»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.
tadzik what, TPF not accepted? 00:05
ridiculous
masak something like that. 00:07
'night, butterflies. 00:08
tadzik 'night
rafl pyrimidine: i did 01:56
tadzik: pitch your ideas anyway - we might be able to make it work still 01:57
moritz good *, #perl6 06:48
sorear good *, moritz 07:05
sorear o/ Teratogen 07:25
Teratogen greetings and felicitations 07:26
tadzik it'd seem that the python foundation did not make it into gsoc either 10:11
tadzik and neither did Battle for Wesnoth! /o\ 10:18
the world has come to an end indeed
moritz whiteknight mentioned on #parrot that we might do some rakudo-related projects through PaFo 10:22
felher BfW did not make it into gsoc? Thats really sad. A new Campaign again or some enhancements to the new planing mode would be really nice (and fixing a strange new multiplayer-bug) 10:25
tadzik fwiw, Python did not make it either 10:26
felher Is there a list of projects that did? I wonder if vim did make it :) 10:40
felher A found it. Sadly, vim isn't in their either. 10:49
s/their/there
fsergot Hi #perl6! o/ 11:56
masak good postnoon, #perl6 12:10
guess Google felt they got to big last year, and are slimming down. 12:11
moritz that's fine as long as they don't slim *us* down :-9
ah well, we shouldn't complain about the absence of benevolence, or however it's spelled 12:12
masak right. 12:17
reminds me of all the tweets around Christmas saying mean things about parents for not getting them that car or that iPad. 12:18
dalek Heuristic branch merge: pushed 202 commits to perl6.org/archive_dir by ranguard 12:52
ranguard moritz: ^^ that's what I'm going to go live with unless you (or anyone else) has objections 12:53
there's a script in archive (convert_pod2html.pl) that does what it's name implies as mowyw doesn't support .pod 12:54
mowyw has warnings (about utf8) but processes all files (except rfc/230.pod and exe/E07.pod, but I made the pod2html script skip them so they just won't exist) 12:55
moritz ok 13:01
ranguard moritz: ok to go live, or ok you'll review :) ? 13:05
moritz ranguard: ok go to live 13:08
ranguard: do I need to run convert_pod2html.pl on the server? or is the result from that already checked in? 13:09
ranguard results already checked in
moritz ok
dalek href="https://perl6.org:">perl6.org: 8ee7a06 | (Leo Lapworth)++ | / (1326 files):
Move into archive folder
 13:10
href="https://perl6.org:">perl6.org: c723573 | (Leo Lapworth)++ | archive/ (663 files):
Merge branch 'master' of ../pw into archive_dir
href="https://perl6.org:">perl6.org: 88b918f | (Leo Lapworth)++ | / (1326 files):
Move the archive into the source directory
href="https://perl6.org:">perl6.org: 0546cc3 | (Leo Lapworth)++ | source/archive/ (6 files):
Remove TT tags and processing
href="https://perl6.org:">perl6.org: 220974a | (Leo Lapworth)++ | source/archive/convert_pod2html.pl:
Add convert_pod2html.pl script, as mowyw does not handle .pod files
href="https://perl6.org:">perl6.org: 8f4a115 | (Leo Lapworth)++ | source/archive/rfc/meta/rfc- (2 files):
Add .html files (generated from their .pod files)
href="https://perl6.org:">perl6.org: ea91703 | (Leo Lapworth)++ | app.psgi:
Add an app.psgi file to make developing easier
href="https://perl6.org:">perl6.org: 5bc5815 | (Leo Lapworth)++ | source/archive/ (593 files):
Add a message at the top of most files that they are achived
masak ranguard++ 13:22
felher ranguard++ 13:25
masak finally gets around to reading the new HPMoR chapters 13:33
ooh, more battles! marvellous.
timotimo hpmor is everywhere o_O 13:34
masak finally, rationality is fashionable ;) 13:40
ranguard moritz: is the mowyw still running? - some pages are live www.perl6.org/archive/architecture.html, but others arn't www.perl6.org/archive/doc/ 13:46
moritz ranguard: it's running, yes. I'll investigate 14:14
ranguard thanks 14:15
moritz perl6.org/archive/architecture is now up 14:17
perl6.org/build.log is the full log
ranguard: is archive/ all latin-1?
ranguard shrugs - but probably 14:18
moritz++ # making it live
moritz ranguard++ # doing all the work 14:19
ranguard :)
I'll go add the redirects from dev.perl.org - this is all yours now - enjoy :) 14:20
moritz \o/ 14:21
I'll have to teach mowyw to apply character encodings on a per-directory base :/
ranguard or just not worry about it until someone complains - this is the archive after all :) 14:22
masak or iconv the directories that aren't utf8. 14:22
ranguard oh, guess the build log is abit messy 14:23
moritz masak: that sounds surprisingly simple and sane :-) 14:24
dalek href="https://perl6.org:">perl6.org: 618d40e | moritz++ | source/archive/list-summaries/200 (68 files):
[archive/list-summaries] convert from Latin-1 to UTF-8
 14:28
masak moritz: nowadays, I almost instinctively consider non-utf8-ness to be a deplorable oversight of some sort. 14:29
much thanks to the indoctrination of this fine channel.
moritz i had the "it's an archive, we shouldn't change more than necessary" reflex, but discarded it as WRONG :-) 14:30
masak the archive was unfortunately preserved for posterity in the wrong encoding when it first reached our caring hands. :) 14:31
see? it's simple. 14:32
dalek href="https://perl6.org:">perl6.org: a25cb6b | moritz++ | source/archive/ (10 files):
recode more stuff to UTF-8
 14:35
ranguard masak: gist.github.com/2060044 <- fles that weren't in the repo (just found them in the rewrite rules) - not sure what we should do with those? 14:36
moritz: ^^ even 14:37
moritz ranguard: would it be much inconvenience to just leave them where they are?
ranguard: if not, that might be the nicest solution... they don't have dev.perl.org/perl6/ URLs anyway 14:38
ranguard moritz: don't think so, I'll mention it to Ask, but there's now no links to them as they were accessed via reditects (github.com/perlorg/perlweb/blob/ma...redirects) 14:41
moritz ah well, then just delete them 14:43
I don't think preserving them is worth the effort 14:44
ranguard hmm, quick win, I could stick links to them on /archive/index.html some of the Presentations might be worth that? 14:49
moritz +1 14:51
dalek href="https://perl6.org:">perl6.org: 2a5ea1e | (Leo Lapworth)++ | source/archive/index.html:
Link to old PPT/PDF etc talks from archive page
 14:58
ranguard redirects on dev.perl.org updated in repo, I've emailed Ask and Robrt to put live 15:09
rindolf Hi all. 15:10
ranguard goes off to hack on puppet for metacpan - have fun with your new archive *wave*
rindolf So I tried niecza. It claims to have a focus on optimisation, but doing -e 'say 2+2;' took about two seconds.
masak rindolf: Niecza runs pretty fast for a Perl 6 impl, but the startup time isn't competitive. 15:12
rindolf: especially not since the bs branch landed in Rakudo. 15:13
rindolf masak: ah.
sorear good * #perl6 16:09
o/ rindolf
rindolf sorear: hi.
colomon o/ 16:10
masak sorear! \o/ 16:14
sorear masak! \o/ 16:16
masak the weekends are a blessing. I have such wonderful devious new uses that I want to put the new, ever-improving Niecza and Rakudo implementations to, but so little time. but weekends, at least some of them, are just brimming with time. whee! \o/ 16:37
fglock masak :) 16:40
colomon is trying to do advanced prep for Ambercon in between St. Patrick's Day gigs, and also has a ton of $work that needs doing. 16:53
flussence perl6: gist.github.com/2063802 18:22
p6eval rakudo 16bf0f: OUTPUT«exception: bar␤bar␤ in block <anon> at /tmp/AqKMXTXt0r:7␤ in block <anon> at /tmp/AqKMXTXt0r:4␤␤»
..niecza v15-4-g1f35f89: OUTPUT«exception: bar␤ at /home/p6eval/niecza/lib/CORE.setting line 1222 (warn @ 3) ␤ at /tmp/m8ACYslQz1 line 7 (ANON @ 5) ␤ at <unknown> line 0 (KERNEL run_CATCH @ 3) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1362…
..pugs b927740: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/p6eval/.cabal/share/Pugs-6.2.13.20111008/blib6/pugs/perl5/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime::Match::HsBridge'␤"␤*** '<HAND…
flussence that's from blogs.perl.org/users/philip_durbin/...nwhen.html - is that output a good or a bad thing?
geekosaur that seems correct for the code in the gist 18:23
geekosaur btw I'm not sure p6 gives that answer for the same reason p5 does; but it occurs to me that the topic in a CATCH block *should* probably be the exception that was caught, *and* at the same time that the global $_ should be available 18:37
so the explanation given may be wrong *and* the person complaining has failed to think things through properly
flussence perl6: my $x = 'foo'; given $x { when /f/ { try { die 'bar'; CATCH { warn $*OUTER::_ }; } } } 18:40
p6eval rakudo 16bf0f: OUTPUT«bar␤bar␤ in block <anon> at /tmp/hsL_Bb7KLq:1␤ in block <anon> at /tmp/hsL_Bb7KLq:1␤␤»
..niecza v15-4-g1f35f89: OUTPUT«Use of uninitialized value in string context␤ at /home/p6eval/niecza/lib/CORE.setting line 1222 (warn @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 230 (Mu.Str @ 10) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/lib/CORE.setting…
..pugs b927740: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/p6eval/.cabal/share/Pugs-6.2.13.20111008/blib6/pugs/perl5/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime::Match::HsBridge'␤"␤*** '<HAND…
au "exception: bar" is probably correct. rjbs's example gives "exception: foo" in Perl5 under "perl -MTry::Tiny -M5.10.0" which is surprising.
the <<This will warn "exception: bar" because...>> is probably just thinko for <<This will warn "exception: foo" because...>>. 18:41
flussence I can understand that, but now I'm confused why rakudo gives "bar" with $*OUTER::_ above...
au *nod* 18:43
geekosaur ah, ok, now that makes sense
fglock au: o/ 18:54
au \o 19:03
sorear o/ au 19:27
flussence: rjbs is warning to avoid given/when in Perl 5 because of a combination of design mistakes in given/when (that have since been fixed, but p5 can't take the fixes for compat reasons), a poor semantic match between p5 and given/when, and the $_ inconsistency 19:29
none of which apply to us
masak it's a pity that Perl 5 can't avoid all the issues with given/when, but maybe it's not so surprising when you're locked in by decades of design already in place. 19:54
meanwhile, Perl 6 refuses to "grow up" in that way before it absolutely has to :)
but in a way -- and I guess this was the point I wanted to get to -- the fact that Perl 5 tries and fails to get a problem-free given/when construct somehow vindicates the Perl 6 effort. 19:56
you can do a lot of cool stuff with Perl 5, but "you can't get there from here". 19:57
doy i'm not convinced that getting a useful given/when in perl 5 is/would have been impossible 19:58
but it was pretty badly mishandled back when it was implemented
masak at least it's fair to say, with hindsight, that a functioning one in Perl 5 won't look the same as the one we ended up with in Perl 6. 20:04
because of different environments.
doy sure 20:05
masak and I'm not referring to the fact that we stranded Perl 5.10 with our initial, broken conception of smartmatching :P
doy (: 20:06
colomon "we stranded"? 20:09
"we" stranded?
colomon knows it was in jest, but hates things that imply Perl 6 developers have all the power... 20:11
... in the Perl 5 / Perl 6 relationship.
flussence maybe the Perl 6 specs should have a whatwg-style warning at the beginning: "implementors of these features not taking part in spec development are likely to have them change out from under them in incompatible ways" 20:13
colomon Actually, being involved in spec development doesn't help much either. ;) 20:19
geekosaur being involved with spec development just makes it more likely you'll change the spec out from under yourself.. 20:20
japhb What facilities are currently available for profiling in Rakudo? To make this concrete, I just did 'time panda list' a couple times, and the shortest time was 23.1 seconds. How do I find the bottleneck(s), whether in panda, rakudo, or parrot? 20:33
flussence japhb: `perl6 --profile` outputs kcachegrind-compatible profiling data 20:40
moritz ... to stderr 20:49
flussence oh moritz, can I prod you to bump rakudo's nqp version again? There's a fix in it now for that compile failure I was getting 20:50
moritz flussence: sure, just a sec 20:51
dalek kudo/nom: b6ca7a6 | moritz++ | tools/build/NQP_REVISION:
bump NQP revision again, to get a build fix. flussence++ and not_gerd++
 20:52
flussence yay, it works! 20:58
moritz \o/
flussence (and I noticed I should've used --parrot-make-option instead of --make-option... that gets it done a lot quicker) 20:59
sdo bonjour 21:04
moritz bon soir 21:05
sdo j'ai besoin d'un commit
cognominal sdo, needs a commit bit to correct my spelling on github.com/perl6/course/blob/master/fr 21:06
sdo moritz j'ai besoin d'un commit bit s'il vous plait
cognominal sdo, this is an english channel and moritz can't guess what you need if you are not precise enough. 21:07
sdo ok man
moritz I need a commit bit please 21:08
I guess the link is github.com/perl6/course/blob/maste...README.pod 21:09
is it ok with you moritz?
cognominal sdo, with this commit bit you get access to all things Perl6 documentation, so you should get a clue :) 21:10
sorear welcome, sdo! 21:11
sdo hello sorear 21:12
moritz sdo: what's your github ID? 21:16
sdo I don't hove one yet 21:17
moritz sdo: you need one, otherwise I can't give you a commit bit 21:18
havenn githubs is wonderbar 21:18
fsergot o/
sdo don't worry I subscribe to github 21:20
sorear o/ fsergot
japhb flussence, moritz, thank you, I will give that a try. 21:21
sdo is it ok with the id doreys 21:22
moritz sdo: you now have commit access. Welcome and have fun! 21:23
sdo thanks 21:24
cognominal thx moritz
sdo moritz
moritz fwiw we now have 94 members in the 'perl6' team on github 21:27
which gives access to nearly all repos under the perl6 organization
moritz -> sleep 21:28
fsergot good night moritz :)
masak 'night, moritz 21:29
japhb So we've beaten 2**6, 3**4, 4**3, 5**2, 6**2, 7**2, 8**2, 9**2, I guess it's time for 10**2.
masak colomon: didn't mean to imply we have all the power. but I've heard accusations, in I-don't-know-how-much jest, that we neglected to tell the Perl 5 people when we updated our smartmatching semantics. 21:48
that update was itself a good example of "fine idea on paper, but didn't work -- here's one that does" 21:51
it looks fine on paper that smartmatching be symmetric.
turns out it's much more useful that it isn't.
dalek kudo/nom: eccc616 | jonathan++ | src/ (5 files):
First crack at PRE and POST phasers. Known NYI: POST handling doesn't get $_ set to the result yet, and one failing POST doesn't abort the rest. Even without that, can pass 17/25 tests in pre-post.t already.
 22:08
jnthn There's a little vacation patch. :)
masak PRE/POST \o/
jnthn++
jnthn Some enterprising soul can fudge pre-post.t if they wish
As for me, dinner :)
japhb jnthn++
masak at 11 PM!? oh wait. :P
jnthn Just gone 7pm here. 22:09
And the culture seems to be to eat later rather than earlier.
japhb jnthn, where is "here"?
masak yeah, you guys in the Americas. always a bit behind.
jnthn japhb: Buenos Aires
japhb Ah, very nice.
Travel Guy is travelling.
:-)
jnthn Yes, my first time here, and I'm liking it plenty.
Especially the steak!
japhb heh
jnthn Got no conference/workshop stuff to do here. Just relaxing. :) 22:10
japhb Oh my that sounds nice. 22:11
jnthn Yeah. Keeps the batteries charged :) 22:16
OK, gone for a bit...or the night :) 22:17
japhb is trying to make (more) sense of the kcachegrind profile output ... 22:21
What causes the same method to appear with multiple addresses, and with '1, '2, '3 at the end?
For example, 0xdb357f8:gimme, 0xea8d1f8:gimme'1
Hmmmm ... perhaps different caller paths? 22:23
Hrmmm, possibly not. 22:24
geekosaur my guess is that there are either multiple call centers in the function, or multiple entry points. the former seems more likely 22:25
japhb "multiple call centers"? 22:26
geekosaur it's usually possible to instrument a function with profile points in the middle of the function, to keep track of various code paths within it 22:27
hrm, I just realized another interpretation 22:28
multis
japhb Well, I'm seeing it light up the same line numbers in the same file.
And the profile appears to be a line-by-line profile. 22:29
japhb geekosaur, try: perl6 --profile `which panda` list 2>my-profile; kcachegrind my-profile 22:32
... assuming you have rakudo and panda installed locally, of course.
geekosaur rakudo yes, panda no, kcachegrind lolno (OS X) 22:33
japhb Hmmm, the 'n is in the original raw profile data, so it's not being created by some kcachegrind magic. 22:34
Does OS X have something that can read cachegrind format?
geekosaur cachegrind is part of the valgrind suite, no? last time I looked valgrind had lots of issues on OS X 22:35
so no valgrind and therefore nobody bothered with tools to read it 22:36
japhb That's unfortunate, because lot of tools use cachegrind-formatted profile output, because it's "what everybody uses". :-/ 22:37
japhb OK, looks like the 'n comes from recursion, as seen in parrot's src/runcore/subprof.c:699 and :812-813 22:55
flussence I think I remember seeing a gtk thing ages ago that could read cachegrind files, but no idea what it's called... 22:57
japhb flussence, alleyoop? 22:58
masak 'night, #perl6 22:59
japhb o/
flussence japhb: looks interesting, but from what I can see it doesn't do profiling stuff... 23:02
arnsholt flussence: Maybe you're thinking of kcachegrind? 23:04
It's KDE/Qt, but it's a very nice tool for looking at cachegrind files
flussence I'm thinking of *alternatives* to that :)
arnsholt Right. Never mind me then =) 23:05
japhb flussence, yeah, agreed. I just was looking at the description in apt. 23:06
arnsholt Incidentally, I saw dtrace.org/blogs/brendan/2012/03/17...me-graphs/ today 23:07
Might be an interesting alternative to the box-thing (the name escapes me at the moment) visualisation kcachegrind does
japhb Ooh yeah, that's much nicer. I always hated the box one, because it is really hard to see what's going on when children use like 95% of the parent's total time, all the way down the stack. You end up with a screen full of tightly nested boxes and no useful information. 23:11
arnsholt Yeah, I'd like to try it 23:13
I'm considering writing a converter for the Lisp implementation I use occasionally to get a proper visualisation 23:14
Text is nice for many things, but some things just are so much better to deal with graphically
japhb quite. 23:15
arnsholt Trying to convince my Lisper friends of this is met with benevolent incomprehension ^_^
japhb I bet! 23:17
sorear What's so unpalatable about Qt? 23:22
gfldex (void *) up and down casts? 23:27
sorear gfldex: people are refusing to use a program just because it uses qt
I've never heard of anyone saying "I ain't touching $thing because it's written in Python and I'm a Perl person" 23:28
gfldex i think that was more about kde what seams to mix badly with osx
shinobicl___ hi... is there a function like ^methods but that shows data-members? 23:42
nevermind,.... already found it... 23:45
^attributes

Powered by Cro and the Raku® Programming Language.

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

Thank you!