»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! | Rakudo Star Released!
Set by diakopter on 6 September 2010.
00:00 devinus left 00:01 donaldh left, risou_ joined
ggoebel o/ 00:03
masak ggoebel: \o
00:04 ggoebel left, risou left
sjohnson oy 00:04
oops.. yo*
00:04 ggoebel joined
masak it was a dark and net-splitting night... 00:05
sorear masak: let me know if you run into anything excessively annoying. 00:06
masak sorear: oh, I will.
sorear masak: a little tip from my work on STD - edit-compile-run cycles go much faster if you add augment blocks to the main program instead of changing the modules
masak nod. 00:07
sorear (my periodic "Merge back" commits are de-monkeytyping of various things)
masak I've recently adopted that style of development when hacking on the Rakudo setting.
00:08 Ikarus left
sjohnson rakudo: say "piggy".reverse 00:08
p6eval rakudo : OUTPUT«piggy␤»
sjohnson rakudo: say scalar reverse "piggy"
p6eval rakudo : OUTPUT«Could not find sub &scalar␤ in main program body at line 22:/tmp/HFQdg3ovr7␤»
masak rakudo: say "piggy".flip 00:09
p6eval rakudo : OUTPUT«yggip␤»
sjohnson heh
thats cute
sweeeeeeeeet
masak sjohnson: I can read your mind, can't I?
sjohnson perl 6!!!
sjohnson yells loudly
you do indeed have a crystal ball programmed in p6
masak sjohnson: strangelyconsistent.org/blog/the-ta...edy-on-irc
sjohnson "Perl 6 -- making life worth living, again!" 00:11
00:12 betterworld left, betterworld joined 00:13 hanekomu left 00:27 Alias__ joined
Alias__ What's the syntax for the evalbot? 00:28
I need to test some corner cases
masak rakudo: say "this is my syntax"
p6eval rakudo : OUTPUT«this is my syntax␤»
sorear perl6: my $f = sub ($x) { say "A" }; $f($f = sub ($x) { say "B" });
p6eval rakudo : OUTPUT«B␤»
..pugs: OUTPUT«A␤»
masak sorear: I'm less surprised about Rakudo's behavior. 00:29
sorear also niecza:, alpha:, elf:, yapsi:, mildew:
masak to me, it makes sense that the (second) assignment is performed before the call binding. 00:30
sorear and nqp:, nqpnet:, perlesque:, std: for the deliberately partial ones
masak: TimToady has already ruled that both are legal; I was just using it as an example of perl6: where the results differed
masak right. 00:31
sorear pugs: say $*PERL
p6eval pugs: OUTPUT«␤»
sorear pugs: say $*VERSION
p6eval pugs: OUTPUT«␤»
sorear pugs: say $*OS
p6eval pugs: OUTPUT«␤»
diakopter pugs: say $*PUGS_VERSION 00:38
p6eval pugs: OUTPUT«Perl6 User's Golfing System, version 6.2.13.15, December 16, 2009␤»
sorear perl6: say $*PERL; say $*PUGS_VERSION
p6eval rakudo : OUTPUT«name rakudo␤version 2010.10-56-g227feae␤␤»
..pugs: OUTPUT«␤Perl6 User's Golfing System, version 6.2.13.15, December 16, 2009␤»
sorear ... we've probably scared away Alias__ . Oops. 00:39
diakopter heh
nothing scares him
masak Alias__ the unscarable.
sjohnson heh 00:42
00:49 sji left 00:53 jaldhar joined
sjohnson Golfing! 00:54
rjbs version FOOORRREEEEEE 01:00
masak rakudo: use v4; 01:01
p6eval rakudo : ( no output )
masak rakudo: use v4; say "fore!"
p6eval rakudo : OUTPUT«fore!␤»
sorear wanted: an IDL for foreign imports, to spec what classes to import and how, for a cross compiler 01:02
01:02 Teratogen joined
Tene IDL? 01:02
01:02 risou joined 01:03 risou_ left
sorear An interface definition language 01:03
masak sleep & 01:05
01:05 masak left
Alias__ rakudo: any(1) 01:06
p6eval rakudo : ( no output )
01:06 QinGW joined
Alias__ rakudo: if ( any(1) ) { say "true" } 01:07
p6eval rakudo : OUTPUT«true␤»
Alias__ rakudo: if ( any(0) ) { say "true" } else { say "false" }
p6eval rakudo : OUTPUT«false␤»
Alias__ rakudo: if ( any() ) { say "true" } else { say "false" }
p6eval rakudo : OUTPUT«false␤»
Alias__ pugs: if ( any() ) { say "true" } else { say "false" }
p6eval pugs: OUTPUT«false␤»
Alias__ Is there a dumper? Like the 'x' in the P5 debugger? 01:08
rakudo: say any(1) 01:09
p6eval rakudo : OUTPUT«any(1)␤»
Alias__ huh...
rakudo: say( any(1) )
p6eval rakudo : OUTPUT«any(1)␤»
sorear rakudo: say any(1).perl
p6eval rakudo : OUTPUT«any(1)␤»
Alias__ rakudo: say any().perl 01:10
p6eval rakudo : OUTPUT«any()␤»
Alias__ rakudo: say any().boolean
p6eval rakudo : OUTPUT«any()␤»
sorear rakudo: say any().Bool
p6eval rakudo : OUTPUT«Bool::False␤»
Alias__ ah, there we go
01:10 skangas joined
Alias__ odd that it's capitalised 01:10
sorear rakudo: say any().nonexistantmethod
p6eval rakudo : OUTPUT«any()␤»
Alias__ But whatever :)
sorear coercion methods are spelled exactly like the target type 01:11
Alias__ Junctions stringify to the code of themselves?
sorear rakudo: say True.WHAT.perl
p6eval rakudo : OUTPUT«Bool␤»
sorear I don't think the low-level stringification of junctions has been specced yet
Alias__ But the booleanification of any() is what I was looking for
List::MoreUtils::any { } () returns undef atm 01:12
I'm thinking that should really be false
(I've taken over List::MoreUtils to clean it up a bit) 01:13
sorear YES THANK YOU
Alias__ Tassilo has finally admitted maintainer defeat 01:14
sorear List::MoreUtils's handling of empty lists is incredibly stupid and has bitten me many times
Alias__ First I'm just cleaning up the test suite and expanding coverage, and fixing cases where the current implementation doesn't confirm to the current documentation
Then I'll move onto actual changes once that's done
But I see this project as a bit like the DBD::SQLite takeover 01:15
I just have to clean up the test suite, get the code packaging and releasing cleanly, and fix the obvious stuff
Then talented people will appear from the air and fix the XS stuff, and I just switch to release manager
01:22 justatheory joined
sjohnson hi 01:24
dalek ecza: 1b6ac44 | sorear++ | lib/SAFE.setting:
Remove several direct calls to context methods in setting
01:31
ecza: 7998ffb | sorear++ | src/ (2 files):
Improve code generation for <?foo> and <!foo>
01:36 Guest71264 left 01:38 kst left 01:39 kst joined 01:48 kanishka joined 02:03 risou_ joined 02:05 risou left 02:22 jferrero left 02:28 kst left 02:29 whiteknight left, kst joined 02:32 envi joined, awwaiid joined 02:39 gimix joined 02:56 kanishka left
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index...._io_tablet 03:01
03:08 plobsing joined
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....dex_tablet 03:18
03:21 plobsing left
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....dex_tablet 03:35
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....dex_tablet 03:40
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....dex_tablet 03:46
03:48 DarthGandalf joined 03:50 plobsing joined 04:06 risou_ left 04:07 risou joined 04:12 satyavvd joined 04:39 rhr joined 04:45 _twitch joined 05:09 orafu left 05:12 orafu joined, risou left 05:35 wamba joined 05:44 kjeldahl joined 05:53 wamba left 06:27 masak joined
masak good morning, #perl6 06:27
lately I've been having an increased awareness of when lichtkind++ is awake. :P 06:28
sorear good * #perl6 06:32
06:34 alc joined 06:37 ggoebel left 06:39 alc left, alc joined
lue hello sorear o/ 06:41
hello masak o/
hello world \o/
masak it's snowing beautifully outside. the sun's not up here yet. it's cold. 06:42
lue it's probably snowing here too. The sun already left. 06:43
sorear it's cold here too. no snow 06:44
lue oh, I forgot to say hello to arsenic-based life. Hello! o/ 06:46
masak that would be the people on reddit. :P 06:47
06:48 _kaare joined
lue Wait until it evolves into beings that can live alongside humans. "NEW Vita-mite™® vitamins contain the 100% of the recommended daily value of Arsenic!" 06:49
masak apparently, according to nihiliad++ on Twitter, the life-form isn't "arsenic-based" so much as having replaced phosphorous with arsenic. 07:00
dalek ecza: daad34d | sorear++ | / (4 files):
Generate better code for <?before \w>
07:02
lue [ Yeah, but it's catchier than "having-arsenic-in-its-DNA-instead-of-phosphorus life" :) ] 07:03
07:04 kst left 07:05 kst joined
masak "perl6 has really turned it up to 11." -- twitter.com/bda_ath_cx/status/10531783657394176 07:06
07:06 wtw joined
lue why does the method in-timezone use the * 60 * 60 bit at the end (in the S32::Temporal example)? Is it to make a complete timezone, or... 07:06
masak which example are you referring to? 07:09
07:09 kst left 07:10 kst joined
lue my $dt = DateTime.new('2005-02-01T15:00:00+0900'); say $dt.hour; $dt = $dt.in-timezone(6 * 60 * 60); say $dt.hour; 07:10
masak lue: "The :timezone signifies the necessary conversion by returning an integer giving the difference from UTC in seconds." 07:12
lue aaaah. 07:14
sorear odd, two successive runs of STD produced *user* times differing by 1.6%
07:15 uniejo joined
sorear I expect that much variability of the real time 07:15
07:21 Util joined, kjeldahl_ joined 07:22 xinming_ joined 07:25 xinming left
lue goodnight o/ 07:32
masak 'night, lue. 07:33
moritz_ good morning 07:36
masak :)
morning.
the Oceanians go to bed; the Europeans rise. 07:37
dalek ok: 872ff96 | hinrik++ | MANIFEST.SKIP:
Don't put README.pod in the release
ok: 0e492fd | hinrik++ | / (3 files):
Update hyperlinks to the Synopses
ok: d44bd65 | hinrik++ | dist.ini:
Update Prereq section in dist.ini
ok: 6628b38 | hinrik++ | Changes:
v0.23

   - Update hyperlinks to the Synopses
masak literal++
moritz_ uhm 07:38
did tadzik put a draft for today's advent post anywhere?
masak in mu, perhaps? 07:39
moritz_ pulls
yes
masak \o/
dalek ecza: 269d6a2 | sorear++ | lib/Kernel.cs:
Fix excessive slow-path use in get_first
07:40
07:40 nymacro joined
moritz_ perl6advent.wordpress.com/2010/12/0...perations/ 07:45
any idea how I can set the author to 'tadzik'? 07:46
masak the paragraph starting "Example usage?" looks mighty weird. 07:47
07:49 ggoebel joined 07:50 justatheory left
moritz_ better now. 07:51
07:51 Bzek left
masak yes. 07:52
"For more info please refer to the *documentation*."
is that meant to be emphasis like that, and not (say) a link somewhere?
which documentation are we talking about? the spec? 07:53
moritz_ makes it a link 07:55
masak moritz_++
sorear am I expected to write anything for this? 07:57
masak s/expected/encouraged/ 07:58
dalek ecza: eebc112 | sorear++ | / (2 files):
Run get_first off trampoline
08:05
08:09 felipe left 08:15 kensanata joined 08:17 alc left 08:21 renormalist joined 08:24 alc joined 09:02 hugme joined, ChanServ sets mode: +v hugme 09:05 hanekomu joined
moritz_ there was a comment on the advent calendar asking if multi MAIN(Int $x, Str $y) { ... } would work 09:12
currently it doesn't 09:13
if I remember the spec correctly
masak which part is the commenter asking about?
moritz_ I think we need some magic string parsing function
masak the 'multi' or the 'Int'?
sorear I think we need a change of perspective
moritz_ masak: multi dispatch on types
sorear I've been thinking a lot about defining subset Int of Cool where /^<[+-]>?\d+$/ 09:14
masak moritz_: as the spec currently stands (at least by implication), you can't pass in an Int.
moritz_ masak: correct
sorear this solves a great deal of issues
masak I mentioned this in my last spec-grumbling post.
moritz_ sorear: and creates a huge list of other problems
masak strangelyconsistent.org/blog/second...done-wrong 09:15
sorear moritz_: like what?
moritz_ sorear: subset types aren't as narrow as nominal types in multi dispatch, for one
sorear: and an Int isn't a string that looks like an Int. An Int is a number.
sorear How is that a problem?
moritz_ sorear: it will surprise lots of people who write signatures, and make type based multi dispatch much harder 09:16
and I guess it's slow.
sorear I'd like to see a specific example of a case that my way makes worse.
sjohnson masak: side question, are you a linux kind of guy?
moritz_ sorear: multi fac(0) { 1 }; multi fac(Int $x) { $x * fac($x-1) } # stops working 09:17
dalek ecza: 9d06eea | sorear++ | src/ (5 files):
Use thunks for parameter default values

This is preparation for using an out-of-line binder.
moritz_ if Int is a subset of cool, this desguars to 09:18
multi fac(Cool $x where 0) { 1 }
sorry
multi fac(Cool $x where /^\-?\d+$/ & 0) { 1 }
multi fac(Cool $x where /^\-?\d+$/ ) { $x * fac($x-1) } 09:19
which both have the same narrowness
=> you're screwed
masak sjohnson: I've used Linux at times. plan to do it again in the future. right now I'm using mostly Mac OS X. 09:20
sorear nitpick: when constraints are involved, MMD falls back to in-order testing, which makes this work
but I see the issue
moritz_ but that's basically Sheer Luck
sjohnson masak: cool. i was thinking of giving linux another try someday
sorear Int as a subset cannot work unless it's made a less constraining subset than 0 in some what
I need to plant this idea in TimToady, see if he can develop anything from it 09:21
masak sjohnson: I feel we're disturbing Real Perl 6 discussion. privmsg me if you want to keep talking Linux.
TimToady I plan to go a different direction with ath
*that
moritz_ anyway, I don't like it from a conceptual point either
sorear the extent to which Perl 6 forces you to care about the difference between strings and numbers bothers me a lot
moritz_ because forcing a Str representation on Int seems wrong
TimToady in fact, I plan to unify what happens with <1 2 3> with what happens to command line args
masak as mentioned before. 09:22
moritz_ TimToady: that was what I was going to propose before sorear++ came up with a different idea :-)
masak so both get the hybrid solution. :(
sorear In Perl 5, 1 and "1" are semantically almost identical
TimToady it's really the same fundamental underlying problem: you are forced to use text to represent a number
sorear the implementation uses a real number for one
moritz_ some kind of general mechanism that turns a string into whatever kind of literal it represents 09:23
sorear the case where Perl 5 breaks down is that 1 | 10 !=== "1" | "10"
moritz_ without going through eval() :-)
sorear Perl 6 fixes that, but breaks 1 === "1" in every other way
TimToady not every other way
mathw But they aren't the same thing
sorear a hypothetical Perl 5 with +| and ~| would allow you to not care 09:24
TimToady Cool covers most of the issues already
the remaining issue of one of MMD, mostly
09:25 tzhs joined
TimToady which Cool would usually solve if people don't specify Str $param, but just $param, or we make some way of Str $param to be looser and recognize ambiguous types from <1 2 3> and such 09:25
sorear right. I think Str, Int, etc being as tight as they are is the root of the problem 09:26
TimToady that is, the Str type could introduce just a bit of the "everything is a string" flavor back into Perl 6
sorear Str current has too much of a representationish flair
TimToady that's the right place to push it, I suspect, not fiddling with numeric subtypes 09:27
but I'm still thinking about it
which is hard at 1:30am :)
sorear I could stand being making Str alone a subtype.
3:30am is when I'm most creative, but I need daybreak to actually understand the ideas :)
Str alone being a subtype would break (1 ~ 1) ~~ Int. Which wouldn't be that bad. 09:28
hrm
moritz_ huh? that's False right now, and would be False then too
sorear in this model, how does Str differ from Cool? 09:29
TimToady the way the spec currently stands, <1 2 3> makes Ints that can("Str"), but perhaps we can strengthen that so that MMD can see it nominally
sorear moritz_: it's True in the p5think model, and I don't think it staying True is useful
moritz_ sorear: in the p5 think, AnyScalar ~~ Int (+maybe warning) 09:33
TimToady what we might want is some mild rewriting of multi (Str $s) -> multi (Cool $s as Str) 09:36
where an explicit sig with Cool would suppress that rewrite, presumably 09:37
we'd still presumably want to make sure the dwimmed number/strings do the round-trip without losing their original format, I suspect 09:39
or you can't have a filename of "1/3" :)
sorear -> sleep 09:42
TimToady EMETOO 09:45
sjohnson heh 09:46
sjohnson fluffs his pillow
09:49 tzhs left 09:50 dakkar joined 10:06 QinGW left 10:26 colomon_ joined, colomon left, colomon_ is now known as colomon 10:30 colomon_ joined, colomon left, colomon_ is now known as colomon 10:32 colomon_ joined, colomon left, colomon_ is now known as colomon
flussence regarding the type thing above, I'm wondering whether it would be better if a signature "(Y $x)" meant "$x.can(Y)" instead of "$x ~~ Y". 10:37
10:38 beerpages joined
beerpages Erstelle dir deine eigene Umfrage - Beispiel: www.beerpages.de/view/2H/Anti-%22Di...20Kampagne 10:38
10:38 beerpages left
flussence with that a strict type check (Foo $x) would become ($x where Foo), ($x where .can('Foo')) becomes (Foo $x), and (Foo $x where .can('Bar')) becomes (Bar $x where Foo). The DWIMmy second one becomes much shorter. 10:44
rakudo: sub t($u where Int) { WHAT($u).say }; t(1); t('1'); 10:46
p6eval rakudo : OUTPUT«Int()␤Constraint type check failed for parameter '$u'␤ in 't' at line 1:/tmp/LX8HqDPsU6␤ in main program body at line 22:/tmp/LX8HqDPsU6␤»
flussence (on an unrelated topic, my test runs are 100% again and they're a quarter hour faster than before! nice) 10:50
moritz_ flussence: \o/ 10:51
flussence there's still 2 hours of room for improvement though ;) 10:52
moritz_ is well aware of that 10:53
flussence I think the hours of typing saved will make up for that eventually. 10:55
.oO( hardware upgrades are cheap, when the hardware's not made of meat )
10:56
10:59 kst left, kst joined 11:01 MayDaniel joined 11:03 timbunce joined 11:05 satyavvd left 11:13 hanekomu left 11:37 alc left 11:51 kst left 11:52 kst joined 11:57 felipe joined, kfo joined 11:58 hanekomu joined 12:20 tzhs joined 12:22 uniejo left 12:36 zby left 12:40 edenc joined 12:44 timbunce left 12:50 ab5tract joined 12:56 daxim joined
takadonet morning all 12:58
colomon o/ 13:01
moritz_ \o 13:02
colomon: are you on schedule for your advent plog post for tomorrow?
colomon My post is about 66% complete. 13:03
Shouldn't be a big problem finding the time to finish it today.
moritz_ while we are at it, are there any more volunteers for advent blog posts?
14 out of 24 slots are still free 13:04
colomon And... hmm... I might save the "big finish" of it and make it a separate post. That would make finishing tomorrow's post a lot easier and provide another post for later. 13:07
moritz_ wfm
mathw hmm 13:09
mathw ponders a suitable subject for a post
Ideas please!
moritz_ mathw: have you seen the subject list? 13:10
mathw: in mu, misc/perl6advent-2010/topic-brainstorming 13:11
mathw: for example X and Z meta operators are a nice topic that hasn't been covered yet
mathw ooh 13:12
that is a nice topic
that I actually know some stuff about too
takadonet www.reddit.com/r/programming/commen...perations/ 13:14
colomon takadonet: are you putting those up on reddit, or is someone else?
takadonet colomon: i'm 13:15
need to get as many people exposed to Perl 6 as possible
mathw yes can I have X and Z metaoperators... let's say on the 6th? then I can write it over the weekend.
moritz_ mathw: sure, you're in 13:16
flussence er
in file-operations.pod: s/dirrectory/directory
mathw \o/ 13:17
my account should still all work from last year I assume
mathw adds thing to his calendar to remind him
dalek : 9c670f8 | moritz++ | misc/perl6advent-2010/schedule:
[advent] mathw++ volunteers for day 6
moritz_ mathw: should work, zes
s/z/y/
flussence: fixed, thanks 13:19
13:22 gimix left 13:32 xinming joined 13:34 MayDaniel left 13:35 xinming_ left 13:38 tzhs left 13:40 Patterner left 13:42 Psyche^ joined, Psyche^ is now known as Patterner 13:44 plainhao joined
mathw moritz_: thanks 13:47
bbkr how to call all matching multi subs in main scope? for multi methods the syntax is $object.*method_name, however for main scope *method_name() is syntax error. is it possible? 14:02
moritz_ doesn't think so
14:11 sji joined 14:19 jaldhar left 14:21 smash joined
smash hello everyone 14:21
phenny smash: 10 Nov 20:02Z <patrickas> tell smash colomon++ fixed the prime scripts, it would be great if you can re-run the benchmarks to generate new graphs
moritz_ hi smash 14:22
smash tell patrickas yes sir, running the scripts right now
phenny: tell patrickas yes sir, running the scripts right now 14:23
phenny smash: I'll pass that on when patrickas is around.
takadonet smash: tell me when you are done sir :) 14:24
smash: really want to see the results
smash takadonet: yes, but it generally takes like 10 hours to run everything 14:25
or something like that 14:26
but i'll let you know as soon as it finishes
i've been a little off perl6 (overwhelmed with work) what's new in town ? 14:30
smash git pull
takadonet been porting over p5 cpan modules over to p6 14:34
perl 6 advent calendar 14:35
perl6advent.wordpress.com/ 14:36
smash nice 14:38
14:58 icwiener joined 15:03 jferrero joined 15:04 cjk101010 left 15:07 cjk101010 joined 15:09 cjk101010 left 15:12 cjk101010 joined
moritz_ wow, the advent calendar had nearly 2.7k visitors yesterday 15:17
even more interesting is that there isn't a single huge source of visitors 15:19
reddit being the biggest with about 200 referrals
but the rest seems to be fairly spread out
takadonet !! 15:20
15:20 sftp joined
takadonet what were our levels last year? 15:21
moritz_ we had a day with ~7k hits due to slashdot last year 15:22
scratch that, 11k views 15:23
15:23 wtw left
moritz_ 45k visits in Dec 2009 15:23
takadonet I should start posting to slaskdot then...
slashdot*
moritz_ not much chance
takadonet why?
moritz_ there was already a /. story about the perl advent calendar
15:24 _twitch left
moritz_ developers.slashdot.org/story/10/12...-For-Geeks 15:24
15:27 sftp_ joined, sftp left, ab5tract left 15:31 kjeldahl_ left 15:34 _kaare left 15:35 am0c left
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index...._6_tablets 15:35
15:35 kst left, sftp joined 15:36 sftp_ left, kst joined 15:37 lichtkind joined 15:39 zby joined 15:40 REPLeffect left
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index...._6_tablets 15:40
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ble_tablet 15:46
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index...._io_tablet
15:51 wamba joined
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ics_tablet 15:51
lichtkind sorry another change in the geenral ordering
takadonet lichtkind: np 15:52
15:55 jfried left
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....l_6_tablet 15:59
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ics_tablet
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....lta_tablet
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ook_tablet
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....kup_tablet
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....dex_tablet
16:01 timbunce joined 16:02 dakkar left, cjk101010 left
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....tor_tablet 16:04
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....gex_tablet
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....nks_tablet
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ble_tablet
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index...._io_tablet 16:09
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....oop_tablet
masak ponders turning those updates into hourly digests 16:10
daxim karma (Herbert Breunung) 16:13
16:14 MayDaniel joined 16:16 alester joined
lichtkind masak: i said sorry in advance, i hope its the last time i have to change the index 16:17
daxim: what to delete the karma :)
want i mean
daxim what good is ++ if no one is there to track it?
masak sure; no problem. but even without flooding, more than one such message per hour doesn't really add much, since there are no summaries. 16:18
lichtkind masak: maybe your right
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ock_tablet 16:20
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ine_tablet
lichtkind sorear: ping? 16:23
daxim: yes but i think it works without braces
daxim karma daxim
nope.
lichtkind karma Herbert Breunung 16:25
karma: Herbert Breunung
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ics_tablet
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....tro_tablet
lichtkind daxim: strange 16:27
ok last one and then I keep silence for today :)
16:27 MayDaniel left
masak Yacc is dead: arxiv.org/abs/1010.5023 16:29
moritz_ already printed the paper
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ory_tablet 16:32
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ign_tablet
16:33 kst left
lichtkind oh i have an ideoa to change it all again, never mind not today :) 16:33
16:34 kst joined
PerlJam reporting changes to perlfoundataion.org seems a bit of a waste to me. 16:35
unless we're just trying to see what lichtkind is doing :)
lichtkind PerlJam: i found the idea of ourly reports fine 16:36
initially i wanted at least some connection to the wiki since there is so low traffic
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index...._6_tablets 16:38
16:43 thundergnat joined
thundergnat Hi perl6 people. 16:45
16:45 hanekomu left
takadonet thundergnat: hey 16:45
thundergnat What is the correct syntax to have a private has in a class? I'm either horribly confused or knoking up againsta bug. 16:46
*hash
takadonet thundergnat: thanks for the help on algorithm:diff! 16:47
thundergnat rakudo: class cat { has %!h is rw; method new { %!h{oh} = "hai"; return self; }; method oh ($a) { return $!h{oh}; } }; my $lol = cat.new; say $lol.oh;
PerlJam thundergnat: class C { has %!hash; }
p6eval rakudo : OUTPUT«Type objects are abstract and have no attributes, but you tried to access %!h␤ in 'cat::new' at line 22:/tmp/5pHyo0CB3U␤ in main program body at line 22:/tmp/5pHyo0CB3U␤»
thundergnat takadonet: no problem
PerlJam: that's what I thought.. see above
PerlJam yeah, I'm looking
16:50 lamstyle left
PerlJam oh, I'm blind. 16:51
you don't have an instance
thundergnat Ok, I'm horribly confused. 16:52
takadonet inside the new function, you do not have an instance yet. You need to call bless yourself
thundergnat oh... duh
PerlJam takadonet++
16:55 lamstyle joined
takadonet thundergnat: are you trying to set a default values for the hash? 16:55
thundergnat rakudo: class cat { has %!h is rw; method new { %!h{oh} = "hai"; return self.bless; }; method oh ($a) { return $!h{$a}; } }; my $lol = cat.new; say $lol.oh;
p6eval rakudo : OUTPUT«Type objects are abstract and have no attributes, but you tried to access %!h␤ in 'cat::new' at line 22:/tmp/zviERzQ3EV␤ in main program body at line 22:/tmp/zviERzQ3EV␤»
thundergnat takadonet: not initially 16:56
I'm actually trying to see if I can port the object interface to Algorithm::Diff and am looking to use a hash to hold the various bits of state. 16:57
Not having much success though. Mostly because of my poor understanding of perl6 objects I suspect. 16:59
I would think I want to use a private hash that is only accessed through internal methods. 17:01
takadonet thundergnat: my knowledge of perl6 object is limited as well
thundergnat My basic attempts at setting up a class with a private hash and using methods to access it have't come to much though. see above 17:02
wolverian I suspect you want BUILD, not new.
takadonet wolverian: one issue with BUILD right now is that default values do not get set 17:03
but i think you would have to use a BUILD to access a private attribute 17:04
thundergnat That sounds like what I might be after. I don't need default values when it is constructed. 17:06
PerlJam thundergnat: what sort of state will be stuck in the hash? Why use a hash rather than normal has-scoped variables 17:07
thundergnat PerlJam: umm, I'm being somewhat lazy and trying to port over a perl5 object without totally redesigning it. 17:08
takadonet anyone have a link to non paste url?
thundergnat It currently saves state in an array
17:09 nymacro left
thundergnat and hsa a host of subs that basically just return insex constants 17:09
errr.. inDex
PerlJam thundergnat: you're suffering false laziness IMHO :) 17:10
thundergnat PerlJam: could be :-/
takadonet rakudo: class cat { has %!hello is rw; submethod BUILD{ %!hello<oh> ='boo'}; method oh ($a) { return %!hello{$a}; } };my $lol = cat.new; $lol.oh('oh').say;
p6eval rakudo : OUTPUT«boo␤»
takadonet thundergnat: is that what you want? 17:11
thundergnat takadonet: possibly, looks promising initially.
17:13 plobsing left
thundergnat takadonet: actually, that looks pretty much exactly what I was trying to acheive. Thanks! 17:14
masak 7 days left to Big Announcement. 17:16
exciting!
takadonet thundergnat: sweet 17:17
While you hack on algorithm:diff, I will continue on Tree::Simple 17:18
17:19 justatheory joined
thundergnat takadonet: I got pretty much the entire procedural interface working with the exception of passing parameters to custom key generators. 17:19
Pushed the changes yesterday.
17:19 dual joined
takadonet thundergnat++ 17:20
thundergnat: i saw !
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ics_tablet 17:22
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index...._io_tablet
PerlJam looks at thundergnat's code
you weren't kidding ... it looks almost exactly like the Perl 5 version 17:23
17:24 _kaare joined
thundergnat PerlJam: tongue in cheek? The commented out code IS the perl 5 varsion. 17:24
takadonet PerlJam: I originally just did a straight up line for line so I can get it to work. 17:25
PerlJam thundergnat: I see that now that I've gotten down that far in my skimming.
well kudos to both of you!
takadonet++ thundergnat++
thundergnat Thanks, later
I got to get back to $work 17:26
17:26 thundergnat left 17:30 Axius joined
dalek ast: 9c5f904 | (Kodi Arfer)++ | S32-temporal/ (2 files):
Added tests for DateTime.perl.
17:42
kudo: 878659f | KodiB++ | src/core/Temporal.pm:
[Temporal] Ensured .perl works for DateTimes with saved offsets.
17:43
17:50 kjeldahl left, Axius left, kjeldahl joined
masak Kodi++ # seemingly Temprally indefatiguable 17:58
bbkr rakudo: "/etc/services".IO.WHAT.say # bug? 18:02
p6eval rakudo : ( no output )
jnthn o/ #perl6
masak bbkr: looks like the usual "error during &say call gets lost somewhere" bug. 18:03
jnthn! \o/
jnthn masak! \o/
flussence that code gives me a ===SORRY!=== locally 18:04
I've noticed the evalbot doesn't show those...
jnthn rakudo: say IO
p6eval rakudo : OUTPUT«IO()␤»
jnthn rakudo: say IO.WHAT
p6eval rakudo : OUTPUT«IO()␤»
jnthn rakudo: say IO.WHAT.say
p6eval rakudo : ( no output )
18:13 Axius joined 18:20 kensanata left 18:30 risou joined 18:31 karb joined 18:32 envi left 18:33 karb left, Axius left 18:51 smash left 18:53 zenog left 18:54 hanekomu joined
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ble_tablet 18:57
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....ics_tablet
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index...._io_tablet
odel: e405f5a | diakopter++ | / (2 files):
[dotnet] implement greedy quantifiers in regexes
19:07
lichtkind jnthn: zdravičko 19:12
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....dex_tablet 19:13
tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....dex_tablet 19:19
19:25 sky|plone joined, sky|plone left
sorear good * #perl6 19:25
masak \o 19:26
sorear lichtkind: pong 19:27
lichtkind ahh
great
sorear: could you dimm the amount of messages about tpf wiki down 19:28
due suckiness of some wikisyntac i have to sametimes change all pages
which results in spam
19:29 wallberg joined
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index....dex_tablet 19:30
masak nom & 19:33
19:33 masak left 20:09 wallberg left 20:14 y3llow_ joined, pothos_ joined 20:16 pothos left, y3llow left, y3llow_ is now known as y3llow, pothos_ is now known as pothos
jnthn Dobry vecer, lichtkind. :-) 20:21
lichtkind ahoj 20:22
jnthn Mas sneh? :)
daxim сне сне сне 20:24
20:24 daxim left
jnthn Was the сне getting heavy on his internet cables? :) 20:26
20:29 plainhao left 20:31 fhelmberger joined, fhelmberger left, fhelmberger joined 20:38 hercynium joined 20:39 MayDaniel joined 20:40 plobsing joined 20:49 MayDaniel left, kst left 20:50 kst joined 20:53 maluco left 21:13 am0c joined 21:14 icwiener_ joined 21:17 icwiener left 21:20 maluco_ joined 21:23 gabiruh joined
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index...._6_tablets 21:27
21:29 s0v joined 21:39 dual left 21:42 dual joined
dalek tpfwiki: (Herbert Breunung)++ | www.perlfoundation.org/perl6/index...._6_tablets 21:43
colomon rakudo: 0, 1, -> $a, $b { $a + $b } ... -> $a, $b { $a + $b > 10000 } 21:52
p6eval rakudo : ( no output )
colomon locally that's
===SORRY!===
Limit arity cannot be larger than 1
is that correct by spec, or just a Rakudo limitation? 21:53
22:01 alester left 22:03 _kaare left
dalek odel: d59b47f | jonathan++ | dotnet/compiler/PAST2DNSTCompiler.pm:
[dotnet] Generalize box_int/box_str to a general box sub that takes an argument. Will make some upcoming opset change bits easier. Also avoid calling dnst_for on something already in DNST.
22:22
22:26 ch3ck3r joined 22:30 ch3ck left
colomon Those of you with permission on the advent blog and time, please take a look at perl6advent.wordpress.com/?p=456&am...eview=true 22:33
I intend to do another round of editing if I have time, but suggestions for improvement would be very welcome.
22:34 timbunce left 22:42 timbunce joined
dalek odel: 4d392d5 | jonathan++ | dotnet/compiler/PAST2DNSTCompiler.pm:
[dotnet] Make unbox_str just unbox, and the place that needs a temp too now is explicit about that.
22:45
odel: ac7f473 | jonathan++ | dotnet/compiler/PAST2DNSTCompiler.pm:
[dotnet] Generalize unbox_str into a generic unbox taking an array.
jnthn wtf...taking a type. Not an array. 22:48
22:51 icwiener_ left
dalek odel: d14ea75 | jonathan++ | dotnet/compiler/PAST2DNSTCompiler.pm:
[dotnet] Kill unbox_int in favor of the generic unbox, and add temp_int in a couple of places where it was needed.
22:52
sorear good * #perl6 23:03
\o/ jnthn!
jnthn sorear: o/ :) 23:07
jnthn has time AND mental capacity to commit again :)
sorear jnthn: suppose you want to access System.Console in mscorlib from Perl6. What use statement will you write? 23:13
23:14 starcoder joined
sjohnson hi! 23:16
jnthn use System:from<dotnet>; I guess
Oh, wait...
use System::Console:from<dotnet>; # maybe... 23:17
23:17 sji left
jnthn Yeah, that's a class 23:17
So I guess the latter. 23:19
sorear Why does it being a class matter?
jnthn sorear: I dunno what I was thinking before... :)
System is just a namespace though so...guess it's more machinary to import all the stuff it wants to export. 23:20
But do-able too I guess.
diakopter I say :from<cts> or :from<clr>
jnthn I imagine most would work.
er
both
Just the first would get you a bunch more things imported.
not cts 23:21
diakopter well, namespaces are spread out across tons of assemblies
jnthn clr maybe though :)
diakopter: Yeah...true.
That makes things...tricky to know what file we want.
diakopter since those are compile-time resolved
jnthn *nod* 23:22
diakopter it really should be by assembly name, or assembly+namespace
or assembly+class
(all three options)
dalek odel: 14268c6 | jonathan++ | dotnet/compiler/PAST2DNSTCompiler.pm:
[dotnet] Eliminate a use of directly calling unbox_str in favor of the unbox sub.
odel: 06ce121 | diakopter++ | dotnet/compiler/PAST2DNSTCompiler.pm:
Merge branch 'master' of github.com:jnthn/6model
23:23
diakopter blast; it missed/masked its other commit 23:24
its other parent commit
jnthn wtf. :/
diakopter maybe that dalek plugin could be taught to recognize such diakopter-- screwups
jnthn diakopter: Oddly, it looks fine in the github commit log.
23:24 s0v left
diakopter oh, weird. 23:25
ok, that's it, time to upgrade TortoiseGit
jnthn diakopter: fwiw, a friend recently talked highly of code.google.com/p/gitextensions/ in comparrison to TortoiseGit. 23:26
Tene "Git Extensions is the only graphical user interface for Git that allows you control Git without using the commandline. It comes with a manual and video tutorials to get you started quickly." -- really? the only one? 23:33
jnthn OH RLY?
Tene: Yeah, I thought that was rather over-stepping the mark-etting. 23:34
23:34 timbunce left, y3llow is now known as y3llow_, y3llow_ is now known as y3llow
Tene Either they're ignorant of the alternatives, in which case I wouldn't trust them to be well-enough-informed to trust their software, or they're just... lying? I'm not sure why? 23:35
dalek odel: a5559d5 | jonathan++ | dotnet/compiler/PAST2DNSTCompiler.pm:
[dotnet] Eliminate some explcit unbox_int calls in favor of the unbox sub. Less code...yay.
Tene Or I guess they could have some much-more-specific meaning in mind, in which case it's just poor communication, i guess.
23:36 y3llow is now known as y3llow_ 23:37 y3llow_ is now known as y3llow 23:38 s0v joined
diakopter std: /(?:a)/ 23:41
p6eval std a194beb: OUTPUT«===SORRY!===␤Quantifier quantifies nothing at /tmp/USvWlnybA_ line 1:␤------> /(?:⏏a)/␤Check failed␤FAILED 00:01 119m␤»
diakopter that error message could be splendified
sorear What I don't want to do is to qx{ monop ... } in the compiler 23:43
at least, not in the short term
once the compiler itself is running on the CLR, I can reflect all I want
jnthn TimToady: I can't find <?[{]> style things documented in S05. 23:44
TimToady: Is it just falling out of <? being lookahead and [...] as a char class? 23:45
sorear Yes
jnthn OK, thanks.
23:46 y3llow left 23:47 y3llow joined 23:50 wamba left 23:52 bitter16 joined, wamba joined 23:58 mberends joined