmp" stub; add V8 copyright message
 21:02
FROGGS hoelzro: not that I want you to cry :o)
hoelzro oh, I realize how much a pain this is =) 21:03
FROGGS that's why I'd vote for a list (>10 items), and after all decisions you decide
pmurias just having simple easy to understand heuristics should be ok 21:04
if a piece of Perl6 code gets highlighted as Perl 5 nothing particularly bad happens
hoelzro exactly 21:05
I'd rather favor Perl 5
moritz aye; start simple, and once there are many reports about false classifications you can still tweak it
hoelzro (for now ;) )
21:08 ilbot6 joined
arnsholt comments out catch blocks for fun and profit 21:09
Where profit is defined as tracking down that pesky bug =)
FROGGS arnsholt: you are a ferengi?
FROGGS .oO( is Perl 6 like umox? ) 21:10
arnsholt Haven't seen enough ST:TNG to get that reference, I'm afraid =)
21:10 cognominal joined
diakopter ds9? 21:10
FROGGS ds9
quark asked a girl (his employee) if she would make umox (sex) for fun and profit 21:11
hoelzro woooo, 189/206 wrong 21:13
moritz hoelzro: maybe invert the condition?
hoelzro heh
arnsholt hoelzro: It's important to have a bad baseline. That way your system looks better in the paper ;)
moritz then you'll just have 17 wrong
hoelzro all I'm doing at the moment is looking for a shebang line and 'use v6' =)
so I have room for improvement! 21:14
FROGGS right, it would be worse if there was no room for it :P
dalek Iish: 7903d63 | moritz++ | lib/DBDish/mysql.pm6:
[mysql] do not screw up when there are question marks in the data
hoelzro not that I would mind with 0/206 ;)
moritz ok, ilbot6 logs into a database again
and doesn't die when question marks are used :-)
arnsholt hoelzro: What kind of split do you have between Perl 5 and Perl 6, OOC?
moritz r: just a test 21:15
p6eval rakudo b972ca: OUTPUT«===SORRY!===␤Undeclared routines:␤ a used at line 1␤ just used at line 1␤ test used at line 1␤␤»
hoelzro well, right now, I have 206 Perl 6 files and 99 Perl 5
moritz wow, even formatting codes and Unicode is correctly logged
I'll leave that running through the night 21:16
sleep&
masak 'night, moritz. dream of more than zero users in businesses and governments.
21:19 fglock left
FROGGS gnight moritz 21:20
21:23 kaare_ left 21:28 pmurias left
hoelzro the Perl 5 corpus could probably use some shoring up 21:28
I just included a smattering of Perl6:: modules
woot, 53/206! 21:29
dalek rlito: d57161d | (Flavio S. Glock)++ | / (2 files):
Perlito5::X64::Assembler - add cpuid, neg, leave, int3
 21:34
hoelzro I think checking for 'use Moose' and the like would be a good way of catching things 21:35
21:59 spider-mario left 22:03 dmol left 22:04 dmol joined
hoelzro ok, 8 failures 22:04
er, 7
all tolerable misses, imo =)
FROGGS cool!
hoelzro pull request created \o/ 22:09
boy, that feels good. 22:11
22:12 dmol left
FROGGS hoelzro++ 22:13
hoelzro += 10 # at least
22:13 dmol joined
hoelzro hooray, fake Internet points =) 22:13
FROGGS :o)
but hey, mind that there is a guy at the other end that is happy too :o) 22:14
hoelzro ;)
I hope that the author accepts the pull request
and doesn't just turn his nose and say "hmmph! Perl!"
FROGGS well, most do, right?
cognominal karma hoelzro 22:15
who handle the karma in #perl6 ? :)
diakopter ilbot
cognominal dalek, karma hoelzro
ilbo, karma hoelzro
tadzik naah, phenny did that
FROGGS ohh man 22:16
diakopter ilbot logs them!
so... moritz' clogs need the feature of displaying current karma :D
cognominal r: say " hoelzro++ " for 1..10
p6eval rakudo b972ca: OUTPUT« hoelzro++ ␤ hoelzro++ ␤ hoelzro++ ␤ hoelzro++ ␤ hoelzro++ ␤ hoelzro++ ␤ hoelzro++ ␤ hoelzro++ ␤ hoelzro++ ␤ hoelzro++ ␤»
cognominal I don't know if many increments are accepted in one line 22:17
is there an integer divide in Perl 6? 22:18
sorear rn: say 18 div 5
p6eval rakudo b972ca, niecza v24-35-g5c06e28: OUTPUT«3␤»
diakopter rn: say Inf div Inf 22:19
p6eval niecza v24-35-g5c06e28: OUTPUT«1␤»
..rakudo b972ca: OUTPUT«Cannot call 'infix:<div>'; none of these signatures match:␤:(Int:D \a, Int:D \b)␤:(int $a, int $b)␤ in sub infix:<div> at src/gen/CORE.setting:2886␤ in block at /tmp/LB3GSCRPoJ:1␤␤»
hoelzro sleep &
diakopter niecza++ I think
FROGGS gnight hoelzro 22:20
sorear karma hoelzro 22:23
22:29 denisboyun left 22:37 leprevost joined 22:49 dmol left 22:50 PacoAir left 23:10 pmurias joined 23:11 baest_ joined 23:12 bakedb_ joined 23:13 sftp left, hoelzro_ joined, `patch` joined 23:14 prammer_ joined, sftp joined, Targen left, thou_ joined, Targen_ joined, Maddingue left, moritz_ joined, thou left, fgomez left, broquaint joined, fgomez_ joined 23:15 Maddingue joined, cosimo_ joined, jtpalmer joined, BinGOs_ joined, ruoso joined 23:16 anocelot joined 23:17 Yappo____ joined, felher joined 23:18 pnu_ joined, snuffles joined, census left 23:19 nwc10_ joined, Some-body_ joined 23:20 autumn joined, Bucciarati joined, cosimo left, Some-body_ is now known as DarthGandalf, __sri joined, shachaf_ joined 23:21 mangala_ joined, xorp joined, shachaf_ left, shachaf_ joined, BinGOs_ is now known as BinGOs, BinGOs left, BinGOs joined, shachaf_ is now known as shachaf 23:23 bloonix left, _sri left, mangala left, pmurias left 23:24 yeltzooo joined, ingyfoo joined, pmurias_ joined, sitaktif1 joined, bloonix joined 23:25 rking joined 23:26 zamolxes joined 23:28 dpk joined 23:30 dpk left, dpk joined
dalek : 4064558 | (Tobias Leich)++ | / (3 files):
for loops with named iterator ok
 23:34
23:36 groky joined 23:37 dayangkun joined 23:39 groky left 23:40 groky joined 23:43 xalbo_ joined 23:44 yx joined, sunnavy_ joined 23:47 Yappo______ joined 23:51 mathw_ joined 23:52 zamolxes left, sitaktif1 left, pmurias_ left, ingyfoo left, Yappo____ left, eiro left, Rix left, awwaiid left, takesako__ joined, jtpalmer_ joined, eiro joined 23:53 Rix joined 23:55 raiph left, araujo joined, xorp left 23:56 zamolxes joined 23:57 xorp joined, jtpalmer_ left, jtpalmer left, prammer_ left, leprevost left, timotimo_ joined 23:58 clkao left 23:59 rom1504 joined, pmurias joined, Maddingue left, yoleaux 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!