»ö« | perl6.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo:, alpha:, pugs:, std:, or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by lichtkind on 5 March 2010.
azert0x hum, or have a nice day :-s 00:00
00:06 yinyin joined 00:12 thisisnick joined, thisisnick is now known as radiation
radiation perl is built into Ubuntu right? 00:12
00:14 radiation left 00:15 pnate left 00:16 lest_away is now known as lestrrat 00:25 ruoso left 00:40 meppl left
dalek kudo: 5219aa5 | (Solomon Foster)++ | src/core/ (4 files):
Move Int.Bool to Real.Bool, tweak definition of Rat.Bool, remove Num.Bool.
00:46
colomon raduko: say 4.Num div 3.Num 00:49
rakudo: say 4.Num div 3.Num
p6eval rakudo 359dbd: ( no output )
00:49 isBEKaml left
colomon rakudo: say 4 % 2.5 01:00
p6eval rakudo 359dbd: ( no output )
01:01 azert0x left
jnthn hacks on a little Rakudo fix before sleep 01:01
01:02 Chillance left
jnthn rakudo: my (Int $x, Num $y); say $x; say $y; 01:02
p6eval rakudo 359dbd: ( no output )
jnthn rakudo: grr, you're broken
p6eval rakudo 359dbd: ( no output )
colomon spec's a little broken, too. 01:04
jnthn colomon: spec or spectest? :-)
colomon (looking at div, %, and mod) 01:05
spec
jnthn Patch away! ;-)
> my (Int $x, Num $y); say $x; say $y;
Any()
tssk...fejl.
colomon only trouble is knowing what the right thing to do is. :)
jnthn: here's the deal, if you don't mind me talking it through with you. 01:06
jnthn Sure, I'm oddly awake given it's 3am, so I'll maybe even follow. :-)
colomon % is defined as $x % $y == $x - floor($x / $y) * $y 01:07
that's for all Real types.
and that makes sense, IMO.
mod is defined as $x mod $y == $x - ($x div $y) * $y 01:08
mod is supposed to fail if the types of $x and $y are different.
div is generic division, also supposed to fail if the types are different. 01:09
jnthn OK.
"types are different" ==> the types have to be the exact same thing?
01:10 whiteknight left
colomon "Not coercive, so fails on differing types." 01:10
jnthn e.g. an Int and an Int is fine, a Rat and a Rat is fine, but a Real and a Real is not?
er
colomon jnthn: that would be my understanding.
jnthn Two things that satisfy Real I mean
OK
eternaleye What if the second arg is a subclass of the first arg's class?
::T, T would match, but potentially be incorrect for that case 01:11
colomon (though that seems a little daft... I mean, it could always just dispatch to / or % if the types differ)
jnthn eternaleye: Yes, that's not quite tight enough.
You'd want something like
colomon anyway, here's the weird bit.
jnthn :(::T1 $a, ::T2 $b where { T1 === T2 })
colomon for integer types, $x div $y == floor($x/$y); 01:12
jnthn Aye, so for integer types div works out just as % would, essentially? 01:13
colomon so for integer types, the definitions for mod and % are the same, just mod guarantees the resulting type is the same as the input type.
jnthn er, mod, yes.
colomon yes.
eternaleye jnthn: Would omitting :: on the second T and using .WHAT in the where potentially be more efficient, due to only running the where in the case where it's a subclass or equal?
01:13 c9s joined
jnthn eternaleye: No, it'd be less efficient under the current implementation. 01:13
eternaleye Huh
colomon what strikes me as weird is that, as I understand it, for non-integer types, div is the same as / 01:14
jnthn eternaleye: Oh, sorry, I mis-read
jnthn reads again
eternaleye :(::T $a, T $b where { $b.WHAT === T })
jnthn Ah, OK
colomon but in that case, for non-integer types, mod always returns 0. 01:15
jnthn But there you're specifying *two* constraints on $b
That it has to be T, and that it has to match the thingy afterwards.
In the long run, the second would be more optimizable.
At the moment, you end up running two checks rather than just one. 01:16
eternaleye Hm, true. It all comes down to the probability of the where clause being run
jnthn colomon: Hmm.
colomon: But div also enforces same type too? 01:17
eternaleye One thing that might be nice would be some sugar for 'subclasses not allowed' .oO( :( ::T $a, T! $b ) ? )
jnthn Of course, so did mod...
eternaleye: I'm sure that in an ideal Perl 6, that'd be possible in a module. :-) 01:18
colomon div and mod enforce the same type, yes.
jnthn colomon: So IIUC, the upshot is that essentially div and mod end up being constrainty rather than coercy versions of / and %? 01:19
colomon jnthn: I think that's the idea.
jnthn colomon: Is that...actually useful? :-) 01:20
colomon but my complaint is that (as I understand the definition of div), mod is useless for non-integer types.
jnthn: I have no idea if constrainty versions are useful. 01:21
01:21 masonkramer_ joined, masonkramer left, masonkramer_ is now known as masonkramer
colomon Going backwards --- div on Ints returns an Int instead of a Rat. 01:22
that makes sense.
jnthn OK, and makes it useful.
Since it gives you a distinct result from / 01:23
So it seems that at least for div it's more useful than "this just constrains" 01:26
colomon right, I'm cool with that.
it's mod that's got me shaking my head.
jnthn Yeah
I can see the problem, but I'm not sure what a fix would be. 01:27
colomon me neither.
jnthn Int $limit = * # parameter declaration FAIL 01:28
(because * !~~ Int)
Just fixed the bug where we didn't check these and...I get spectest fails due to a few bits of sloppiness. :-)
01:43 alester joined 01:47 masonkramer left 01:56 lestrrat is now known as lest_away 02:05 lisen joined
lisen @arnsholt 02:05
what is this 02:07
some people do? 02:08
02:08 lisen left 02:10 [mark] joined
jnthn Hmm...still a couple of fails. I'll finish up the patch after sleep. o/ 02:12
colomon \o 02:14
sleep soon here, too, but I'm trying to get out a blog post first.
02:14 lest_away is now known as lestrrat 02:21 am0c joined
colomon loliblogged: justrakudoit.wordpress.com/2010/06/...operators/ 02:24
dalek kudo: a1193fe | (Solomon Foster)++ | src/core/ (2 files):
Generic version of infix:<%> for Real, and change the generic infix:<%> to
02:26
02:39 plainhao left 02:45 masonkramer joined
lue hai 03:13
Are there any LHFs I might be interested in? [even ones I might not be :)] 03:15
03:32 hudnix left 03:53 agentzh joined 04:05 masonkramer_ joined 04:06 satyavvd joined 04:28 tedv left 04:41 c9s left, c9s joined 04:56 jaldhar_ joined, _jaldhar left 05:05 jaldhar_ left, _jaldhar_ joined
eternaleye lue: Well, rt.perl.org/rt3/Public/Search/Simpl...ml?q=[LHF] has some (go for the ones with 'open' status) 05:10
05:12 _jaldhar_ left 05:13 _jaldhar_ joined
lue thank you 05:16
05:17 masonkramer_ left 05:23 pnate joined
lue afk 05:46
05:46 Wolfman2000 left
[Coke] 4 05:51
05:54 justatheory left 06:02 uniejo joined 06:05 kaare joined, kaare is now known as Guest54945 06:08 alester left, rgrau left 06:12 ive joined 06:20 BrowserUk joined, finanalyst joined 06:21 satyavvd left 06:28 shrub joined 06:31 Su-Shee joined 06:37 plobsing left 07:11 _jaldhar_ left, _jaldhar_ joined, skids left 07:12 skids joined 07:13 Trashlord joined
sorear what happened to pugssvn? 07:27
07:30 ive left
moritz_ good question; I'll investigate 07:44
feather3 shows me the motd, but doesn't give me a shell 07:46
07:47 dju left 07:48 dju joined 07:50 pugssvn joined
pugssvn r31151 | pmichaud++ | [t/spec]: Fudge a new test in S12-class/basic.t (RT #65022) 07:50
r31152 | pmichaud++ | [t/spec]: Fudge a new test in S02-builtin_data_types/hash.t
r31153 | sorear++ | [viv] Explicitly use Encode in output. This ensures that the prototype of Encode::encode_utf8 is visible at BEGIN time, which was previously guaranteed by using Cursor, but broke after the Cursor split.
07:50 moritz_ sets mode: +v pugssvn
pugssvn r31154 | pmurias++ | [mildew] dumping an array of object 07:50
r31154 | (for example the contents of registers) works
r31155 | masak++ | [S32-str/encode.t] refudged
r31156 | colomon++ | [t/spec] Add Real tests for .Bool, infix:<+>, infix<*>, and infix</>.
r31157 | colomon++ | [t/spec] Tests for infix:<%>.
07:50 ajs joined
moritz_ rakudo: try { exit 0 }; say "alive" 07:52
p6eval rakudo a1193f: ( no output ) 07:53
moritz_ rakudo: try { exit }; say "alive"
p6eval rakudo a1193f: ( no output )
moritz_ \o/
07:56 wallberg joined 08:01 [particle] left 08:02 rgrau[work] joined, [particle] joined
moritz_ www.reddit.com/r/perl/comments/cbey...nt_to_see/ 08:05
Su-Shee ah. so the short answer is "all of them, as usal" ;) 08:07
moritz_ :-) 08:08
Su-Shee my choice is a new, native one based on clutter/cairo xul and css based. ;) 08:10
mathw well you can write that
I'll stick to using GIR to generate GTK+ bindings
08:12 _jaldhar_ left
Su-Shee noone for the poor qt folks? ;) 08:12
08:12 _jaldhar_ joined
mathw I don't use Qt 08:17
I'm sure someone will want it though
moritz_ Qt used to be GPL-only unless you pay
is that still the case?
mathw no 08:18
that hasn't been the case for a long time
Su-Shee no, they changed that in the last century. ;) 08:19
moritz_ I'm *so* 19ths :-) 08:22
yesterday I've learned that I'm a "digital native"
as opposed to those born before 1980, who are called "digital imigrant"
of course it makes sense to have such a hard, arbitrary boundary </sarcasm> 08:23
08:23 dakkar joined
Su-Shee then don't read social science on twitter level. ;) 08:25
moritz_ :-) 08:26
pugssvn r31158 | moritz++ | [t/spec] small unfudges for rakudo 08:28
sorear moritz_: it's a good thing rakudo can ALREADY do Tk! 08:29
moritz_ sorear: aye 08:30
Su-Shee *shiver*
moritz_ it seems that wxwidgets already acts as an abstraction layer for GUI toolkits 08:31
so when you interface wxwidgets, you already have gtk support too
Su-Shee yes. and another bad looking, clunky gui application giving "GUI" a bad name. ;) 08:32
moritz_ what makes wxwidgets bad looking? 08:33
I mean more than the usual GUI toolkit :-)
08:33 BrowserUk left, BrowserUk joined 08:35 [particle]1 joined
Su-Shee moritz_: it's supposedly "native" looking to either the OS or the toolkit binding, but it's usally slightly off. it's not really nice to code and doesn't support actually thriving for a really good gui. 08:35
08:36 [particle] left
moritz_ so, why is it less awesome than DBI for databases? 08:38
is it because the APIs are too different in structure?
Su-Shee err.. well that's not just comparing apples to pears but apples to .. steaks :) 08:39
moritz_ well, both can be tasty
mathw GUI toolkits don't seem to deal with abstraction layers very well
they've all got their particular things
Su-Shee a good gui is usally more than an abstraction in the visuals of a button or a menu.
mathw and to make a GUI look good, you usually need full access to the core library 08:40
Su-Shee exactly.
mathw also because different platforms and toolkits have different UI guidelines and different things work well
and their widgets never precisely map in capabilities or behaviour
Su-Shee also, there's at least two schools of "GUI" - being _extremely_ flexible and able to tweak down to the last pixel or "trying to be standarized" 08:41
mathw I like standardised to a certain extent
but sometimes a good specialist UI is better
Su-Shee indeed. also, the standarized tks got old over the last 5-8 years. and not in a good way. features are missing which web nowadays can do. and that's bad. 08:42
08:42 jjore left
Su-Shee not to mention that programming GUI is complex and requires much work. 08:43
moritz_ that I know :-)
Su-Shee and you really _need_ a widget set either being _very_ good or - if not - then very flexible so you can do your own stuff. 08:44
moritz_ well, I think with Perl 6 we're still in the "embrace" phase from "embrace and extend"
Su-Shee also there seem to be only a handful of variations how widget sets are used (programming-wise).
ok and then there is X ;) 08:45
mathw I don't think we're going to revolutionise GUI programming any time soon 08:47
Let's deal with the underlying language first :)
Su-Shee and your audience is extremely ungrateful. ;) everybody wants a really good gui, smooth to use, fast, not too shiny and not too shabby, extremely flexible but also totally simple.. ;)
mathw although of course GUI programming with Perl 6 will be joyful
08:48 jjore joined
moritz_ I fear that writing a GUI toolkit wrapper isn't -Ofun (except when you can automate most of it) 08:49
08:49 eternaleye left
mathw that's why I want to do GTK+ using GIR 08:49
although there's still unfun stuff
you have to deal with memory issues
but first we need NCI sorting out
so I'm not starting now :) 08:50
08:50 Sarten-X left
Su-Shee and with every other implementation of perl 6 using something else underneath you have to probably do it all over again ;) 08:50
moritz_ that's some I'll worry about when we actually have competing implementations 08:54
mathw yes 08:58
08:58 Sarten-X joined 09:00 BrowserUk left 09:02 eternaleye joined 09:03 masak joined
masak oh hai, #perl6! 09:03
moritz_ oh hai, masak
scottaaronson.com/blog/?p=449 has been a joyful read 09:04
masak reads
moritz_ "Indeed, many of the conjectures (or, in physics lingo, “results”) in this area that the physicists derived using their hocus-pocus methods, were later rigorously confirmed by the mathematicians"
09:05 _jaldhar_ left, jaldhar joined
moritz_ it's true; many papers in physics rely on hairy asumptions, so it's a small miracle that not more of them turn out to be outright wrong later on 09:09
masak we have a similar division in the computer world between people who use algorithms/heuristics which haven't been proven efficient, and those who spend their time proving algorithms efficient. 09:12
09:13 clintongormley joined
masak or any other property besides efficiency. correctness, even. 09:13
moritz_ aye
sadly most programming time isn't about algorithms
masak true. 09:14
moritz_ if you're good, most programming time is about organizing data; if you're bad, it's about (mis)understanding language and library features, and suffering from not properly organizing the data 09:16
at least that's my impression from following perlmonks 09:17
masak I still think about the von Neumann bottleneck sometimes. whether it's real, whether we've overcome it in some concrete sense, or whether we will.
moritz_ from my limited experience I'd say it's real, and it's somewhat lessened by various cache tricks 09:20
actually that's wrong; I've seen lots of cases where I thought that fetching memory is the bottleneck 09:21
but I've never really seen a proof for that
masak I find caching deeply interesting. a friend of mine works with such things. 09:24
Su-Shee if gui is involved,most programming time (about 40%) is about the gui. :/ 09:25
Gothmog_ moritz_: Try valgrind/callgrind/kcachegrind, if you're in doubt.
09:27 _jaldhar joined 09:28 jaldhar left
moritz_ Gothmog_: I will, at some point; I've tried, but haven't really understood what those tools told me (except valgrind, which I found exceptionally useful for finding memory leaks and corruptions) 09:29
jnthn morning, #perl6 09:31
masak morning, jnthn! \o/ 09:34
Gothmog_ moritz_: Use valgrind --tool=callgrind for collecting the data and kcachegrind to view it & enjoy fancy graphs ;-) 09:35
09:35 szabgabx joined
moritz_ fancy, yes. But what do they mean? 09:35
my screensaver also produces fancy graphs, if I configure it appropriately :-) 09:38
Gothmog_ I cannot answer that, as there are different ones. :)
moritz_ I likely need to invest more time 09:39
09:52 _jaldhar left, jaldhar_ joined 09:53 agentzh left
masak it's amazing how archaic some of the build solutions are in some of the older Perl 6 projects. 09:56
masak cleans out html-template 09:57
09:57 jaldhar_ left
moritz_ like, json was until yesterday 09:57
09:57 jaldhar_ joined
masak oh? json got updated? 09:57
moritz_ yes, I removed the Configure script 09:58
masak excellent! the Makefile.in is still there, though.
moritz_ oh
needs removing too, I guess
masak moritz_: have you tried ufo on the repo? 09:59
moritz_ what is ufo?
jnthn What's the current best practice for this? Do we document it anywhere?
er, I mean, Well Thought Out Practice ;)
cosimo masak: does the Digest.pm you have in november repo work with current rakudo? I can't get it to work...
masak moritz_: wow! you removed it independently of ufo? :)
moritz_: github.com/masak/ufo
cosimo: current Rakudo? no. 10:00
10:00 JimmyZ joined
masak cosimo: November still runs on alpha, and nothing else. 10:00
cosimo masak: thanks,
masak jnthn: actually I envisioned WTOP to do some of what ufo now does. 10:01
jnthn masak: There's something vaguely amusing in a UFO providing a WTOP. :-)
masak jnthn: sometimes one needs influences from outside to get going :P 10:02
arnsholt masak: Is "forinsecal" intentional? 10:03
masak arnsholt: I had to look long and hard for a fitting word. so yes, it's intentional. 10:04
it means 'alien'.
moritz_ masak++ # VERY nice README
arnsholt Oh, right =D
moritz_ masak: you should blog that, at some point
masak moritz_: thanks :) I probably should.
jnthn Wow, it actually is a word...
10:05 yinyin left
moritz_ masak: I'd like to change ufo's first line to #!/usr/bin/env perl6 10:07
masak: so that it picks up the first perl6 in $PATH
instead of the one in /usr/local/bin, which doesn't exist on my machine
any objects? what's your preferred way of patching? 10:09
*objections
10:13 ispeak_ joined 10:22 lestrrat is now known as lest_away
moritz_ rakudo: say { :one-and-some{1}, :almost-pie(3.3) }.WHAT 10:24
p6eval rakudo a1193f: ( no output )
10:24 am0c left
jnthn rakudo: broken? 10:24
p6eval rakudo a1193f: ( no output )
jnthn :( 10:25
colomon morning, all 10:30
is there any reason we should define infix:<div> for non-integer types?
10:34 JimmyZ left 10:38 skids left 10:39 am0c joined 10:42 ajs left 10:47 jaldhar_ left, jaldhar_ joined 11:00 szabgabx left 11:06 lest_away is now known as lestrrat
masak moritz_: you now have commit rights to ufo. 11:16
11:18 ruoso joined
masak moritz_: not only the shebang line of the 'ufo' script would need to be changed, but also the location of the PERL6 binary. 11:18
moritz_: I've been thinking of putting in a command-line option to enable Makefiles to be created with 'alpha' as the binary. maybe I should generalize that to accepting an arbitrary path for which perl6 binary to use in the make process. 11:19
11:19 ispeak_ left
jnthn masak: General flag sounds useful. Encouraging people to keep using alpha less so. :-) 11:23
moritz_ masak: PERL6=perl6 is the current state, and works here
masak moritz_: oh, good. 11:24
so maybe --use-alpha is enough, and then we can deprecate it.
moritz_ commit pushed to ufo 11:25
masak moritz_++ 11:26
11:31 Lorn joined
masak mathw: ping! 11:36
moritz_ masak: you can remove json from the 'Repositories that still use human technology 11:37
' list
11:37 finanalyst left
masak already did that. 11:37
moritz_ oh :-)
masak github.com/masak/ufo/commit/19423bc...b7ce0ea409
11:38 shrub left 11:39 IllvilJa joined
masak ok, so there's this person 'ank' who doesn't particularly like Perl 6. he replies to me on Twitter sometimes with comments mainly intended to hurt. yesterday he wrote in an old comment thread on one of my blog posts: use.perl.org/comments.pl?sid=44783&cid=72022 11:40
I have many things I could reply to that, but I don't think it's a good idea for me to write to the guy at all. it'll only make me sad, long term. if someone else wants to reply to him, feel free. 11:41
just don't waste the whole day on him. 11:42
moritz_ masak: ignore it
ruoso bom dia #perl6
moritz_ doesn't make sense to reply to personal attacks from strangers on the internet
masak moritz_: indeed. I am ignoring him.
but I'm kinda flattered to be counted in a group of three along with TimToady and chromatic. :) 11:43
also a bit flattered that he apparently reads my blog posts and their comment threads.
colomon Wait, I get to be one of masak's homies?! Awesome!
masak :P 11:44
colomon I love how he says "won't acknowledge it." Well, no, it's directly contrary to one of the core Perl 6 design ideas.
11:44 rv2733 joined
moritz_ it's actually contrary to perl 5 design ideas too 11:44
but perl 5 wasn't quite as consequent as Perl 6 is (hopefully :-)
masak the author of the comment also authored www.ank.com.ar/jade7/ -- a joke which ends up being mostly tasteless and immature.
moritz_ he seems to invest quite some energy 11:45
unlike the traditional troll 11:46
masak yes. apparently he doesn't feel he's getting through to TimToady, so he chooses other ways to express himself.
sbp don't feed the... 11:49
masak right. that's why I'm letting off a little steam by discussing him here. 11:50
I'm still getting used to being the target of troll attacks. :)
sbp :-)
11:52 wallberg left, envi^home joined
moritz_ masak: comes with your new fame 11:53
11:53 xomas left
colomon Actually, the funny thing about the comment is how he tries to sneak "well-defined" in there. 11:53
masak in Moscow, I heard the phrase 'the famous masak". I replied that they must be mistaken, and that I'm not famous, just loud. 11:54
colomon I mean, I'd argue that p6's operators are well-defined precisely because there are so many of them.
there's a difference between being famous and being loud? ;)
moritz_ masak: famous = loud + somebody hears you being loud
masak moritz_: :) 11:55
moritz_ traditionally people got famous by other people talking about them
masak moritz_: people are most likely thinking of you when they hear my name. :)
moritz_ with the internet it's possible to be famous by talking, and having readers
colomon moritz_: ooo, good point. loud + no one cares is pretty much just a pathetic troll. ;)
moritz_ masak: why should they? 11:56
masak "masak? isn't that the guy with the p5->p6 blog posts? he's awesome!"
moritz_ :/
masak "I heard he wrote that nifty JSON::Tiny module"
moritz_ masak: did people actually say it, or are you making that up now?
jnthn operators are like...prepositions?!
masak moritz_: I'm making that up now. 11:57
sbp "and that I'm not famous, just loud"
moritz_ masak: thought so
sbp what's the difference? :-)
masak moritz_: but I see people mixing us up here on the channel now and then.
moritz_ jnthn: akshually they are like verbs... but I don't have to tell you that
masak moritz_: so most likely you're the famous one, and I'm just the loud one. :P
colomon hell, I've mixed you up a few times myself.
sbp oh, I didn't see the fame algebra beneath
moritz_ masak: I'm mixing us up too... :-)
jnthn moritz_: Yeah, I know. Linguistic grasp fail... 11:58
masak moritz_: me too!
colomon it happens when there was a discussion I wasn't really paying attention to... sometimes I can only remember that one of the M's was involved.
moritz_
.oO( the ^m crowd )
masak mmmm... m.
jnthn The nick space seems to collect around some common prefixes. "m" is one of them, but "pm" also is. :-)
colomon or perhaps, the M & M's. ;)
masak :D 11:59
moritz_ co also
jnthn Wow, yes.
"jn" is pretty lonely though. 12:00
masak considers changing his nick to 'jnasak'
jnthn
.oO( probably because those are the start of a word are unpronouncable )
colomon okay, new expression: M & M & co.
jnthn masak: :P
moritz_ :-)
12:01 ive joined
moritz_ due to unforseen circumstances I'll have less p6 time this week 12:02
need to buy a "new" car, and that's a nasty business
masak moritz_: good luck! 12:03
moritz_: do you want someone else to do the weekly challenge?
moritz_ masak: that would be awesome
colomon I had an idea for the challenge, but I don't know if I'll have the time today to set it up.
masak seems I just volunteered. dang :P
moritz_ I also have an idea 12:04
masak nice. maybe nopaste it?
cognominal mazak++ ufo++
moritz_ the book has sidenodes, which are rendered less-than-awesome
cognominal but I had to go to the source to understand what it was about.
moritz_ could be a challenge to come up with latex code for nice side bar rendering 12:05
+ as a bonus hack it into Pod::PseudoPod::LaTeX
cognominal And the lack of decent syntactical highlighting makes difficult to skim perl 6.
moritz_ vim.
mathw masak: pong 12:06
cosimo moritz_: I thought of doing the very same thing for synopsis documents and smartlinks
the way they are interspersed with the synopsis text is making it a bit difficult to follow the text 12:07
I can try to have a look at that, if someone else cares
moritz_ I do care
and I think that the inline, hilighted test files can go away 12:08
all that js load makes searching the pages very slow
cosimo i just volunteered to have a look then :)
moritz_ \o/ 12:09
pugssvn r31159 | colomon++ | [spec] Change div and mod to be operators on integer types.
masak mathw: would you like to join in the revolution? github.com/masak/ufo 12:14
moritz_ perl6: say 'alive'
p6eval rakudo a1193f: ( no output )
..elf 31159, pugs: OUTPUT«alive␤»
masak cognominal: I agree about the syntax highlighting making it difficult to skip Perl 6 code. 12:15
moritz_ :~$ ./p/bin/perl6 -e 'say "hi"'
Segmentation fault
masak cognominal: anything you would have liked to see in the ufo README that wasn't there?
mathw masak: I love the README 12:16
masak :)
colomon loliblogged: justrakudoit.wordpress.com/2010/06/...v-and-mod/
masak colomon++
mathw masak: yes I'll look at that 12:17
some time
umm
not now
masak mathw: no worries.
mathw reviews his schedule and realises 'sleep' doesn't really feature until wednesday night
masak worst thing that can happen is that you'll be the only one remaining on the 'human tech' list :P
colomon: 'only working on integer types'. what's the failure mode -- coercion, or runtime error? 12:18
colomon I'd argue runtime error. 12:19
cognominal masak: not really, I just hope that all your tools will be integrated in a whole, each being a point in the big whole. Right now, it is like seeing a pointillist painting point by point even if each point is a world in itself.
masak colomon: because operators in Perl generally tend to coerce rather than error out.
colomon The spec already specifies that those two don't coerce. 12:20
mathw masak: from the look of it, it doesn't look like it's going to be too hard to do the convert anyway
I take it you have to rerun ufo if your directory structure changes
colomon I think they have a very specific use, and if you're not there already, you shouldn't be using them.
though I can see an argument that they should handle mixing different integer types.... hmmm... 12:21
masak cognominal: it's the small, rapid solutions which seem to succeed in the long term. I agree about it being important to unify them, but since no-one can see the whole yet, that will have to wait a bit. :)
mathw: no, it's just a question of removing the old files.
and maybe 'make install'-ing ufo.
12:22 colomon_phone joined
mathw ooh 12:23
so you don't check in the generated Makefile
colomon_phone laptop just ran out of juice
mathw your build process is "ufo && make"
masak aye.
cognominal masak: emergent design... something in between dartwinist evolution and intelligent design. Quite appropriate for most human endeavors. 12:25
masak cognominal: I often find that parts of the Perl 6 effort are similar in structure to the Perl 6 effort itself. that probably goes for many communities. but Perl 6 as a whole is definitely emergent/whirlpool, with implementations informing the spec etc. 12:26
12:27 plainhao joined 12:28 |Jedai| left 12:29 pmurias joined
pmurias ruoso: ping 12:29
moritz_ moritz.faui2k3.org/tmp/parabel-small.png # the new logo for Math::RungeKutta
12:29 colomon_phone left
masak mathw: some projects under the proto umbrella have their Makefile committed. I see nothing fundamentally wrong about that; it's perhaps not always convenient that local changes show up as a dirty working dir. but that's not so bad. 12:30
12:30 skids joined, colomon_phone joined
masak the principle behind ufo is exactly the same as with proto originally: "this can be done by hand, but it's, like, boring and stuff". 12:30
mathw also a certain level of standardisation is desirable 12:31
moritz_ rakudo: say 3
mathw especially for simple pure-Perl 6 libraries like Form
p6eval rakudo a1193f: OUTPUT«3␤»
colomon_phone \o/ 12:32
12:32 meppl joined
colomon_phone just remembered he should be able to launch a spectest run from his phone. 12:33
ruoso pmurias, pong 12:34
moritz_ rakudo: say { :one-and-some{1}, :almost-pie(3.3) }.WHAT
p6eval rakudo a1193f: OUTPUT«Hash()␤»
pmurias ruoso: hi
moritz_ rakudo: say { :one-and-some{1}, :almost-pie(3.3) }.<one-and-some>.WHAT 12:35
p6eval rakudo a1193f: OUTPUT«Block()␤»
moritz_ that's curious, ain't it?
jnthn Hmmm. :-)
pmurias ruoso: when testing the m0ld frontend for mildew i found a bug in smop that causes a segfault
jnthn moritz_: It kinda makes sense. :-) 12:36
moritz_ jnthn: but only kinda
jnthn Yeah.
pmurias ruoso: i added a new entry to the RI which also dumping the smop object so it's possible to look at their internals 12:37
masak moritz_: looks like a bug.
moritz_ masak: agreed 12:38
masak moritz_: so... what do we do? submit it?
jnthn Question is what should it do?
masak rakudo: say { :a{1} }<a>.WHAT
p6eval rakudo a1193f: OUTPUT«Block()␤»
moritz_ masak: yes
masak submits rakudobug 12:39
jnthn er
moritz_ jnthn: construct a hash, and warn (odd number)
jnthn But {1} would be a block, not a hash.
masak agreed.
jnthn: the context should probably override it to always be a hash there.
moritz_ jnthn: yes, but :colonpair{postcircumfix} is a special syntactic form
jnthn I think it's correct. 12:40
S02:
a => {...} :a{...} :a({...})
colomon_phone spectest running, \o/
pmurias ruoso: here's an example of how the dumps look like pastie.org/994860
jnthn a => {...} will have {...} as either a block or a hash depending on what's in it. 12:41
moritz_ jnthn: hum. Yeah. How... suprising
jnthn moritz_: Well, if the rule is "what follows the name behaves exactly like the same circumfix would anywhere else"...it's not so surprising.
:comparator{ $^a.foo <=> $^b.foo } is kinda quite too. :-) 12:42
12:42 proller_ left
jnthn *cute 12:42
Thing is, would we expect
:comparator({ $^a.foo <=> $^b.foo }) to be a block there?
moritz_ you're right
pmurias ruoso: should the definitions of functions in smop headers have export? 12:43
masak so it should be a block?
pmurias like export int foo(void);
moritz_ it's just a thought I have to get used too
masak so, no bug report.
jnthn masak: I'm not seeing anything in the spec that suggests otherwise.
masak: And something that suggets it's fine that it is.
12:43 SmokeMachine joined
jnthn I'm sure we'll get clarification when somebody backlogs. :-) 12:43
It did make me do a double take. 12:44
12:44 bluescreen joined 12:45 bluescreen is now known as Guest50900, proller joined
mathw :a({1}) looks like a pair with a block in the value slot to me 12:46
:a{1} could easily be the same
or not
moritz_ \o/ JSON::Tiny is down to 6 failing tests on master
mathw now I've caused doubt in myself 12:47
jnthn mathw: In a sense, it'd feel odd for it not to be the same too, though.
mathw actually it feels odd that it even parses
12:47 JimmyZ joined
jnthn Welcome to Perl. ;-0 12:47
mathw I thought colonpairs were supposed to use () :) 12:48
moritz_ nope
jnthn Nah, there's a load of variants
moritz_ rakudo: say :ab[3, 5].perl
p6eval rakudo a1193f: OUTPUT«"ab" => [3, 5]␤»
jnthn Basically, any dream^Wcircumfix will do.
12:48 Guest50900 left
moritz_ recommends reading S02 or the book at some point :-) 12:48
jnthn Even better, once you can define your own circumfixes they'll parse there too under current STD. :-)
mathw moritz_: It's been a while since I read S02 all the way through
so if [] makes lists, shouldn't {} make whatever {} makes elsewhere :) 12:49
moritz_ that's what we're discussing
mathw I think I've caught up now
moritz_ and after throwing my previous opinion over board I agree with jnthn++
mathw it would indeed make sense 12:50
masak mathw: {} makes blocks/closures.
pmurias ruoso: is there a timespan when you're accesible on irc or is email prefered?
colomon_phone surely [] makes arrays?
mathw masak: does it not also make hashes?
masak mathw: but sometimes there 'collapse' down into ordinary hashes.
s/there/these/
mathw okay
well that can still happen
pugssvn r31160 | pmurias++ | [mildew] move the c emitting from AST::Block::SSA to Mildew::Backend::OptC 12:53
pmurias ruoso: with regards to missing the milestones i plan to crush the bug with the data dumping todayish and try to replace scons this week to meet the release SMOP to CPAN milestone 12:56
ruoso: do you think that's a sane plan? 12:57
masak moritz_: did you say you had an idea for the weekly challenge? 12:59
moritz_ masak: yes, I've described it in the backlog
(fix some latex code for $book)
masak oh; I remember now. 13:00
moritz_ there are other ideas in misc/helpnow/README too
13:00 rgrau joined 13:01 mberends joined
moritz_ web-ish stuff has usually received high participation 13:02
jnthn How about something like "port X module from Web.pm to run on master"? 13:03
masak EMORETHANWEEK
moritz_ that would be nice too, for a module where we expect that to be quite smooth
jnthn masak: *one* module.
moritz_ provided there's a small enough module
13:03 Guest50900 joined
jnthn masak: Not the whole Web.pm. 13:03
13:04 takadonet joined
takadonet morning all 13:04
masak hm. possibly. but it's hardly an independent newbie task.
takadonet: \o
13:04 mantovani joined, am0c left, mantovani left, jhuni left
masak someone could try it outside of a weekly challenge, perhaps. I'd be delighted if someone got Squerl working on master, frex. 13:05
the schedule at the end of misc/helpnow/README appears abandoned. 13:06
moritz_ it is
ruoso pmurias, email is preferred this days... it has been quite hard to be here in a predictable fashion
moritz_ since nobody else committed to taking a date in advance, it was useless for scheduling 13:07
masak moritz_: can I remove it?
moritz_ masak: sure
masak: perlgeek.de/blog-en/perl-6/contribu...ounce.html has an up-to-date list of tasks that have been used already
ruoso pmurias, yes... it seems like a sane plan... but it would be cool if you could mail me more detailed information on where you are now, what you see ahead and so on 13:08
pugssvn r31161 | masak++ | [misc/helpnow/README] volunteered, removed stale schedule 13:09
13:09 [mark] left 13:12 [particle]1 is now known as [particle] 13:16 Guest54945 left
pmurias ruoso: ok i'll try to write something 13:18
13:19 am0c joined
pmurias ruoso: what i'm implementing atm, is ++BACKEND --option-for-backend ... ++/BACKEND as invented by S19 13:19
so we can have ++BACKEND --trace ++/BACKEND and ++BACKEND --valgrind ++/BACKEND 13:20
ruoso I see... is that particular feature close to finish? 13:21
pmurias yes
ruoso okay... then I'd suggest getting back to "releasing SMOP/mildew to CPAN" per se
and adding extra features later, if time is available 13:22
mberends hi #perl6. After a little more work on FakeDBI, I plan to write some serious production code to use it: a little accounting or bookkeeping program, for my own small business. Anyone who is interested will be welcome to begin collaborating when the work begins, maybe as early as next week.
moritz_ loves to hear of production uses of Perl 6 13:23
pmurias ruoso: the things that blocks me in releasing SMOP is that the code mildew turns m0ld into segfaults
moritz_ mberends: how close is FakeDBI to being a subset of perl 5 DBI?
13:23 Woody2143 left
mberends moritz_: it's already CRUD complete, the rest is all optional bells and whistles 13:24
pmurias ruoso: so i want to port of the debuging aids from the old mildew to the new refactored one 13:25
moritz_ mberends: my question is more: does it contain incompatible API changes, apart from things that are NYI or not planned or due to the difference between Perl 5 and 6
jnthn Sheesh, what is it with people in Lund and loud horns today... :S
moritz_ (imagine a question mark at appropriate place)
mberends moritz_: no differences between P5 and P6. It really does work to simply replace -> with . and $arr[...] with @arr[...] and keep everything else unchanged so far. 13:27
moritz_ mberends: ok, thanks 13:28
PerlJam Where is FakeDBI?
moritz_ hugme: show FakeDBI
hugme moritz_: sorry, I don't know anything about 'FakeDBI'
moritz_ hugme: show fakeDBI
hugme moritz_: sorry, I don't know anything about 'fakeDBI'
moritz_ hugme: show fakedbi
hugme moritz_: the following people have power over 'fakedbi': P⁣erlJam, T⁣imToady, [⁣particle], c⁣olomon, j⁣nthn, m⁣asak, m⁣berends, m⁣oritz_, p⁣michaud, s⁣orear. URL: github.com/mberends/fakedbi/
PerlJam heh ... "github" would have been enough of an answer for me :) 13:29
jnthn moritz_: (production use) I'm increasingly using it to do little bits to help along with $dayjob.
Small quick throwaway hacks more than anything. 13:30
13:30 uniejo left
jnthn But it's decidedly being useful. 13:30
moritz_ that's good news 13:32
mberends jnthn: more than 50% change that I'll go to www.perlworkshop.be/bpw2010/
*chance, gah
jnthn What did it change...oh. :-)
50% change from a 5% chance you'd go woulda need less awesome. ;-) 13:33
13:34 Woody2143 joined
mberends how can I say this and not be misunderstood? ;) 13:34
jnthn mberends: Somebody determined enough will always fine a way to misunderstand it. ;-)
mberends: I think they're still accepting talks, BTW, if you get to the point of it being more likely than not. 13:35
Schedule is pretty sparse.
13:35 JimmyZ left
jnthn Wow. There's a talk about "making out". 13:35
moritz_ is bpw only one day?
jnthn moritz_: Aye. 13:36
moritz_: But I suspect there's tasty Belgian beer on Friday *and* Saturday evenings. ;-) 13:37
mberends I'll probably submit a talk about Perl 6 database interfaces to BPW
jnthn mberends: +1 13:39
mberends jnthn: ^^ were you threatening to *fine* me?
jnthn :-) 13:41
mberends: I think I just submitted to waffle about Rakudo *
Which means I'll probably just waffle about Awesome Stuff That Works.
moritz_ hm, waffles are tasty
mberends waffle FTW 13:42
jnthn What other word could I use do describe a talk at a *Belgian* workshop? ;-)
dalek kudo: 6f0d81a | (Solomon Foster)++ | src/ (2 files):
Initial implementation of infix:<mod>. In the long run, should maybe be in
pugssvn r31162 | pmurias++ | [mildew] ++BACKEND ++/BACKEND support 13:47
13:51 colomon_phone left 13:52 proller left 13:54 proller joined 13:57 plobsing joined 13:58 wasy left 14:03 PacoLinux joined
pugssvn r31163 | pmurias++ | [mildew] added the --dump option to dump the frame after every instruction 14:04
14:05 gbacon joined 14:12 am0c left 14:18 alester joined 14:21 pmurias left 14:24 ash__ joined 14:27 am0c joined, sorear left 14:29 rgrau[work] left 14:30 hudnix joined 14:31 colomon left 14:32 sorear joined 14:40 _kaare joined 14:41 plobsing left, kraai left 14:45 patspam joined, masonkramer left, masonkramer joined 14:49 cognominal left 14:51 cognominal joined
masak hugme: show fakedbi 14:51
hugme masak: the following people have power over 'fakedbi': P⁣erlJam, T⁣imToady, [⁣particle], c⁣olomon, j⁣nthn, m⁣asak, m⁣berends, m⁣oritz_, p⁣michaud, s⁣orear. URL: github.com/mberends/fakedbi/
masak moritz_: in the hugme output above, my client shows an unprintable character after each nick's initial character. why is that, you think? 14:52
14:52 _sri left
masak I see nothing similar in the IR clogs. 14:52
mberends: that accounting/bookkeeping idea sounds cool! I can't promise tuits, but I'll make sure to follow along, so I hope it'll be on a reease early/often schedule. :) 14:53
14:53 _sri joined
masak mberends: also, I might try to make fakedbi work with Web.pm Very Soon. unless Tene++ beats me to it. 14:53
14:54 _jaldhar_ joined, jaldhar_ left
jnthn masak: Because your client doesn't know how to display INVIZIBLE SEPARATOR 14:54
mberends masak: :-)
jnthn masak: (which was added to stop doing mass highlights of everyone when querying hugme)
masak jnthn: oh, so it's deliberate to avoid nick highlighting? I shoulda realized that.
right.
jnthn Yeah 14:55
.u INVISIBLE SEPARATOR
phenny U+2063 INVISIBLE SEPARATOR (⁣)
masak "I see invisible separators!" o.O
jnthn EINVISIBILITYFAIL
masak Emacs++ 14:56
rakudo: say "j⁣nthn".chars
p6eval rakudo 6f0d81: OUTPUT«6␤»
masak awesome. :)
now only I will know the truth, and the rest of you will never know when to expect an invisible separator. :> 14:57
rakudo: say "j⁣n⁣t⁣h⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣n".chars
p6eval rakudo 6f0d81: OUTPUT«18␤»
jnthn Oh no...I can see the April Fool's bug report already;.
masak :D
rakudo: say "j⁣n⁣t⁣h⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣􏿽xE2􏿽x81 14:58
p6eval rakudo 6f0d81: OUTPUT«===SORRY!===␤Confused at line 11, near "say \"j\u2063n\u2063t"␤»
skids
.oO(fortunately April is a long way off)
masak rakudo: say "j⁣n⁣t⁣h⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣􏿽xE2􏿽x81
p6eval rakudo 6f0d81: OUTPUT«===SORRY!===␤Confused at line 11, near "say \"j\u2063n\u2063t"␤»
masak seems it can't handle very many INVISUHBLE SEPARATUHRs in a row. 14:59
jnthn masak: I don't see the ending quote here... :-S
rakudo: say "j⁣n⁣t⁣h⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣�"
p6eval rakudo 6f0d81: OUTPUT«j⁣n⁣t⁣h⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣�␤»
jnthn hmm.
14:59 ajs joined
masak rakudo: say "j⁣n⁣t⁣h⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣n".chars 15:00
p6eval rakudo 6f0d81: OUTPUT«108␤»
masak that's better.
jnthn rakudo: <a(⁣b(⁣c>.elems.say
p6eval rakudo 6f0d81: OUTPUT«1␤»
masak nice try :)
jnthn say ?("a(⁣b" ~~ /\s/) 15:01
rakudo: say ?("a(⁣b" ~~ /\s/)
p6eval rakudo 6f0d81: OUTPUT«0␤»
jnthn Ah
Oh of course, it's not white*space* because it's INVISIBLE.
:-)
masak right. INVISIBLE implies zero-width.
skids
.oO(life before unicode was so much simpler)
15:02
masak no, the simplicity was just outside of the standard instead.
er, s/simplicity/complexity/
skids Honestly here in this country back in the early 90's I may have gone whole months between seeing umlauts. 15:03
jnthn That's nothing, in the country I was in during the early 90's I went YEARS without seeing umlauts. 15:04
masak sees umlauts all day 15:05
jnthn Me too now.
Su-Shee starts even hearing umlauts.. ;)
jnthn Well, that's a mark above where I am right now. 15:06
masak jnthn: you're sitting under an umlaut? :)
jñthñ 15:07
jnthn Yeah, and it's driving me dotty!
masak hm, probably not umlauts... :)
jnthn heh, I learn something every day... en.wikipedia.org/wiki/Metal_umlaut 15:08
skids
.oO("Spüriʊüs Ümläüts" would be an awsome hair metal band name.)
masak :) 15:09
jnthn :D
pmichaud "Metal Umlaut" would be a good name for a software release. :-) 15:10
good morning, #perl6
jnthn morning, pmichaud
masak morning, pmichaud
jnthn looks in despair at a CSV file he's meant to be able to parse 15:11
masak jnthn: you're not reinventing the CSV parser, are you?
jnthn It allows *unescaped* quotes inside a quoted string.
masak jnthn: that's CSV standard, last I looked. double quotes inside quoted strings.
jnthn masak: Ew. 15:12
masak I know. I was surprised too.
jnthn masak: I guess it's meant to look ahead after it to check that it doesn't have a delimiter there?
masak en.wikipedia.org/wiki/Comma-separat...cification
it's there among the examples.
jnthn: guess so. don't remember how I did it in the Text::CSV module. 15:13
masak cheks
ooh, nice! 15:14
regex quoted_contents { <pure_text> ** [ <[,]> | \s | '""' ] }
jnthn masak: Ah, no, what I have isn't like:
1997,Ford,E350,"Super ""luxurious"" truck"
It's
1997,Ford,E350,"Super "luxurious" truck"
masak o.O
masak cries
right. I was ambiguous when I said 'double quotes' above. 15:15
I should have said 'double double quotes'.
jnthn Yes. :-)
That's reasonable and sane.
pmichaud or even "quoted double quotes" :-)
masak in what sense are they quoted, apart from being inside a quote environment? 15:16
15:16 colomon joined
masak it's more like they're escaped, and the escape character, insanely, is also a quote character. 15:17
jnthn "" is quite common
(as a way to write a " inside a "..." quoted string) 15:18
Even C# does it in @"...." strings.
skids And yet '' is not, you find as a newbie, when playing with sh.
colomon anyone out there relying on div's behavior in rakudo wrt negative operands? it doesn't match spec atm. 15:19
masak isn't
colomon masak++ ;)
masak hardly relying on div. 15:20
15:20 colomon left
diakopter re gui toolkits, on mono/.NET there is gtk2 15:21
and asp.net, which is a gui toolkit in a manner of speaking 15:22
15:25 macdaddy joined, macdaddy is now known as Guest58430
diakopter looks around for sorear 15:27
15:31 xinming left, xinming joined
masak gist.github.com/428829 # this week's contribution to Perl 6 15:50
comments welcome. if there are none, I'll publish that one soonish.
jnthn glances it 15:52
masak: looks fine
s/visually inspiring/visually appealing/ ? 15:53
or "aesthetically pleasing" :-)
masak thanks. changing.
TimToady U t/spec/S02-builtin_data_types/hash.t
U t/spec/S32-num/real-bridge.t
U t/spec/S32-list/minmax.t
bleah
EWRONGMOUSEBUTTON
masak TimToady: use nopaste for two lines or more :P
moritz_: I'm prepared to either post the challenge on my blog, or have you post it on yours, or both. whatever you think is best. 16:01
16:09 masak left 16:13 Ross joined 16:19 justatheory joined 16:24 hercynium joined 16:25 IllvilJa left 16:34 rv2733 left 16:42 ashleydev left 16:50 _jaldhar_ left
moritz_ masonkramer: post in on yours, I'll write a short note on mine with a link 16:54
erm, meant masak
16:54 cdarroch joined, cdarroch left, cdarroch joined
diakopter phenny: tell sorear i'll be back online mid-afternoon probably 16:57
phenny diakopter: I'll pass that on when sorear is around.
16:58 rv2733 joined 17:04 dakkar left 17:16 SmokeMachine left
sorear messages clear 17:26
phenny sorear: 16:57Z <diakopter> tell sorear i'll be back online mid-afternoon probably
17:27 saaki left 17:29 ashleydev joined 17:33 cosimo left 17:35 cosimo joined 17:36 SmokeMachine joined 17:39 patrickas joined 17:40 cosimo left
patrickas hello 17:40
17:40 saaki joined 17:41 cosimo joined
moritz_ hi 17:41
sorear hi
17:42 ashleydev left
moritz_ phenny: tell masak to please blog the challenge on his own blog 17:42
patrickas I got the one liner at the end of day23 to work with latest rakudo (it previously only worked with pugs) should I update it ?
phenny moritz_: I'll pass that on when masak is around.
moritz_ patrickas: best add it, so that we have two versions 17:43
and comment appropriately
patrickas ok I'll do that, I also got a couple of questions about it 17:44
is %(...){'x'} supposed to work if '...' is a code that can produce a hash ? 17:46
rakudo: say %(<a b c> Z <1 2 3>){'b'}
p6eval rakudo 6f0d81: OUTPUT«Method 'hash' not found for invocant of class 'GatherIterator'␤ in main program body at line 11:/tmp/93TXb9LxCR␤»
17:47 ruoso left 17:51 ruoso joined
patrickas actually I take that back .. the old one liner parses with pugs but produces garbage. 17:52
anyways did I hear you needed someone to port some old code to work with latest master ? 17:54
18:03 envi^home left 18:05 mberends left 18:09 patrickas left
[Coke] p4 18:12
18:18 dakkar joined 18:25 alester left 18:32 rokoteko left 18:35 dduncan joined 18:37 dduncan left
pugssvn r31164 | moritz++ | [t/spec] test hash vs. block distinction with leading ; and placeholders (RT #74050) 18:50
18:51 azert0x joined 18:52 estrabd is now known as estrabd_afk
pugssvn r31165 | moritz++ | [t/spec] < > returns an empty Parcel aka Nil (RT #73772) 18:56
moritz_ std: multi a($:b) { } 19:00
p6eval std 31163: OUTPUT«===SORRY!===␤You may not use the : twigil in a signature at /tmp/ai7BsmU96s line 1:␤------> multi a($⏏:b) { }␤ expecting twigil␤Parse failed␤FAILED 00:01 108m␤»
jnthn moritz_: The patch that adds a note about possible named parameter syntax confusion is OK to add, imho.
(just read the ticket a moment ago)
moritz_ I'll spectest it (not that it shoudl cause any harm...) 19:01
"patch does not apply"
huh
applying it manually works just fine 19:02
if only git-patch was as good as the merging stuff in git... 19:03
Tene git-patch? you mean git-apply? 19:05
moritz_ yes
cognominal rakudo : my &quoted = token (Str $s where { $s.chars == 1 } ) { $s [ '\\' $s | <?before $s> . ]* $s }; say "'a'" ~~ quoted("'") 19:06
rakudo: my &quoted = token (Str $s where { $s.chars == 1 } ) { $s [ '\\' $s | <?before $s> . ]* $s }; say "'a'" ~~ quoted("'")
p6eval rakudo 6f0d81: OUTPUT«Not enough positional parameters passed; got 1 but expected 2␤ in <anon> at line 1:/tmp/jJt0wELyJw␤ in main program body at line 11:/tmp/jJt0wELyJw␤»
cognominal hum
jnthn, any idea of what I am doing wrong? 19:07
jnthn cognominal: It won't work like that. 19:08
say "'a'" ~~ /<&quoted("'")>/
That should work
cognominal I am always burned by trying to put directly rules without the / / for a match, probably because it works in Perl 5. 19:10
And I am a bad student, I keep making the same mistakes
PerlJam "it works in Perl 5"?
somehow I think you're comparing apples and oranges
19:11 pyrimidine joined, ShaneC joined
cognominal here what I means : perl -e ' $a = qr|a|; print 'a' =~ $a' 19:12
1
you know, I don't need to do a /$a/ in the right
PerlJam rakudo: $a = rx/a/; say 'a' ~~ $a; # like this? 19:13
cognominal I can't argue if it would be meaninful in perl 6, but my memory muscles keeps doing that in Perl 6
p6eval rakudo 6f0d81: OUTPUT«===SORRY!===␤Symbol '$a' not predeclared in <anonymous>␤»
PerlJam oops, forgot the my 19:14
rakudo: my $a = rx/a/; say 'a' ~~ $a; # like this?
19:14 Ross left
p6eval rakudo 6f0d81: OUTPUT«a␤» 19:14
cognominal ho
PerlJam like I said, apples and oranges. If you use the equivalent constructs, it works equivalently :)
cognominal yes. 19:15
I will try to understand later why the first form as for a seconde parameter, when the token is defined outside a class. I am not 100% convinced that the right behavior. 19:17
I am stubborn that way, hitting all the walls to learn the correct path 19:18
19:20 Ross joined
Tene cognominal: rules and tokens are methods. you didn't give it an invocant. 19:25
19:25 ashleydev joined
cognominal rakudo : / <&quoted("'") > / 19:25
rakudo: / <&quoted("'") > /
p6eval rakudo 6f0d81: OUTPUT«===SORRY!===␤regex assertion not terminated by angle bracket at line 11, near "(\"'\") > /"␤»
cognominal I don't see them as methods when defined outside a class. 19:26
rx is just alternate syntax for a rule so why does it behave correctly? 19:27
rakudo: / < &quoted("'") > / 19:28
p6eval rakudo 6f0d81: OUTPUT«===SORRY!===␤Confused at line 11, near "/ < &quote"␤»
19:30 plainhao left
dalek kudo: b258f5c | (Stéphane Payrard)++ | src/Perl6/Actions.pm:
Awesommify error message for $:a in signature
19:38
19:38 rv2733 left
arnsholt Does array slicing work in Rakudo yet? 19:40
PerlJam rakudo: my @a = "a".."z"; say @a[5,6,8,12,15];
p6eval rakudo 6f0d81: OUTPUT«fgimp␤»
19:41 SmokeMachine left
ash__ rakudo: my @a = ((2, 1), (3,4)); say @a[1;1]; say @a[1;*]; # that second slice seems wrong... 19:44
p6eval rakudo 6f0d81: OUTPUT«1␤2134␤»
ash__ actually that first slice is wrong too 19:45
hmm
moritz_ rakudo doesn't do any multidim stuff
ash__ maybe those don't work yet
ah 19:46
got ya
jnthn I didn't realize we parsed those... 19:47
ash__ rakudo: my @a = 'a' .. 'z'; say @a[* ~~ /<[a..e]>/] # would be neat if that worked, is that the right syntax? 19:48
p6eval rakudo 6f0d81: OUTPUT«a␤»
19:48 Guest50900 left
moritz_ wonders how hard it would be to add minlength and maxlength info to PAST::Regex nodes 19:48
arnsholt rakudo: my @a = (1,2,3,4,5); @a[1,*].perl.say;
p6eval rakudo 6f0d81: OUTPUT«(2, (1, 2, 3, 4, 5))␤»
cognominal moritz++ made my day by accepting my patch 19:49
arnsholt rakudo: my @a = (1,2,3,4,5); say @a[1,*];
p6eval rakudo 6f0d81: OUTPUT«212345␤»
arnsholt rakudo: my @a = (1,2,3,4,5); @a[1,+@a].perl.say;
p6eval rakudo 6f0d81: OUTPUT«(2, Proxy.new())␤»
moritz_ cognominal++ # submitting patches 19:50
pugssvn r31166 | moritz++ | [S32] fix a grammero 20:01
20:01 Su-Shee left 20:03 Psyche^ joined 20:04 Guest50900 joined 20:07 Patterner left, Psyche^ is now known as Patterner
diakopter moritz_: I like the spello in the commit msg :P 20:11
moritz_ diakopter: for some reason I can't quite put my finger on I always spell "grammar" with "e" :-) 20:13
20:14 pmurias joined
moritz_ rakudo: say &Nil().WHAT 20:15
p6eval rakudo 6f0d81: OUTPUT«Parcel()␤»
moritz_ rakudo: say &Nil().hash.WHAT
p6eval rakudo 6f0d81: ( no output )
moritz_ cognominal++ has the new record for crashing rakudo in a minimal number of steps 20:17
before it used to .HOW.HOW (8)
20:17 colomon joined
moritz_ now it's %() (3) 20:17
where crash = segfault 20:18
ash__ %() (3)? is that casting nothing as a hash then asking for the 3rd element?
moritz_ no, just %()
3 is the number of characters 20:19
ash__ ooh got ya
lol, yup seg faults here too
cognominal moritz_, I am very good at breaking toys. 20:22
moritz_ cognominal: you should team up with masak :-)
20:22 ashleydev left
moritz_ cognominal: want a new toy to break? check out Math::RungeKutta 20:22
cognominal as a production programmer, I am a plague because I alway hit many bugs 20:23
dalek meta: r322 | diakopter++ | trunk/Sprixel/ (8 files):
[perlesque] implement constructors/methods/fields/events/properties on generic
moritz_ github.com/moritz/Math-RungeKutta
PerlJam moritz_++ 20:24
moritz_ book LHF: github.com/perl6/book/issues#issue/4 20:25
(if it's valid, no idea) 20:26
dalek ok: 1eb7efd | moritz++ | src/basics.pod:
[bsics] s:g/games/matches/ to be less confusing
20:35
pugssvn r31167 | pmurias++ | [smop] dumping the interpreter and bools works 20:39
dalek meta: r323 | diakopter++ | trunk/Sprixel/ (3 files):
[perlesque] bugfixes
20:43
20:49 xomas_ joined 20:54 SmokeMachine joined, pragma_ left 20:56 jferrero joined 21:02 kensanata joined 21:03 pragma_ joined, pragma_ is now known as Guest84666 21:07 Ross left 21:09 ashleydev_ joined 21:10 pmurias left 21:14 Guest50900 left 21:30 SmokeMachine left 21:33 whiteknight joined, clintongormley left 21:35 clintongormley joined 21:38 _kaare left 21:39 m635b joined 21:40 pyrimidine left 21:41 m635b left, Guest84666 left, Guest84666 joined 21:42 Guest84666 is now known as pragma_ 21:45 ive left
arnsholt Right. It's Str, not String >.< 21:48
21:52 szabgabx joined
arnsholt What's the idiom to convert a list of alternating keys and values to a hash again? 21:54
(Or is it just %hash = @list?) 21:55
Juerd Yes 21:56
Arrays called list are confusing 21:57
arnsholt Oh. Right
True. I tend to think of arrays as lists. It's a functionalism I suppose
sbp @cecinepasunearray 21:58
Juerd n'est pas
sbp danke
Juerd De rien :)
arnsholt Is a Hash the same as a List of Pairs?
(I'm guessing no) 21:59
diakopter sorear: hio
Juerd arnsholt: No, but it can flatten to one
22:00 masak joined
masak ahoy! 22:00
phenny masak: 17:42Z <moritz_> tell masak to please blog the challenge on his own blog
arnsholt But with non-deterministic order, right?
masak will do.
arnsholt Bah. I'll just do this file munging with a grammar. That's why we have them after all =D
Juerd arnsholt: Depends on how determined you are 22:01
22:02 xomas_ is now known as xomas, rgrau_ joined, azert0x left 22:03 rgrau_ left
Juerd arnsholt: The order will (iirc) be the same as the one for .keys and .values, but it can change with every mutation 22:03
masak lol i blogged: use.perl.org/~masak/journal/40385
jnthn lolmasakblogged! 22:07
masak heh; chromatic responded to ank's comment: use.perl.org/comments.pl?sid=44783&cid=72023
chromatic++
jnthn A very chromatic++ reply. :-) 22:08
22:08 justatheory left, justatheory joined
masak I'm glad to be thought to belong to the same stubborn conspiracy as such an eloquent and concise fellow. 22:10
22:14 skids left
jnthn finally has $scary-live-demo for dayjob together...hopefully. 22:15
Far too much that could possibly go wrong. :-)
masak "it's a million-to-one chance, but it might just work!" 22:17
22:18 dakkar left
masak arnsholt: a list of pairs is all you need to make a hash, and it can easily pretend to be one, but it isn't one. a list is ordered, a hash never is. 22:19
sorear masak: chromatic is quite human 22:20
arnsholt masak: Yeah, that's what I thought
masak sorear: certainly. but for a person being about as load as I am, he makes much better use of his voice and his choice of words. :) 22:21
diakopter load -> loud? 22:24
or loaded?
jnthn :-D 22:25
masak I rest my case.
yes, *loud
22:31 fridim joined 22:40 gbacon left
arnsholt Is making the ast with inlined code in the grammar supposed to work, or am I doing it wrong? 22:43
masak is not sure he understands the question 22:46
which grammar?
arnsholt I'll cook up an example. That should clearer 22:47
grammar Foo { token TOP { <foo> { $<foo>.ast.perl.say } }; token foo { foo { make "OH HAI" } } }; Foo.parse("foo"); 22:48
Er,
rakudo: grammar Foo { token TOP { <foo> { $<foo>.ast.perl.say } }; token foo { foo { make "OH HAI" } } }; Foo.parse("foo");
p6eval rakudo b258f5: OUTPUT«Any␤»
22:48 clintongormley left
arnsholt See what I'm trying to do? 22:49
jnthn Yeah - I suspect it should work. 22:50
masak aye.
same her. 22:51
s/her/here/
sorear arnsholt: doing anything involving $<foo> in inlined code is expected to be weird because $/ isn't defined during regex execution 22:52
there are some rules in S05 about it. incomprehensible ones.
pugssvn r31168 | cosimo++ | [docs] Try to make synopsis html docs less cluttered 22:53
r31168 |
r31168 | - Index page now hides lots of intro text by default
r31168 | - Improved default css a bit for synopsis documents
r31168 | - Smartlinks now right-aligned
r31168 | - Replaced small/full links with icons
r31168 | - Icon links are currently broken. Needs fix!
masak was just about to refer to S05, but chooses to consult it instead
cosimo++!
cosimo let's see if I broke something... :) 22:54
i guess the next sync will happen in an hour or something?
arnsholt sorear: Oh, right. Thanks for the clarification 22:55
cosimo would be cool to see if the docs are ok before going to sleep 22:57
i really need to
:)
22:59 jferrero left 23:00 skangas left
cosimo if there's breakage in the synopsis docs at perlcabal.org, you know it's me. I will take a look tomorrow morning and fix the fixable. 23:03
23:03 rgrau left 23:05 orafu left 23:07 orafu joined 23:12 skangas joined, lestrrat is now known as lest_away
lue ohai 23:16
masak \o 23:18
what's the semantics of $2 in a regex with fewer than three parentheses? 23:19
rakudo: say "foo" ~~ /f(.)$2 o/ 23:20
p6eval rakudo b258f5: OUTPUT«foo␤»
masak alpha: say "foo" ~~ /f(.)$2 o/
p6eval alpha 30e0ed: OUTPUT«Null PMC access in get_string()␤in Main (file <unknown>, line <unknown>)␤»
masak I have to confess I like alpha's reaction better in this case.
masak submits rakudobug
std: say "foo" ~~ /f(.)$2 o/
p6eval std 31168: OUTPUT«ok 00:01 109m␤»
jnthn Great, we spend all our time removing null PMC accesses...
:P
masak :) 23:21
jnthn masak: I guess what you're saying is you want an error?
masak aye.
jnthn Or at least a warning.
It's one of those things we should be able to spot and barf over at compile time though, I guess.
masak since this is something that we can statically check for...
right. 23:22
arnsholt Are strings still the only valid type for keys?
masak in Rakudo? yes.
arnsholt That's what I thought. Hopefully it explains why my code refuses to work =)
lue anyone catch wind of Apple's new HTML5 demo? :) 23:23
sorear yes
masak Apple is the company that makes those Mac computers, right?
jnthn masak: iThink so.
masak I like those. they look nice.
jnthn likes his ThinkPad, but more for its robustness than its aesthetics. 23:24
lue Here's a tip: don't use Safari :)
It's a joke.
masak one I don't get. 23:25
rakudo: say "foo" ~~ / f $0 (.) /
p6eval rakudo b258f5: OUTPUT«fo␤»
jnthn Getting stuff to display decently on Safari is a lot less bother than it is in a certain other browser.
masak: heh
sorear you mean IE4? 23:26
masak jnthn: that's why it must be wrong.
23:26 bjarneh joined
jnthn sorear: IE6 ain't much fun. 7 and up is, admittedly, better. 23:26
23:26 ashleydev_ left
jnthn sorear: Last time I was doing any HTML layout complex enough to be non-trivial, IE6 was still big enough in terms of user base to have to care about. 23:27
sorear: I do look after one legacy codebase at $dayjob that has a few references to workarounds for IE4 vs NS4 though. :-)
23:27 meppl left
lue jnthn: it still is, amazingly 23:27
23:27 ashleydev joined
Juerd IE6 userbase is still way too big. :( 23:27
arnsholt Whee! Halfway to a trained bigram HMM =D
lue That's why I've vowed never to actively take time to support IE. Takes too much time. 23:28
Juerd I'm enjoying doing html5 and jquery but IE6 really makes me angry.
jnthn jquery is about the only thing I find fun in client side web dev these days.
sorear problem: making coordinated changes to viv and STD is a real bother 23:29
Juerd I like two tiny things about html 5 that can still make me smile
1. I can actually remember the doctype
lue Amen to 1. 23:30
Juerd 2. <meta charset="..."> is so much nicer.
(than http-equiv garbage)
lue For ye who have not seen: www.apple.com/html5/ 23:34
Tene looks like something like 7%, and there's the question of what correlates with that 7%, such that they would or wouldn't visit your site. 23:35
lue And remember: only Safari supports HTML5. No other browser supports even a single tag </sarcasm>
lue wants to hack into their scripts and get pass the Safari Only code. 23:36
*past
rakudo: multi sub circumfix:<A T>($a) {$a;}; multi sub circumfix:<C G>($a) {$a;}; say A "hi" T; say C "hi" 23:37
sorear I've seen blog posts from people who've dissected the source of the demo
p6eval rakudo b258f5: OUTPUT«error:imcc:syntax error, unexpected '\n'␤ in file 'EVAL_1' line 36574794␤===SORRY!===␤syntax error ... somewhere␤»
sorear it's actually quite legitimate that they only allow Safari, because what it's actually demonstrating is HTML5's vendor namespace feature
23:38 jhuni joined
lue It doesn't matter. You don't say "Look at our support for HTML5" and then make it safari-only 23:38
I ought to tell them about their false advertisment 23:39
i.e. "...Not all browsers offer this support...."
jnthn lue: Didn't do custom circumfixes just yet...but that error message is crap.
arnsholt Heh. Adding parametrised type constraints to my code caused Rakudo to segfault
Whoops
lue If you notice, that was base pairings, and I wanted a mismatch or missing "G" error :) 23:40
jnthn arnsholt: ugh.
masak lue: you could try the same trick with strings or arrays instead, perhaps. 23:41
23:41 cdarroch left
lue rakudo: say /<digit>+/.WHAT 23:41
arnsholt jnthn: Yeah. I'm trying to extricate the bug from my code
p6eval rakudo b258f5: OUTPUT«Regex()␤»
dalek kudo: b497776 | (Solomon Foster)++ | src/core/Int.pm:
Change infix:<div> to return floor($a / $b) as in the spec. (Same as before for
23:42
lue rakudo: say Regex.^methods
p6eval rakudo b258f5:
..OUTPUT«ACCEPTSBoolwrapassumingunwrapACCEPTScountaritynewmultinameclonesignaturedoperlreturnsofassumingcallwithStrsecasectrimcotanhtrim-leadingcislogsechlog10randtruncatesprintfbytessqrtsinasincoshsuccsignasinhacosechabsucceilingunpolarfloorsamecasesubstrasechordroundflipsplitmatchacoshaco…
lue O.o
arnsholt rakudo: multi sub foo(Array of Str @foo) {}; foo(<helo>); 23:43
p6eval rakudo b258f5: OUTPUT«_block39␤»
arnsholt Oh, wait. I'm not up to date
jnthn Array of Str probably won't work too well ATM
Also though, Array of Str means every element in @foo would be an Array, if it did work. 23:44
arnsholt Point
(In this case I even want Array of Str @arr)
Even if I wrote it Array of Array of Str @arr before you pointed that out
But type parameters are parsed but not checked, no? 23:45
jnthn arnsholt: They are checked, AFAIK. 23:46
In most places
arnsholt Oh, goodie. Last time I poked that they weren't 23:47
jnthn rakudo: my Int @foo; my Str @bar; sub x(Int @foo) { }; x(@foo); x(@bar);
p6eval rakudo b258f5: OUTPUT«Nominal type check failed for parameter '@foo'; expected Positional[Int] but got instead␤ in 'x' at line 11:/tmp/SRISyWz6jV␤ in main program body at line 11:/tmp/SRISyWz6jV␤»
jnthn but got...FAIL. 23:48
arnsholt Heh
lue rakudo: say Regex.new("<digit>+")
p6eval rakudo b258f5: OUTPUT«too few positional arguments: 2 passed, 3 (or more) expected␤ in main program body at line 1␤»
lue rakudo: say Regex.new("<digit>+").perl
p6eval rakudo b258f5: OUTPUT«too few positional arguments: 2 passed, 3 (or more) expected␤ in main program body at line 1␤»
lue :(
masak g'night, #perl6! 23:49
jnthn o/
lue goodnight
jnthn lue: say regex { <digit>+ }
23:49 masak left
jnthn well or (regex { ... }).perl 23:49
lue rakudo: my $a = "<digit>+"; say regex { $a } # likely fail 23:50
p6eval rakudo b258f5: OUTPUT«_block49␤»
lue rakudo: my $a = "<digit>+"; say (regex { $a }).perl # likely fail
p6eval rakudo b258f5: OUTPUT«{ ... }␤»
lue rakudo: my $a = "<digit>+"; my $b = regex { $a } ; say "1337 h4x0r" ~~ $b
p6eval rakudo b258f5: OUTPUT«␤» 23:51
lue hrm..
rakudo: my $a = "<digit>+"; my $b = regex { $a } ; say '1337 h4x0r $ays hi' ~~ $b 23:53
p6eval rakudo b258f5: OUTPUT«␤»
lue I mustn't comprehend regexes :)
sorear lue: you appear to be under a critical misunderstanding of what variables in regexes do
23:53 tedv joined
jnthn lue: I think you want <$a> 23:53
lue obviously :)
rakudo: my $a = "<digit>+"; my $b = regex { <$a> } ; say '1337 h4x0r $ays hi' ~~ $b 23:54
p6eval rakudo b258f5: OUTPUT«1337␤»
jnthn <$a> treats $a as regex code
$a matches it literally
lue And Now....
sorear reread the section starting at S05:0991
lue rakudo: multi sub circumfix<☃ ☄> ($a) { regex { <$a> };}; say '1337 h4x0r' ~~ ☃"<digit>+"☄ 23:55
p6eval rakudo b258f5: OUTPUT«===SORRY!===␤Malformed multi at line 11, near "sub circum"␤»
lue so. close.
sorear you mean circumfix:< 23:56
lue rakudo: multi sub circumfix:<☃ ☄> ($a) { regex { <$a> };}; say '1337 h4x0r' ~~ ☃"<digit>+"☄
p6eval rakudo b258f5: OUTPUT«error:imcc:syntax error, unexpected '\n'␤ in file 'EVAL_1' line -2102517350␤===SORRY!===␤syntax error ... somewhere␤»
lue dankon sorear, that was it. 23:57
Now it's just ye olde circumfix bug.
jnthn lue: Rakudo doesn't know how to handle those just yet.
It's a medium difficulty patch.
lue
.oO(It's the one thing keeping ☃☄ from being reborn)