»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.
lue ohai o/ 00:00
masak lue: hi! I have good news.
00:01 jfried left
lue news of good? What be it!? 00:01
00:01 jhuni joined
masak Tardis is up and running. 00:02
lue :D
masak and I'm making a Yapsi release today where I mention as much. 00:03
provided github unbreaks, that is.
otherwise I'll probably just go to bed and do the release tomorrow.
lue offers git some unbreak medicine 00:04
s/git/github/
masak would that be medicine against unbreaking? because I don't need that. 00:06
lue the title for github's page on Tardis is amusing: "masak's tardis at master - GitHub" :) 00:09
masak yep. that's my tardis. if you clone it, you can have one, too.
...and we have unbreak! \o/ 00:10
masak proceeds with the release
diakopter wonders what was wrong with the Hugbit
lue
.oO(and github's Royal Fanfare™ erupts on Royal Unbroken Trumpets™.)
masak and it's uploaded: github.com/masak/yapsi/downloads 00:11
be the first in your household to get one!
lue What does tardis work on currently? 00:12
masak lue: everything that Yapsi works on.
(I'm assuming you mean "on what inputs")
masak composes announce mail to p6c
lue now's as good a time as any to get yapsi :) 00:14
00:15 stkowski left
masak p6c mail's away. 00:15
masak concocts blog post
00:15 stkowski joined 00:16 timbunce left
lue "the code has the same volume (in cubic centimeters) as its packaging." One question: what kind of volume? Physical, virtual, etc.? 00:17
masak physical.
(and etc.)
and virtual.
00:18 kanishka left
diakopter decommute& 00:20
00:22 greatflamingfoo left
lue how's the documentation grant going? 00:23
masak the Pod one? 00:24
00:24 simon_ joined
lue yes. 00:26
masak it's eagerly awaiting my finishing up the Web.pm grant, which should be any month now.
00:26 simon_ left
lue rakudo: say "ABC" ~~ /<alpha>+/ 00:26
Tene >.>
p6eval rakudo 5f5bae: OUTPUT«ABC␤»
lue rakudo: say "ABC" ~~ /<A::B>/
p6eval rakudo 5f5bae: OUTPUT«===SORRY!===␤regex assertion not terminated by angle bracket at line 22, near "::B>/"␤»
lue lets out big noooo... 00:27
masak lue: part of the problem is that the A::B syntax assumes the the rule is 'our'-scoped, something which feels odd nowadays. 00:28
lue The only other WTDI I can think of is Grammar.parse() 00:30
masak 'WTDI'? 00:31
colomon .parse is what you have to use if /<A::B>/ isn't available, right?
masak well, we definitely want to have a way to express the /<A::B>/ thing, but making rules 'our'-scoped feels a bit unjustified. 00:32
lue The tale end of TIMTOWTDI 00:33
masak blog post! strangelyconsistent.org/blog/yapsi-...1-released
lue s/tale/tail/ # homophone idiot
masak lue: ah, I see :)
00:34 stkowski left 00:35 kanishka joined
masak goes to bed 00:38
'night, #perl6! 00:39
colomon \o
00:39 masak left 00:51 rgrau_ left 00:55 ash_ joined 01:00 QinGW joined, ggoebel left 01:05 dnl left 01:07 greatflamingfoo joined 01:09 greatflamingfoo left 01:22 ggoebel joined 01:33 dju_ left 01:42 Chillance left 02:02 molaf_ joined 02:06 molaf left 02:09 whiteknight left 02:15 nymacro joined 02:17 lidden left 02:22 meppl left 02:42 xinming left 02:44 xinming joined 02:49 ggoebel left 02:52 nymacro left
diakopter 03:02
03:08 kavakava left 03:10 agentzh joined 03:14 eternaleye_ left, eternaleye_ joined, eternaleye left, eternaleye joined 03:21 eternaleye_ left 03:31 satyavvd joined 03:48 jaldhar joined 03:52 patspam left 03:55 nymacro joined 04:34 hudnix left 05:22 justatheory joined 05:26 pythonian4000 is now known as pythonian4000afk 05:40 jnthn left, jnthn joined 05:41 _ilbot2 left 05:43 _ilbot2 joined, ChanServ sets mode: +v _ilbot2 05:49 kaare joined, nymacro left, kaare is now known as Guest64750 06:21 colomon left 06:37 justatheory left 06:46 pythonian4000afk left, pythonian4000afk joined 06:47 wtw joined 07:01 uniejo joined 07:07 meppl joined 07:15 sftp left 07:19 kanishka left 07:21 colomon joined 07:34 kanishka joined 07:48 cjk101010 joined 07:49 broquaint left 07:51 nimiezko joined 08:00 xabbu42 left 08:05 Bzek_ left 08:07 daxim joined 08:09 Bzek joined 08:27 flatwhatson_ left 08:34 Guest64750 is now known as kaare_
jnthn monring, #perl6 08:46
08:46 pythonian4000afk is now known as pythonian4000, pythonian4000 left 08:50 mavrc joined
moritz_ morning, * 09:04
09:05 ggoebel joined
frettled \o/ 09:06
moritz_ reads the yapsi release announcement. masak++ 09:07
09:08 REPLeffect left, timbunce joined 09:14 hugme joined, ChanServ sets mode: +v hugme 09:21 Raynes left, starcoder left, exodist left 09:22 exodist joined 09:24 REPLeffect joined, starcoder joined 09:40 xabbu42 joined, xabbu42 left, xabbu42 joined 10:07 Axius joined, QinGW left 10:09 masonkramer left, masonkramer joined 10:12 broquaint joined 10:32 ggoebel left 10:34 frettled left, masak joined
masak oh hai, #perl6! 10:34
another glorious day in November!
moritz_ indeed 10:35
10:36 sjn left 10:38 frettled joined, sjn joined
jnthn masak: huh, it's foggy and rainy here :P 10:38
masak there's something of an inner light to it all, though :) 10:39
10:39 icwiener joined
masak actually, the weather up here is behaving. relatively speaking. 10:39
sky is gray, but not wet. temperature stays just above freezing. 10:40
jnthn I looked at the forecast for Malmo this week and every day said "crappy".
flussence :( horizontal rain here. 10:45
jnthn Wow, at least mine is verticalish. :-) 10:49
flussence I'm going to be travelling without net access some time this week, what'd be the best way to get an offline p6 reference? 10:50
moritz_ flussence: git clone the spec, the book and my perlgeek.de repo
10:53 nymacro joined
flussence thanks moritz_++ :) 10:54
10:56 agentzh left 10:57 Axius left 11:02 Axius joined 11:07 baux joined 11:09 xabbu42 left, xabbu42 joined 11:10 xabbu42 left
mathw Hello wonderful people! 11:25
11:28 Alias__ joined 11:30 Alias_ left 11:33 dnl joined 11:50 satyavvd left
takadonet morning all 11:57
12:03 nimiezko left
takadonet rakudo: multi fib($n where { $n < 3 }) { 1 } ;multi fib($n) { fib($n-1) + fib($n -2) }; say fib(2); 12:08
p6eval rakudo 5f5bae: OUTPUT«1␤»
takadonet rakudo: multi fib($n where { $n < 3 }) { 1 } ;multi fib($n) { fib($n-1) + fib($n -2) }; say fib(10);
p6eval rakudo 5f5bae: OUTPUT«55␤»
bbkr when I test MiniDBI I get error message that libpq / libmysqlclient libraries could not be located. do I need *-dev packages for this to work? 12:09
12:10 masonkramer left
arnsholt I don't think so. You on Linux or OS X? 12:10
12:11 Axius left
bbkr Debian 12:11
arnsholt Hmm. Should work as long as you have the mysql/postgres packages installed 12:12
Does locate libpq return anything
Or locate libmysqlclient
bobkare istr I needed libpq-dev 12:13
arnsholt I wouldn't know. Worth trying though 12:14
bbkr I have libpq and libmysqlclient installed. I'll try on my home machine, where I can experiment - to install dev packages.
I confirm that libpg-dev and libmysqlclient-dev packages are required for MiniDBI to work on Ubuntu/Debian 12:28
mberends bbkr: correct, the dev packages are necessary because of the way Zavolaj links to libraries. 12:29
bbkr thanks! time for testing :) 12:30
12:34 ggoebel joined 12:41 Bzek left 12:43 Bzek joined 12:45 bluescreen joined 12:46 bluescreen is now known as Guest86803 12:48 hudnix joined
masak I believe the upcoming change in Yapsi will require a bit of design. isBEKaml, tadzik, and patrickas might be interested in participating. I certainly would like to have someone to juggle ideas with. 12:48
it's about adding first-class closures, something that we need for a lot of other things. 12:49
12:50 timbunce left
colomon takadonet: that's a valid way of calculating fib, but it's also insanely slow. 12:51
masak let me guess, it's O(fib(N))? :)
colomon loliblogged: justrakudoit.wordpress.com/2010/11/...ght-along/ 12:52
masak \o/ colomon++
colomon masak: something like O(fib(N)), yeah. :) 12:53
masak it's easy to show: the algo bottoms out in a bunch of 1s.
(I don't know what you're talking about, by the way, I'm just guessing) 12:54
(but that's what usually happens)
12:54 hatseflats joined
takadonet colomon: I was having trouble with a bug with fib but it works now 12:56
12:57 timbunce joined
colomon takadonet: it's just that if you actually want to use it for anything, you want to do something like 1, 1, *+* ... * to calculate them, it's vastly more efficient. 12:59
13:01 plainhao joined
takadonet colomon: I just was showing a co-worker the power of the multi and condition statement in the header 13:01
my presentation last friday was really well received
masak \o/ 13:02
takadonet Our lab policy is that you can use any language you want to do your work as long as you can justify it 13:03
so small script I can use perl6 but .... any really data parsing, not a chance 13:05
masak makes sense.
13:05 xabbu42 joined
takadonet hopefully in time I can use perl6 on a daily basis 13:08
13:16 hanekomu joined
masak rakudo: use MONKEY_TYPING; augment class IO { method readline { $!PIO.readline_interactive("> ") } }; say $*IN.readline 13:18
p6eval rakudo 5f5bae: OUTPUT«> Land der Berge, Land am Strome,␤»
masak oh right. I forgot. Austria :)
13:18 tzhs joined
mathw hah 13:19
masak rakudo: use MONKEY_TYPING; augment class IO { method readline { $!PIO.readline_interactive("> ") } }; my $f = $*IN.slurp; say $*IN.readline
p6eval rakudo 5f5bae: OUTPUT«> Null PMC access in type()␤ in main program body at line 22:/tmp/dU72vefrsV␤»
mathw aaaw
13:19 envi left
masak I get a Null PMC access when using $!PIO.readline_interactive and someone does ^D 13:19
I'm hesitant to report it as a bug, because it's not very exposed and I need to use MONKEY_TYPING to expose it. 13:20
so I'll frame it like this: is there something I can do to not get a Null PMC access here when someone does ^D?
13:21 Patterner left
masak rakudo: use MONKEY_TYPING; augment class IO { method readline { $!PIO.readline_interactive("> ") // Nil } }; my $f = $*IN.slurp; say $*IN.readline 13:21
p6eval rakudo 5f5bae: OUTPUT«> ␤»
masak that seems to do the trick. huh.
I suppose it's not a reasonable thing to add this automatically on "foreign" method calls. 13:22
13:22 MayDaniel joined
masak largely due to there not being a clear delineation of what's "foreign" here. 13:22
13:23 envi joined
masak also, if someone were to package the above functionality in a module, I'd gladly use that module. 13:23
jnthn masak: (closures) You may want to read the one or two patches I wrote that added those to 6model 13:25
masak url?
13:27 sftp joined
jnthn github.com/jnthn/6model/commit/1bb0...c7808493d8 and github.com/jnthn/6model/commit/13a7...2c339d0b3f and github.com/jnthn/6model/commit/c04f...b20349633d 13:27
colomon takadonet: understood. certainly is a fine example of "where". :) 13:29
masak jnthn: thank ye. will peruse. 13:30
s/ye/thee/
jnthn Puzhalusta.
masak :)
takadonet colomon: Really nice since you get both end cases fib(0) and fib(1) in one shot
jnthn ...that looks so wrong not in cyrillic... :-)
takadonet plus the first 'iteration'
masak jnthn: aye. I don't see auto-palatalization in non-cyrillic. 13:32
my brain automatically adds it when I read cyrillic.
13:33 Psyche^ joined, Psyche^ is now known as Patterner
jnthn Same :-) 13:33
masak jnthn: before the refactor in August, Yapsi only had outer context. right now it only has caller context... :) as soon as we introduce 1st-class closures, we'll need both. 13:34
(because it's when you start moving closures around, and start to refer to them from faw away, that OUTER:: and CALLER:: start to differ.)
jnthn I had outer and caller from day 1 in 6model otherwise I couldn't get lexicals to work. :-) 13:36
Adding closures is mostly about cloning.
Happily, I don't have to worry about when to clone in 6model 'cus the PAST tree that I have already contains that information though :-) 13:37
masak is "DNST" your version of "PAST"? :)
jnthn No
I use PAST
DNST is lower level.
It's more like POST
PerlJam DNST == Duke Nukem Syntax Tree? ;)
jnthn PerlJam: oh no...that connection never occurred to me! :P
masak do you think of Kirsten Dunst ever time you write DNST? will you from now on? :P
jnthn who...
masak Mary Jane in Spider Man. 13:38
jnthn masak: oh wow.
masak en.wikipedia.org/wiki/Kirsten_Dunst
jnthn ...thanks. :P
yes yes, I just google image searched.
masak Puzhalusta. :)
jnthn: why couldn't you get lexicals to work without both outer and caller? they work fine in Yapsi... 13:39
jnthn masak: In Yapsi do you ahve so far an opportunity for the caller *not* to be the outer? 13:40
masak no.
jnthn OK, that's why you've not run into the issue. :-)
13:41 nymacro left
jnthn I did have that possibility before I made closures work properly. 13:41
masak where? how?
13:41 kaare_ left
jnthn When I compiled loadinit methods, or signature construction, for example. 13:42
s/methods/blocks/
13:42 MayDaniel left
masak ok; we don't have either in Yapsi yet. 13:42
jnthn OK. I can see how you got away with it.
I had to support loadinit from very early on.
You can probaly do both of those in Yapsi without needing loadinit, perhaps. 13:45
(That's the goal for nqp/6model/rakudo too...)
(But iiuc you're interested in exploring the compile time / runtime relationship in Yapsi.) 13:47
masak TimToady: based on what you said yesterday, should the following only say "B"?
rakudo: class A { method foo { say "A" } }; class B is A { method foo { say "B"; lastcall } }; B.new.*foo
p6eval rakudo 5f5bae: OUTPUT«B␤A␤»
masak jnthn: yes, though that branch of experimentation has for the moment faded into the background. 13:48
jnthn would be curious to know that too.
masak I spent some time thinking about it yesterday. 13:49
jnthn masak: My guess is perhaps no because in that case you're using a dispatcher that tries to invoke everything, rather than playing the deferal game.
masak: I'm curious how nextsame and .* interact.
masak: I guess I'd rather they didn't. :-)
WEll
Maybe :-)
Or maybe they should. Hm. Can argue that both ways. :-) 13:50
masak it seems to me that sometimes a 'nextsame' or similar in a method doesn't actually *know* whether it's talking to an MRO dispatcher or a dot-star dispatcher.
and I think that's a problem in some ways.
jnthn Or a multi-dispatcher. :-)
masak right.
jnthn Though you probably know 'cus it's marked "multi"
masak aye. 13:51
in the case of wrappers, it's also clear-cut, since we know we're writing a wrapper.
but it seems to me there's some possibility of confusion in the case of dot-star.
and dot-plus.
and confusion isn't really what we need in that area :) 13:52
jnthn That's why I ponder that .* should cause an invocation with its own candidate list from the POV of if we'd actually called that candidate.
For all of the candidates. 13:53
masak whatever model we land on, I'm happy if that model is consistent and easy to cram into a normal-sized head.
jnthn So it's like you'd really done B.new.B::foo; B.new.A::foo
masak in which case B\nA\n above is correct, right? 13:54
jnthn If we go that way, yes.
masak and 'lastcall' is ineffecutal.
jnthn Right.
masak and TimToady was wrong yesterday.
jnthn lastcall; lastcall; otoh... :-)
masak ewww!
jnthn Well, I'm positing more than saying anyone is right/wrong. :-)
masak I hope that doesn't become a recommended pattern. 13:55
it's enough to talk to one dispatcher in a method. :(
13:55 amkrankruleuen joined
amkrankruleuen Hi 13:55
jnthn it's been suggested in spec that lastcall takes an argumennt saying what kinda dispatcher you're interesting ihn terminating.
masak amkrankruleuen! \o/
13:56 uniejo left
masak jnthn: maybe lastcall is sugar for $*MULTI_DISPATCHER.lastcall or sump'n. 13:56
or rather $*DISPATCHER.lastcall, and $*DISPATCHER is bound to the inntermost dispatcher. 13:57
jnthn It'd maybe be semantically like a contextual, yes
But with a way to say "I want to look for the nearest dispatcher that is a multi dispatcher" or "I want to look for the nearest one that is a method dispatcher"
masak right. so you'd need both $*DISPATCHER, and individual contextuals for the specific dispatchers. 13:58
jnthn That could work.
masak haz a happy
something I said wasn't completely insane! \o/
jnthn I maybe don't object to really having those as contextuals either...
13:59 drbean left, drbean joined
jnthn well 13:59
a $*DISPATCHER anyway.
a bunch of different ones could get icky
masak it would provide a way to talk to individual dispatchers. 14:00
a bit like labels provide a way to leave individual blocks.
jnthn It means having lots of slots in every block though.
And isn't very extensible for when $user decides to go dispatcher writing. 14:01
14:01 Plastic_Chicken joined
masak troo. 14:02
14:04 clkao left
masak rakudo: class A { method foo { say "A" } }; class B is A { method foo { say "B"; lastcall } }; my $b = B.new; my @candidates := .WALK(:name<foo>); $b.*@candidates 14:07
p6eval rakudo 5f5bae: OUTPUT«===SORRY!===␤Can not use .* on a non-identifier method call at line 22, near ""␤»
masak if this worked would 'lastcall' have any effect? would 'nextsame' in B.foo produce the output B\nA\nA\n or just B\nA\n? 14:08
14:09 clkao joined
masak clkao! \o/ 14:10
14:13 ash_ left 14:17 tzhs left
jnthn masak: Yeah, that's a problematic case because we have a list of methods rather than being able to produce a list of dispatchers, I guess. 14:17
masak exactly.
so there's a line to be draws between the two. 14:18
jnthn While the recent spec changes massively cleared up multi-dispatch, I think we may have a little distance left to run on dispatchers generally.
masak and some of what you said above about .*foo sounded like drawing the line so that .*foo ends up on the same side as .*@candidates
14:18 Plastic_Chicken left
jnthn Particularly this nested stuff. 14:19
masak aye.
jnthn And what .* and friends really mean.
masak sludgy, to be sure.
er. slushy.
jnthn Same stuff. :-) 14:22
14:24 amkrankruleuen left
masak sometimes I feel, even apart from this whole bug-submitting gig, that my job in the Perl 6 world is going over and standing next to some feature and saying "guys, this isn't very good yet". 14:25
14:28 amkrankruleuen joined
colomon I'd say +1 except I totally don't want you to feel that you are limited to that. 14:28
but it's a hugely important job, and you've done well at it.
14:30 xinming left
masak thanks. :) 14:32
I didn't expect to end up in that role, since I'm not a very plaintive person to begin with. 14:33
14:33 plobsing left
colomon In fact, you've done well enough at it I probably should work hard at convincing you of my own complaints -- I'm sure you'd do a better job airing them than I have. 14:33
masak but I guess I could frame myself as the grain of sand that polishes the rock. 14:34
colomon grain of diamond, dude. ;)
masak hardly :P
14:35 xinming joined
mkramer I've also learned a lot of perl 6 from following your postings masak 14:35
masak the blog postings? cool. 14:36
I've learned a lot of Perl 6 from submitting bug reports.
mkramer by the way, "a lot" is relative here 14:37
"a lot of what I know about perl 6" would have been more accurate
masak that's still more than most Perl programmers, relatively speaking. :) 14:38
mkramer tragically, yes 14:39
masak don't view it as tragic; view it as an opportunity for you -- to capitalize (monetarily or otherwise) on your relatively vaster knowledge of something useful. 14:40
14:41 justatheory joined
masak rakudo: say "P earwle s6o mies".comb[(0, 2 ... 16), (1, 3 ... 15)].join 14:51
p6eval rakudo 5f5bae: OUTPUT«Perl 6 is awesome␤»
14:51 risou joined
mkramer cute 14:52
I'm personally very excited to use perl 6 as a production language. In particular if we can get async right 14:56
masak rakudo: sub indices(@list) { my @sorted = @list.sort; my %s; gather for @list -> $e { my $i = %s{$e} // 0; ++$i while @sorted[$i] ne $e; %s{$e} = $i + 1; take $i } }; say indices("Perl 6 is awesome".comb).fmt 14:59
p6eval rakudo 5f5bae: OUTPUT«4 6 13 10 0 3 1 9 14 2 5 16 7 15 12 11 8␤»
masak rakudo: say " 6Paeeeilmorssw".comb[<4 6 13 10 0 3 1 9 14 2 5 16 7 15 12 11 8>].join
p6eval rakudo 5f5bae: OUTPUT«Perl 6 is awesome␤»
15:00 y3llow left, y3llow joined 15:12 nymacro joined 15:15 justatheory left 15:19 MayDaniel joined 15:20 wtw left 15:36 MayDaniel left 15:42 gbacon joined 15:43 nymacro left, timbunce left 15:48 _kaare joined 15:50 timbunce joined, masak left 15:57 cjk101010 left 16:09 iWork joined, iWork left, cdarroch joined, cdarroch left, cdarroch joined 16:10 stkowski joined, risou left 16:11 MayDaniel joined, risou joined 16:15 \xF0 left 16:16 \xF0 joined 16:17 wooden joined, wooden left, wooden joined 16:20 starcoder left, starcoder joined, patspam joined 16:23 Alias joined 16:26 Alias__ left 16:28 Plastic_Chicken joined 16:36 lamstyle left, florz left, [Sec] left, simcop2387 left, Grimnir_ left, wolverian left, c1sung left, sorear left, wolverian joined, sorear joined, [Sec] joined, simcop2387 joined, c1sung joined 16:39 lamstyle joined 16:41 tadzik joined
tadzik ~~ 16:41
16:42 florz joined
takadonet tadzik: yo 16:43
TimToady takadonet: also, you're assuming you want your fib sequence to start 1,1 rather than 0,1, which some folks like 16:48
moritz_ colomon++ # impressive progress with ABC 16:49
slavik1 1 1 2 3 5 is a zip code in brooklyn :)
colomon moritz_: you're too kind.
but I am having a lot of fun with it.
moritz_ colomon: no. I looked at the PDF, and was impressed.
colomon moritz_: I've been dreaming about having Perl ABC tools for at least five years now. But it wasn't easily practical without grammars. 16:51
moritz_ colomon: reminds me of Math::Model, which I've also envisioned long ago 16:52
but which needed signature introspection (or heavy workarounds) to work properly
16:53 mavrc left
tadzik feels like playing "The Star of Rakudo" 16:54
colomon I still need to get around to recording it. :)
tadzik I don't recognize the ~ symbol over some notes, what is it? 16:55
colomon It's a turn symbol.
Though in this context, it's supposed to indicate an Irish-style roll.
tadzik I've never seen it
interesting 16:56
colomon It's basically a way of breaking a note into three distinct notes, but by using your fingers rather than your breathing / tongue.
(On a fiddle, it may be "using one bow-stroke"? strings are weird.)
16:57 baux left
colomon it's a very standard technique in Irish traditional music 16:57
tadzik I have to say that this PDF really impressed with Lilypond’s output.
Have you accidentally a word?
colomon doh!
tadzik++ # fixed now, I hope 16:58
arnsholt tadzik: Yeah, Lilypond is very pretty
tadzik colomon++ # every time I stumble upon a sentence like this, the first thought is "damn, my english isn't that good" 16:59
16:59 jql left
colomon arnsholt: he was just quoting me. but yes, after initial doubts, I'm now thinking Lilypond generates good looking sheet music. 17:00
17:01 Grimnir_ joined 17:04 mkramer left 17:05 amkrankruleuen left 17:06 mkramer joined, Plastic_Chicken left, timbunce left 17:07 amkrankruleuen joined 17:12 timbunce joined 17:16 mkramer left 17:19 timbunce left, guidj0s joined 17:25 kanishka left, timbunce joined 17:35 kanishka joined 17:38 dnl left 17:43 mberends left 17:44 Raynes joined 17:48 frettled left, frettled joined 17:50 daxim left 17:51 envi left 17:52 kcwu left 17:53 kcwu joined 17:54 mkramer joined 17:55 jaldhar left 17:57 jaldhar joined 17:58 Guest86803 left 18:03 donaldh joined 18:11 molaf_ left 18:14 Guest86803 joined 18:18 donaldh left 18:27 tadzik left 18:28 tadzik joined 18:32 kanishka left 18:34 jaldhar left 18:39 araujo left
colomon #phasers in 20? 18:40
jnthn erm 18:41
oh...dammit
yes
Timezone change.
18:45 araujo joined 18:48 wamba joined, Helios left, wamba left 18:49 wamba joined 19:01 araujo left 19:04 araujo joined 19:07 timbunce left, jql joined 19:08 xabbu42 left 19:09 xabbu42 joined 19:11 mberends joined, Ross joined 19:13 xabbu42 left 19:14 timbunce joined, molaf joined 19:15 Ross^ joined, Ross left 19:19 tzhs joined, M_o_C joined 19:20 M_o_C left 19:21 masak joined
masak ahoy! 19:21
takadonet masak: hey
mberends ahoy masak, from Amsterdam.pm! 19:22
masak \o/
mberends: tell Amsterdam.pm I said hi :)
tadzik hi masak! 19:23
19:26 meppl left
pmichaud did Amsterdam.pm drink the yapc::eu auction beer yet? 19:28
szabgab hi pmichaud 19:29
pmichaud szabgab: hello -- got your messages but still trying to find good time to respond 19:30
(and I have to leave in 4 mins but will be back in 60)
szabgab perl6-ers, has anyone submitted a talk proposal to the Main tracks of FOSDEM?
I was told that there might still be a chance to submit a talk quickly
mberends pmichaud: yes, last month, where masak++ also gave us a little talk :) 19:31
szabgab so, pmichaud if there is no talk submitted yet and you submit one and get accepted
pmichaud I've thought about submitting a talk, and would love to go, but I'm concerned that we might have a repeat of last year where I have to cancel at the last moment
szabgab I'll make sure PEG will finance it - we already have the money 19:32
pmichaud if that's the case, I'll definitely submit a talk
and I'll insure the travel costs again
szabgab please do it ASAP as the dead-line is oevr 2 weeks ago
pmichaud I really enjoyed my last trip to fosdem/brussels
I'll do it this evening
any particular topic you'd like to see emphasized/covered? 19:33
szabgab and pls also send it to me by mail so I can forward them as well
pmichaud: it is up to you
you are the master
you know way better how to get people enthusiastic
pmichaud okay.
jnthn szabgab: I may be able to attend to.
szabgab jnthn: that would be awesome 19:34
pmichaud jnthn: I recommend submitting a talk also, if you can
jnthn If pmichaud++ is going, may do it for meeting up.
pmichaud if something comes up with me and travel, you could then cover my talk
jnthn pmichaud: OK. Let's try and not overlap too much then. :-)
szabgab actually I already started to ask companies to sponsor a Perl 6 hackathon at FOSDEM
pmichaud p6 hackathon at fosdem would be awesome
I need to check with my wife and make sure that's a good time for travel (should be)
have to run pick up kids from school.... bbiaw 19:35
jnthn k
szabgab k
19:36 envi joined 19:39 plainhao left 19:45 dnl joined 19:51 tzhs left
tadzik szabgab: are you planning more Perl 6 screencasts? 19:52
masak is the default .new method a multi? 19:58
jnthn Not today. 19:59
Not sure if we want it to be?
20:00 hanekomu left
masak well, sometimes I have the urge to declare a 'multi method new' in my own classes -- getting both the regular one and my variant. 20:00
I was hoping the new multi-dispatch changes might allow for that. 20:01
szabgab tadzik: planning yes, time - I don't have 20:03
for some reason lots of paid work started to flow in my direction which is of course good, but does not leave a lot of time for other things 20:04
and I want to focus on getting PEG up and running 20:05
sorear yay! I'm backlogged! good * #perl6
20:05 guidj0s left
masak szabgab: that's good news. sounds like you're being busy in a good way. Perl 6 will still be here when things settle down :) 20:05
szabgab yeah, it is not running away 20:06
so to speak :)
masak ouch :)
20:07 wallberg joined
tadzik sorear: o/ 20:07
szabgab which reminds me I should update rakudo on my netbook before I leave to Amsterdam in 1.5 days
masak wallberg! \o/
20:17 timbunce left
wallberg masak: hi! 20:18
masak wallberg: how's $work? and $life in general? 20:26
sorear Who'se wallberg?
masak he's a Perl person here in Uppsala, where I live. we talk about booting up an Uppsala.pm mongers group, but we never seem to get around to it. 20:27
wallberg haha, yes exactly.
jnthn
.oO( Uppsala means treehouse, the upp meaning "up" and sal(a) meaning "room" )
20:28
masak jnthn: now you're just making stuff up. :) 20:30
tadzik and Warsaw.pm seems to be dying a bit :/
jnthn Dang, caught.
wallberg well, life and work keep me busy atm. I commute to Stockholm on a daily basis, have two crazy kids at home and a spouse who is on her final year of graduate studies. I struggle to find time for the monthly or so LUG as well :-)
jnthn tadzik: Oh no! It can't die until I first visit to drink wodka with you all! :/
...though preferably it'll keep going after that too... :-) 20:31
tadzik jnthn: we usually go for beer. And the problem is, when the main Monger moved to London, there's no one to organize meetings
wallberg masak: are you in Uppsala on thursday?
masak wallberg: no, I have an appointment with Austria.
jnthn tadzik: Beer is good too :-)
tadzik jnthn: we could make one new workshop if you come :) The last one was nice 20:32
wallberg masak:?
masak wallberg: apparently, they need me again.
masak rolls eyes
wallberg masak: haha
masak rakudo: .say for $*IN.lines
p6eval rakudo 5f5bae: OUTPUT«Land der Berge, Land am Strome,␤Land der Äcker, Land der Dome,␤Land der Hämmer, zukunftsreich!␤Heimat bist du großer Söhne,␤Volk, begnadet für das Schöne,␤vielgerühmtes Österreich,␤vielgerühmtes Österreich!␤␤Heiß umfehdet, wild umstritten␤liegst dem Erdteil du inmitten,␤einem
..stark…
jnthn tadzik: Sounds fun :-)
wallberg masak: so you are leaving for a while=
masak: ?
masak wallberg: Wed/Thu
wallberg masak: do you like sushi? 20:33
masak wallberg: yes. very much.
wallberg masak: how we discuss Uppsala.pm over some sushi on thursday the 11th of Nov? 20:34
masak: how about...
masak sounds nice.
I'd like that. 20:35
wallberg masak: goodie, www.yukikossushi.se/ it is then :-)
tadzik oh, the land of Korpiklaani 20:36
masak \o/ my favourite sushi place :)
wallberg masak: excellent. 12:00 is ok? 20:37
masak yes.
masak writes that down
wallberg masak: its a date :-)
masak \o/
if you write it 2010-11-11T12:00+01:00, it's even an ISO 8601 date :P 20:38
jnthn I so couldn't be bothered to write all that out though... 20:39
20:43 ggoebel left
masak there are various abbreviations. 20:44
all of them very sensible, actually.
the above is just the full form.
20:46 meppl joined 20:50 MayDaniel left
masak seems I didn't pay attention when moritz_++ submitted rt.perl.org/rt3/Ticket/Display.html?id=78510 20:58
anything that affects JSON affects pls, too. that's what I get for not wanting any dependencies.
oh well; it gives me something to November-blog about today. :) 20:59
heh -- I managed to apply moritz_++' json commit to the pls branch of the proto repository :) git rocks. 21:03
21:03 patrickas joined, risou left
patrickas o/ 21:04
tadzik \o
wallberg masak: lol Google Calendar claims that our lunch on the 11th is in 37 hours from now 21:05
masak wallberg: I should never have brought Tardis back online... :/
patrickas masak: I time traveled for the first time through a perl6 program today \o/
masak patrickas: you did? great!
patrickas: any first impressions?
wallberg: this is what it's like on this channel, every day. it's awesome :) 21:06
patrickas masak: I gotta do this more often!
masak :)
patrickas that was my first impression :-)
wallberg :-) 21:07
masak patrickas: cool. maybe you'll help me to conspire over how to give Yapsi first-class closures?
I bet Yapsi and Tardis will have to evolve in lock-step on that feature. it's a fundamental one.
patrickas masak: I'd love to but that seem way out of my league! 21:08
21:08 aesop_ left
patrickas masak: which is exactly why I want to take a stab at it :-) 21:08
masak patrickas: secret fact: writing a compiler looks scary, but it's actually quite nice.
sorear Why are people so afraid of writing compilers? 21:09
It's fun!
masak I agree.
one does need to think clearly about things, yes. but it's not particularly difficult or mentally taxing.
well, sometimes it is. right before a refactor, for example :)
21:09 alester joined
tadzik I have a Perl 6 toylisp compiler lying around, shame I have no idea about how a compiler should be written :) 21:09
masak tadzik: wanna help? 21:10
:)
tadzik masak: with what?
masak Yapsi.
tadzik well, I may if I'll be able to
patrickas masak, sorear: I have no doubt... I was hoping to get at least a whole day to familiarize myself with yapsi...
tadzik I remember reading the code some time ago
masak: wanna see my attempt?
masak sure! 21:11
21:11 aesop joined
masak always nice to read Perl 6 code. 21:11
tadzik wklej.org/id/411688/
masak reads
ok, since there's three of us here (patrickas, tadzik, masak), let me take some time away from my November blogging to explain what I want to work in Yapsi.
masak prepares gist
21:12 aempirei joined, aempirei left, jhuni left
masak tadzik: cute. that looks like a compiler to me. 21:12
it's not an interpreter, since it produces some code rather than execute the AST. 21:13
tadzik ywis
it produces a working PIR
sorear Attempting to classify tools into compilers and interpreters is folly
masak tadzik: I just reached that conclusion too :) 21:14
cool. very cool.
tadzik well, I don't think the algorithm to run... hmm, example needed
masak tadzik: maybe you'd like to create a PIR backend for Yapsi. that'd be nice to have.
sorear There is, to my knowledge, only one interpreted mainstream programming language from the last 45 years
masak BASIC?
sorear Tcl 21:15
masak ah.
sorear every BASIC I've seen had a compile-time tokenizer
tadzik like if we have this (say (sum 5 2)). So (sum 5 2) has to be executed first, so code for it is have to be produced and printed first. And the result has to be available for say, which is later
sorear there's a tradeoff between doing stuff now and doing stuff later
but it's a continuous one
tadzik and I didn't really new how to achieve this 21:16
sorear trying to find the dividing line between compilers and interpreters is like trying to find the point on the Antartctic coast where the Atlantic and Pacific meet
21:17 justatheory joined
masak tadzik: doesn't recursive-descent parsing (which Perl 6 gives you) do that for you? 21:17
tadzik: anyway, there's ample code for doing that (and other necessary tricks) in the Yapsi codebase. 21:18
tadzik masak: well, the thing is, I have to know the available regi... or maybe not
21:18 kensanata joined, Guest86803 left
sorear "Let's Build a Compiler" by Jack Crenshaw, ca. 1985, is required reading 21:18
masak what you need to do is annotate the AST nodes on the way up, yes. but that's just a SMOP.
tadzik so you say this should be able to be implemented in just grammars and actions? 21:19
sorear copies are readily available on the internet
masak sorear: FWIW, I'm reading it now after you mentioned it the other day.
Tene masak: let me know what you think of it.
tadzik I don't really understand this last sentence of yours, masak
masak tadzik: I'm saying that's all you need, yes.
Tene: ok.
tadzik masak | what you need to do is annotate the AST nodes on the way up, yes. -- I don't get it 21:20
masak tadzik: (say (sum 5 2)). a recursive-descent parser starts with the small pieces, so it does (sum 5 2) before it does (say ...)
tadzik yes 21:21
masak tadzik: that's your execution order right there.
tadzik hmm, right
masak tadzik: now you just need a way for the little expression to communicate (a register) to the bigger expression.
that's what `make` is for.
tadzik so make is gonna return a register with the result? 21:22
masak if it does, that information will be available from the action method which handles (say ...)
and that seems to be what you need.
tadzik make sense. Bah, why do good ideas always come in late hours 21:23
masak :)
I can tell you again tomorrow morning if you prefer.
tadzik so in my case, stuff like builtin_say, builtin_sum will be just those action methods?
and my compile sub is actually a recursive-descent implementation of some sort? 21:24
masak looks again
well, the recursive-descent parsing is done by the Toylisp grammar. 21:25
21:25 wamba left
arnsholt tadzik: Subrule calls like <leest> do recursive descent parsing 21:26
masak but yes, if I'm correct, then you could expand your action methods so thet &compile, &builtin_say and &builtin_sum are all baked into the Toyactions class.
tadzik mhm. Experimenting time
arnsholt masak: You're entirely right, AFAIK 21:27
sorear out 21:28
21:28 kanishka joined
masak it's quite odd how top-down recursive descent translates into bottom-up action methods. 21:28
21:28 snake_ joined
moritz_ well, *all* action methods are bottom-up really 21:29
21:29 mberends left
arnsholt Yeah, you could do the same with a YACC parser 21:30
masak not necessarily. if you can promise that you won't backtrack out of anything (or if it doesn't matter), then you could theoretically trigger action methods at rule entry. then it would be top-down.
21:31 MayDaniel joined
arnsholt The difference is which derivation the parse gets 21:31
Well, true. But those actions wouldn't have access to the results from actions in subrules (for obvious reasons)
moritz_ which is why Perl 6 has the action methods at the end 21:32
masak well, but, does that mean that a bottom-up parse would have top-down action methods? 21:33
moritz_ no
all action methods are bottom up 21:34
in a top-down approach, the buttom rules still finish before top rule
masak ah. 21:35
tadzik heh, I can compile hello world :) 21:37
21:37 donaldh joined
colomon sorear: I'm under the impression that classic Microsoft Basic (as found, say, on a Commodore 64) is strictly interpreted. 21:38
but I've not seen their source, and admit I could be wrong.
21:39 stepnem left 21:44 stepnem joined
colomon I guess they might have parsed the code as it was entered, but the internal representation was the source code, for all intents and purposes. 21:45
21:46 icwiener left
moritz_ at least the early Basic interpreters didn't support any means to store a byte code 21:47
that only came in QuickBasic
21:48 Ross^ left
moritz_ -> sleep 21:48
21:49 Ross joined, Ross left, Ross joined, meppl left
masak sleep well, moritz_. dream of signed CLAs. 21:50
21:50 Alias left
sjohnson masak! 21:52
masak sjohnson! \o/
sjohnson oh knowledable perl6 blogger
masak you're too kind. 21:53
slavik1 yeah, masak doesn't know anything 21:54
:P
masak that's more like it.
slavik1 masak: how do I do that weird complicated thing I dreamed about 10 years ago, in Perl6? 21:55
patrickas slavik1: there is an operator for that!
actually there is an op for that! 21:56
sjohnson i think i had the same dream
masak slavik1: your question becomes less and less decent every time I re-read it... :)
and sjohnson just made it worse! :P 21:57
frettled It's almost straight out of a Biosphere song…
slavik1 haha!!!!!!!!!!!!!!!
wasn't the intent, but still good
frettled (www.youtube.com/watch?v=8IHZKm-gZAA)
sjohnson this is going to sound pretty lame, but i actually have more dreams about p6 than i do about girls 21:58
masak anyway, I don't think you can do that in Perl 6 without first doing 'use MONKEY_GROPING;'
frettled (I think, maybe I misremembered the song)
dukeleto sjohnson: there are commit bits for that ;)
rjbs masak++ # less and less decent
masak sjohnson: girls come and go, Perl 6 is forever.
jnthn Oh my, I come back from a walk and the first thing I see is "monkey groping"...
masak 哈哈 21:59
frettled :D
jnthn Do I want to know how we got to this... :-)
dukeleto masak: i heard that GORILLA_GROPING was going to be in the next Rakudo dev release, is that still planned?
masak jnthn: no, you don't. do yourself a favour and don't backlog :)
rjbs Only in Ubuntu "Groping Gorilla"
frettled hehehe
masak 哈哈
frettled masak: you may be too late, he's not responding 22:00
jnthn Is that the release that comes after "Touching Tapir"?
sjohnson :)
masak the more I listen to you, the less I want to know about your dreams.
jnthn
.oO( Maybe *this* is what Ubuntu needs to "go mainstream" ;-) )
I've only had two dreams recently. They both sucked. 22:01
22:01 justatheory left
jnthn In one I couldn't find the laundry room in my appartment block. 22:01
masak you told me about that one. 22:02
jnthn I think it's still preferable to groping monkeys though.
masak that was the same morning I had dreamed about some pleasant monkey groping.
jnthn I...didn't remember it being with a monkey? 22:03
masak no, it was with a cute girl.
jnthn
.oO( Are we at #perl level yet? :P )
masak and it was actually very innocent. but still nice.
jnthn Very cinematic, but saldy not 18-rated. 22:05
22:05 l00rl4nd joined
masak right. it was more like a romantic novel. 22:06
jnthn <disturbed>can we go back to talking about Perl 6 now?</disturbed>
flussence rakudo: say <yes no>.pick(1) 22:07
22:07 MayDaniel left
p6eval rakudo 5f5bae: OUTPUT«no␤» 22:07
jnthn :P
frettled ROFL
jnthn Fine, monkey choking it is then.
er
flussence I'm gonna see if I can do anything about that binary parsing idea I had a few days ago...
jnthn GROPING
argh!
22:08 justatheory joined
flussence is there any way to introspect an object and get a list of its properties, in the order they're defined in the source? 22:08
jnthn flussence: .^attributes should give you them in the right order.
flussence: We depend on that because we use it for the order to initialize them too.
e.g. so has $.a; has $.b; has $.c = $.a + $.b; 22:09
22:09 l00rl4nd left
masak flussence: you just started a meme. now we can do 'rakudo: say <yes no>.pick(1)' on every rethorical question. :) 22:09
jnthn flussence++ # The topic. He's on it.
22:10 Bzek_ joined
flussence (I seem to have terrible luck, last time I tried a random output line like that it made rakudo declare it didn't exist...) 22:10
tadzik how about just .roll? 22:11
that's how we roll
jnthn rakudo: my @rick = <you never gonna give up>; say @rick.roll
p6eval rakudo 5f5bae: OUTPUT«never␤»
flussence rakudo: <pick roll>.pick(*).roll(1).say 22:12
p6eval rakudo 5f5bae: OUTPUT«roll␤»
jnthn rakudo: my @rick = <you never gonna give up>; say @rick.roll(*)
flussence ok, roll it is.
p6eval rakudo 5f5bae:
..OUTPUT«(timeout)ayounevernevernevergonnaupgonnagonnagonnayougonnanevergivegonnaneveryougivegiveyougivenevergonnagonnanevernevergonnaupupupupgiveyouneveryouupyouyouyougonnagonnaupupgiveupgiveyougiveneveryouyounevergonnayougivenevergonnaupyouneveryougivegonnagonnagivegonnayouupgivegiveupnev…
22:12 Bzek left 22:13 jaldhar joined
masak patrickas, tadzik: here's what I'm currently thinking for Yapsi: gist.github.com/660396 22:15
oh, and blog post! strangelyconsistent.org/blog/novemb...ing-action 22:16
looking good, two days down, 28 to go. :) 22:17
jnthn Quick, somebody extend November!
;-)
patrickas well if julius and augustus ceasar can do it, we can't we? 22:18
22:20 Ross left
patrickas 'night all o/ 22:37
22:37 patrickas left
dukeleto gives jnthn a standing ovation for rick-rolling #perl6 22:39
masak gives jnthn a golf clap
frettled better than claps, I suppose 22:41
jnthn That'd be an unfortunate consequence of a hole in one. 22:43
sjohnson the clap 22:44
22:45 amkrankruleuen left
jnthn Aye, getting the clap multiple times would be...well, wrothy of a golf clap I guess. 22:45
sbp phenny: "哈哈"?
phenny sbp: "Haha" (zh-CN to en, translate.google.com)
jnthn phenny: "啤酒"? 22:46
phenny jnthn: "Beer" (zh-CN to en, translate.google.com)
22:46 snake_ left
jnthn phenny: "трубкозуб" 22:47
Aw. :(
sbp you need to append a "?" 22:48
jnthn phenny: "трубкозуб"?
phenny jnthn: "aardvark" (ru to en, translate.google.com)
jnthn \o/
sbp++
diakopter .g трубкозуб 22:49
phenny diakopter: ru.wikipedia.org/wiki/%D0%A2%D1%80%...1%83%D0%B1
sbp good engraving thereon
masak phenny: "Jag kan äta glas och det gör inte ont."? 22:50
phenny masak: "I can eat glass and it does not hurt." (sv to en, translate.google.com)
masak seems to work fine.
bobkare heh, rather ugly fail from google though 22:51
jnthn ? 22:52
Looked correct to me.
masak phenny: "Efter dagens diskussion, hur ska jag rensa de tafsande aporna ur min hjärna?"?
phenny masak: "After today's discussion, how should I clean the groping monkeys out of my brain?" (sv to en, translate.google.com)
masak jnthn: maybe a phrase to learn by heart.
jnthn Can't wait to use that expression at work... :P
masak hah
jnthn "aporna" - porn?! 22:53
masak "the monkeys"
jnthn I know but...
Etymology... :-)
bobkare jnthn: last time I ate glas in sweden it tasted lots better than glass (icecream)
sbp cognate to ape by any chance?
masak yes.
jnthn oh!
I wish I could see these things. 22:54
masak German has "Affe", I think.
tadzik masak: reading
sbp no you're lucky, otherwise you start thinking that an apiary is a hive for apes
22:56 wallberg left
jnthn phenny: "Hur läge ska du kyla en person som har blivit brännskadad!"? 22:58
phenny jnthn: "How mode, cooling a person who has been burned!" (sv to en, translate.google.com)
masak jnthn: "läNge" 22:59
tadzik has got his Uni C++ project by email. First though: oh, easy, one class 3 roles. Oh wait...
jnthn masak: Oh. 23:00
masak: That makes much more sense. :-)
(Test data in a $dayjob project...apparently contains spelling mistakes....)
masak :)
jnthn It actually does have a picture of a person on fire too :-) 23:01
As well as the epicest bbq on fire I've ever seen.
Please tell me this test question isn't as stupid as I think it is...
phenny: "Klicka på brandsläckaren i bilden!"? 23:02
phenny jnthn: "Click on the extinguisher in the picture!" (sv to en, translate.google.com)
jnthn ...how could anyone get *that* wrong?!
masak you'd be surprised.
some people click on the bathtub in the next room. or the cat under the table.
jnthn True. Given the state of this codebase, people's "achievements" should probably surprise me less... 23:03
23:11 mberends joined
masak mberends! \o/ 23:11
23:12 donaldh left 23:16 ggoebel joined
mberends ahoy masak! the Amsterdam.pm group now knows all about 6model, I gave an unplanned half hour talk :) 23:18
23:20 stkowski left
jnthn mberends: ooh, nice 23:20
mberends: How was it received?
mberends quite nicely, although some took the long term view: Haskell, Parrot, CLR, JVM, what next? 23:22
masak not sure I understand what "what next?" refers to here.
are they asking why we target many things?
23:23 cdarroch left
masak or are they sincerely wondering if we have more things we want to target? 23:23
mberends the implication seemed to be that P6 is still in such a state of flux, that it has not settled down on a specific platform. 23:24
masak that is, to some extent, true. 23:25
jnthn I'm not sure I understand the notion of "settle down on a specific platform"
masak it's also, I think, a very p5-centric view :)
jnthn If there's multiple implementations, it's unlikely they'll all target one platform.
masak right.
jnthn Heck, 6model is already trying to work on 3. :-)
Which, I grant, is quite enough to be taking on at the moment.
23:26 rgrau_ joined
mberends markov was worried that by giving most attention to the high level language issues, some low level primitives such as inter thread communication were being overlooked. 23:26
masak that's a reasonable worry. 23:27
jnthn Concurrency is probably the next big thing I want to work on after the meta-model work.
And async.
masak in many ways, the top-down approaches have been more successful than the bottom-up approaches in the Perl 6 world. and it shows.
jnthn And parallelism.
mberends and that such things need to be addressed early on, because it's much more difficult to add nicely afterwards. 23:28
masak unless you count Rakudo as a bottom-up effort.
mberends: indeed.
23:28 wamba joined
jnthn But without some earlyish implementation to play with, we'll end up with abstraction astronautism rather than soemthing that works. 23:28
masak I bet that's happened several times with Parrot already, to various degrees. 23:29
at least Rakudo seems to be informing Parrot design and refactors nowadays.
jnthn masak: It's unlikely Parrot will be my first resort for prototyping anything from hereonin.
I expect the concurrency stuff to be explored on another backend first and then, as you suggest, inform Parrot design. 23:31
23:31 _kaare left
masak I'm not so sure that's a bad thing. it sounds sane to me. 23:33
jnthn I don't consider it a bad thing.
masak in fact, most of my refactors nowadays start with a small "build from scratch" side project, which I then try to merge in.
keeps things simpler.
jnthn Yes, the isolation helps. 23:34
tadzik rakudo: Date.new('2010/09/08') # LTA error message
p6eval rakudo 5f5bae: OUTPUT«Method 'year' not found for invocant of class 'Str'␤ in 'Date::new' at line 6914:CORE.setting␤ in main program body at line 22:/tmp/wJ1V8_ZKfA␤»
masak submits rakudobug
masak resists fixing it right away 23:35
23:35 whiteknight joined
frettled resistance is futile 23:35
23:36 __david__ left
tadzik is tempted to fix it just to see how it should be fixed and why 23:36
masak there are a couple of .new multi methods in the Date class. offhand, I don't see any of them matching the above invocation.
tadzik I wonder if it's the one in #452 23:37
as it's the only one which takes a string 23:38
masak right, but that regex should fail.
tadzik ywis
masak tadzik: anyway, I'd start there too. make it emit debug output in the where clause or something.
tadzik a constructor using strptime could be nice, no? 23:39
masak don't overcomplicate core, please :)
many things "could be nice"...
tadzik alright, alright :)
that's Datetime::Parse job
masak nod 23:40
and now, I really should sleep a little. early plane to catch.
tadzik hmm 23:41
masak 'night, #perl6. you've been wonderful.
tadzik 'night masak
23:41 masak left
tadzik rakudo: multi foo(Str $date where { $date ~~ /^ <[0..9]>**4 '-' <[0..9]>**2 '-' <[0..9]>**2 $/) { say 'complex foo' }; multi foo(Str $asd) { say 'simple foo' }; foo('2010-08-08'); foo('2010/08/08'); 23:42
p6eval rakudo 5f5bae: OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 22␤»
tadzik rakudo: multi foo(Str $date where { $date ~~ /^ <[0..9]>**4 '-' <[0..9]>**2 '-' <[0..9]>**2 $/}) { say 'complex foo' }; multi foo(Str $asd) { say 'simple foo' }; foo('2010-08-08'); foo('2010/08/08');
p6eval rakudo 5f5bae: OUTPUT«complex foo␤simple foo␤»
23:45 justatheory left