»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! | tinyurl.com/p6contest
Set by moritz_ on 28 December 2010.
00:07 felliott joined 00:09 Kodi left 00:31 kst left, kst joined 00:38 kfo_ joined 00:41 kfo left 00:43 molaf__ joined 00:46 molaf_ left 00:48 cdarroch left 00:57 QinGW joined 01:02 kst left 01:03 kst joined 01:18 felliott left 01:24 felliott joined 01:31 starcoder left 01:32 felliott left 01:38 starcoder joined 01:42 arlinius joined 01:50 dukeleto left, dukeleto joined 01:54 _jaldhar joined 01:55 drbean joined, kst left 01:57 kst joined 02:07 whiteknight left
flussence I was poking around the rakudo source and found "as" used inside a signature, what does that mean? The only place in the spec I can find mention of it is for return value types. 02:08
(and it's not the most greppable of keywords...)
Tene flussence: coerced to type 02:09
rakudo: sub foo($n as Str) { say $n.WHAT }; my $x = 1; say $x.WHAT; foo($x); 02:10
p6eval rakudo 723986: OUTPUT«Int()␤Str()␤»
flussence oh... that's useful to know :)
Tene flussence: like that: ^
colomon woah, I didn't know that was implemented yet. 02:14
Tene colomon: Neither did I. ;)
colomon rakudo: sub foo($n as Str) { say $n.WHAT; say :$n.perl; }; my $x = 1; say $x.WHAT; foo($x); 02:15
p6eval rakudo 723986: OUTPUT«Int()␤Str()␤"n" => "1"␤»
colomon rakudo: sub foo($n as Rat) { say $n.WHAT; say :$n.perl; }; my $x = 1; say $x.WHAT; foo($x);
p6eval rakudo 723986: OUTPUT«Int()␤Rat()␤"n" => 1/1␤»
colomon rakudo: sub foo($n as Complex) { say $n.WHAT; say :$n.perl; }; my $x = 1; say $x.WHAT; foo($x); 02:16
p6eval rakudo 723986: OUTPUT«Int()␤Complex()␤"n" => Complex.new(1, 0)␤»
colomon dang, it's good.
flussence rakudo: sub foo($n as IO) { say :$n.perl }; my Int $x = 1; foo($x); # wondering how it breaks 02:18
p6eval rakudo 723986: OUTPUT«"n" => IO.new(PIO => Any, ins => Any, autoflush => Any, path => "1", stat => IO;Stat.new(path => "1"))␤»
flussence oh, probably won't do that with builtin types.
rakudo: class A {}; sub foo($n as A) { say :$n.perl }; my Int $x = 1; foo($x); # wondering how it breaks
p6eval rakudo 723986: OUTPUT«Unable to coerce value for '$n' from Int to A; no coercion method defined␤ in 'foo' at line 1:/tmp/gcRL90y4nT␤ in main program body at line 22:/tmp/gcRL90y4nT␤»
flussence yey
s/e/* 02:19
rakudo: sub foo($x as Int where { $_ > 0 }) { say $x }; foo("0x10") 02:20
p6eval rakudo 723986: OUTPUT«16␤»
flussence ooh, I am *so* going to use that. 02:21
02:23 leprevost left
flussence Ick. (-1).sqrt == NaN and (i ** 2) is -1 + (float rounding error)i. 02:34
Tene rakudo: say i * i 02:42
p6eval rakudo 723986: OUTPUT«-1 + 0i␤»
[Coke] rakudo: say pi * e 02:44
p6eval rakudo 723986: OUTPUT«8.53973422267357␤»
[Coke] rakudo: say pi e
p6eval rakudo 723986: OUTPUT«===SORRY!===␤Confused at line 22, near "say pi e"␤»
02:47 cafesofie joined 02:53 kst left, kst joined 02:55 felliott joined 03:00 jferrero left 03:29 am0c joined 04:03 agentzh joined 04:07 mtk left 04:16 mtk joined 04:25 cotto left 04:27 Su-Shee left 04:28 Su-Shee joined 04:32 GinoMan joined 04:46 Chillance left 05:04 envi joined, felliott left 05:19 coldhead left 05:20 stifynsemons joined, stifyn joined 05:21 zevpl left 05:31 jfried left, jfried joined 05:37 cotto joined, tomaw left, jfried left 05:40 stifyn left, stifynsemons left 05:43 Patterner left 05:46 Psyche^ joined, Psyche^ is now known as Patterner 05:49 woosley joined 05:50 JimmyZ joined 05:52 jfried joined 05:54 JimmyZ left
dalek ecza: f1f77ce | sorear++ | / (2 files):
Implement .method automatic $_ form
 05:54
06:12 kaare_ joined 06:18 cjk101010 joined 06:19 f00li5h left 06:31 mberends left 06:33 dju left 06:40 orafu left, orafu joined 06:45 dju joined, dju left, dju joined 06:47 ponbiki left, ponbiki joined 06:59 wtw joined 07:05 drbean left, ravic joined 07:07 tomaw joined
ravic i got a perl qquestion. and i m unable to login into #perl 07:07
m ashamed to ask it here as its nt perl6. can any help me dodging into #perl
perigrin why are you unable to login to #perl? 07:08
moritz_ ravic: there's also irc.perl.org with lots of channels
perigrin including a #perl-help
ravic on it. 07:20
'how to copy a file to all 10 remote machines simulataneoulsy?' -> need somehelp in this context
sorear you don't need Perl for that 07:21
man bash_builtins
look up "for"
man scp
moritz_ or search for ntfs :-)
erm, nfs 07:22
moritz_ not quite awake
sorear -> sleep
plobsing moritz_: if you could do that using only ntfs, I'd be impressed 07:26
07:30 cosimo left 07:31 cosimo joined, PZt left
moritz_ plobsing: not only nfs, but nfs for distributing the file 07:34
ravic nfs doesnt work out for me. architechure limitations 07:37
yes scp can push file
rsync as well
but u need a way to push it parllelly 07:38
plobsing ravic: have you heard of job control? your shell should have it. 07:41
07:50 coldhead joined
Tene ravic: man parallel 07:52
ravic: man bash, /^JOB 07:53
ravic: man xargs, /-P 07:54
08:18 kst left, kst joined 08:23 cafesofie left
TiMBuS is there a way to unload or overwrite a module in perl6? 08:24
jnthn morning, #perl6 08:25
TiMBuS: You could use a different version of it in a nested lexical scope, maybe. Dunno how far that helps. Guess if you only use it within a given scope then it's "unloaded" outside that scope. But that's not unloaded as such I guess, only not in scope... 08:26
TiMBuS that could eventually get a bit bulky.. 08:27
08:28 bacek left
jnthn TiMBuS: What's the use case? 08:31
08:32 bacek joined
TiMBuS jnthn, irc bot, modules are used for callbacks 08:33
so being able to load/unload features would be mighty handy
08:36 Mowah joined
TiMBuS or, if you need a more exact usage example: require 'the/module'; modulelist.push(the::module.new) 08:40
i guess hiding the name works 08:41
08:41 PZt joined 08:47 daxim joined, Mowah left 08:48 masak joined 08:49 kensanata joined, drbean joined
masak morning, zebras. 08:49
daxim now that sixers have a sizable toolchain, how about applying for the QA hackathon? interested/does something need fixing with facetime? www.nntp.perl.org/group/perl.qa/201...12675.html 08:51
masak daxim: that's a splendid idea. daxim++ 08:52
daxim: especially since we have lots of tricks to learn from the Perl 5 people who have gone before us.
rakudo: my $foo = [5]; for $foo { .shift }; say +$foo 09:03
p6eval rakudo 723986: OUTPUT«0␤»
masak \o/
09:11 ravic left 09:27 REPLeffect left 09:29 REPLeffect joined 09:35 drbean left 09:36 timbunce joined 09:47 noganex_ is now known as noganex
masak twitter.com/supernovus/status/29666617625419776 -- I know the feeling. I've solved this by searching out problems where CPU isn't the bottleneck. 09:47
09:56 timbunce left
jnthn is speed away from using Rakudo for more stuff 09:58
masak ...which is why Niecza is interesting to me. it seems to run faster, which can be a plus even with fewer features than Rakudo. 09:59
jnthn *nod* 10:00
10:02 QinGW left
jnthn The Rakudo 6model branch will be made in a week or so. 10:02
10:02 timbunce joined
masak that's excellent news. ++jnthn! 10:09
jnthn We'll also get the multi semantics updated in with that lot. :) 10:11
masak \o/
finally!
jnthn nqp-rx/nom supports them already ;)
masak yes, I've tried them. :) 10:12
jnthn ooh :) 10:13
10:13 tzhs joined
jnthn It's still dead slow in nqp-rx though. 10:13
I didn't port the multi cache from /nom
er
to /nom
from nqpclr :)
10:14 dakkar joined
masak jnthn: see, even you are confused by the naming! :P 10:14
jnthn :P 10:15
jnthn looks forward to pmichaud++'s forthcoming post to clear it all up :)
10:17 justatheory left 10:20 cognominal joined
masak jnthn: www.angryduck.com/pictures/2010_11/..._Drunk.jpg :P 10:23
jnthn That fish looks like it's fin-ished plenty of beer. :P 10:27
masak :P 10:35
10:43 f00li5h joined, f00li5h left, f00li5h joined
daxim www.oreillymaker.com/ # have fun 10:46
masak :)
daxim old idea, though, I think I used this one the last time: web.archive.org/web/20020124124713/...oracovers/ 10:49
mathw jnthn: I read your blog post about 6model with great interest.
jnthn daxim: Oh, cool :) 10:50
mathw: Glad it was interesting. :) 10:51
jnthn tries not to swear too loudly at TFS 10:52
mathw It was also rather exciting
does prove the value of the dispatch cache doesn't it :) 10:53
jnthn mathw: Very much so
mathw: It was exacberbated by the fact that I hadn't done anything much to optimize the non-dispatch-cache case.
mathw mmm 10:54
10:54 dukeleto left
jnthn (e.g. it shoudln't have been quite that much slower anyway) 10:54
mathw well hopefully there'll be some wins there in the future :)
jnthn (but I'd not prioritized that bit yet as I expected to be hitting the cache lots :))
10:54 dukeleto joined
mathw well, what's the rule? make it work, then make it fast 10:57
hopefully :)
with a load of caveats about trying to avoid implementations which are impossible to make faster
jnthn *nod* 11:02
11:03 dju left 11:04 dju joined 11:10 coldhead left
mathw jnthn: you're learning Russian, IIRC? 11:12
jnthn Trying to :)
mathw I'm trying to learn how to pronounce it
I joined a choir, and the first piece we're doing has four movements in Russian (and two in Latin) 11:13
It's... got some interesting sounds
dukeleto ipv6 just landed in Parrot 11:14
jnthn mathw: :)
And all the palleted things
jnthn bbs - lunch 11:15
arnsholt And lots of weird sibilants
mathw mmm
masak mathw: I agree, singing in Russian is fun.
mathw so of course when we have a whole load of consonants that don't exist in English
Prokofiev puts them on a load of very fast notes
arnsholt There are some Polish placenames where I acknowledge only the German versions (because I can pronounce the real Polish names =) 11:16
masak mathw: live with it, anglophone. :P
mathw pfft
I can handle some of them better than my immediate neighbours in the tenor section
because they're in Lojban :)
masak :) 11:17
mathw and I have no idea what the man next to me thought he was singing last night
but it wasn't what we've been taught
masak mathw: consider how well-suited the Cyrillic orthography is to the Russian phonology.
mathw certainly seems to fit it better than English is served by Latin alphabet 11:18
I'm trying to wean myself off the transliteration, because that's really misleading
arnsholt The problem with English isn't as much the alphabet as the orthography 11:19
mathw I'm also trying to avoid the temptation to sing the German translation, which is also printed in my edition of the score
it's all fun though
arnsholt Essentially, the Great Vowel Shift and the codification of written English occured in the wrong order =)
mathw heh yes
we also didn't mangle our loanwords enough
arnsholt (Unfortunately for anyone born after Shakespeare) 11:20
Yeah, that's a pain as well
But less so for me, since I know both French and Latin =)
masak mathw: no, you definitely shouldn't pay attention to the transliteration. :/
11:21 woosley left
mathw Ultimately it's going to be better to learn how to pronounce what's written in Russian 11:22
the transliteration takes too long to interpret at the speed this thing goes
At least I have no existing assumptions about how to pronounce things written in Cyrillic
Rather than having to remember what the transliterator meant when he wrote "l(y)oo"
masak :( 11:23
noganex reminds me somehow of the time when i learned thai...
gfldex i was teached russian in school and i can tell you that english transliteration is a joke
masak thai doesn't even have a standardized transliteration...
joke or not.
mathw gfldex: we figured that out :)
11:24 jpike left
mathw we hired a Russian-speaker to help with the learning 11:24
noganex masak: i remember driving in the wrong direction for ~40 miles because i misread a sign ;)
mathw his pronunciation is nothing like you'd think from the transliteration, so it's really not very helpful at all
11:24 woosley1 joined
noganex and thai has a transliteration. 11:25
masak oh?
gfldex around my hometown in eastern germany there are quite a few town with slavic names left 11:27
that helped a lot :)
Tschopau, Tschochern and the like
11:27 masak left
arnsholt Thai script is Brahmic, so it should be transcribable with something similar to IAST I'd think 11:28
11:29 drbean joined
daxim all hail ISO-9 11:30
11:35 agentzh left 11:42 tzhs left 11:43 masak joined
jnthn mathw: Yes, learning the Cyrillic is probably a better way to go, if you have the time to do so. 12:05
sECuRE hm, i’m currently trying to use my patch from #77992 on the current rakudo. i get a NULL PMC access in set_integer_native() in some of my code (the line numbers don’t seem to be precise). how can i debug this further? 12:25
masak sECuRE: a very good question. 12:31
sECuRE: what I end up doing is trying to get the failing code down to its bare minimum.
sECuRE :-\ 12:32
flussence
.oO( I like how Vim just magically makes sense sometimes - transposing 2 characters is "xp" )
masak flussence: right, and two lines is "ddp". 12:33
flussence it's the way 2 independent actions spell out a mnemonic for the thing putting them together does :) 12:34
masak oh, "x" can mean "trans-". cute. :) 12:39
daxim xh, "o" can means "cis-". grave. :( 12:44
moritz_ sECuRE: start rakudo with the --ll-backtrace option
(needs very latest parrot)
sECuRE moritz_: alright, will try
12:46 bluescreen joined
moritz_ it gives you a PIR level backtrace 12:46
which is more fine grade than the normal ones 12:47
sECuRE p.nnev.de/1700 < this is what i get
moritz_ do you have a quick link to the getaddrinfo code in IO::Socket::INET 12:48
sECuRE rt.perl.org/rt3/Ticket/Attachment/7...do-A.patch 12:49
moritz_ sECuRE: must be an assignment of the form $P0 = $I0 12:50
sECuRE hm, ok 12:51
moritz_ where the RHS can be anything that returns an int
sECuRE ok
and you’re sure it’s $P0, not maybe $P1? 12:52
moritz_ not at all
can be any PMC container on the left
sECuRE to be clear: are $P0 etc. only for PMCs? 12:53
moritz_ if you write $P0 = someInteger, theset_integer_native vtable is called 12:54
on the PMC that's previously in $P0
sECuRE uhm, oh. i misunderstood the concept, then, i think
moritz_ s/theset/the set/
and set_string_native for strings etc.
12:55 accel joined, accel left
takadonet morning all 12:55
flussence I'm not convinced a dual-stack-supporting thing should be called INET, especially after seeing this - www.reddit.com/r/perl/comments/f7f3...6_in_perl/
12:55 accel joined
masak takadonet: \o 12:55
sECuRE flussence: yeah, but let’s focus on getting ipv6 support merged at all, right? ;)
flussence sECuRE: of course :) 12:56
moritz_ #perl6 is always a funny mixture of "how can I achieve $x now?" and "Can't we do much better in future, somehow?" 12:58
daxim duct-tape and bugs 12:59
sECuRE moritz_: have a look at my method getaddrinfo() please. i need to call sock.'getaddrinfo'(…) with the arguments provided. how would i use find_lex correctly?
moritz_ sECuRE: Perl 6 level functions always need PMCs as arguments 13:01
13:01 tzhs joined
moritz_ s/functions/routines/ 13:01
sECuRE: it would be much easier to ony create sock in Q:PIR, return it from the PIR block, and then do the actual call in Perl 6 code 13:02
my $sock = Q:PIR { %r = root_new ['parrot';'Socket'] }
( plus newlines, I think)
and then if pir::isnull($sock) { raise error } else { $sock.getaddrinfo($hostname, $port, ...) } 13:03
or something along these lines
hm, maybe the null check needs to be inside Q:PIR too
but not the whole call
13:04 y3llow left 13:05 y3llow joined
sECuRE moritz_: ok, i’ll try 13:07
13:12 plobsing left
sECuRE compiling core.pir takes ages :\ 13:13
masak yes. :/ 13:14
it's likely to get better with jnthn++'s improvements.
13:17 MayDaniel joined 13:20 felliott joined
sECuRE moritz_: i now reduced it to the following code: p.nnev.de/1703 13:22
moritz_: new backtrace here: p.nnev.de/1704
13:23 bluescreen left, bluescreen joined 13:26 stifyn joined, stifynsemons joined
sECuRE moritz_: hm, now i changed it to p.nnev.de/1705 and get p.nnev.de/1706 13:30
bluescreen }
13:33 y3llow left, y3llow joined
moritz_ sECuRE: sorry, just had a bit local stress, will try to take a look 13:39
sECuRE no worries
just trying to compile a minimal %r = root_new …, like you suggested
moritz_ sECuRE: sholdn't you have %r = sock somewhere? 13:40
sECuRE ok, it survives the call now and errors in the open() itself. probably a similar error there
yeah, that was what i changed just now ;)
13:45 accel left
moritz_ unless sock goto ERR # maybe should be "unless null sock goto ERR" 13:45
13:45 plobsing joined 13:48 plainhao joined 13:51 mtk left
mathw jnthn: concert's on the 16th of March... I may have managed to memorise the words by then, which would be far more helpful. 13:51
there aren't, in the grand scheme of things, very many of them
jnthn mathw: :) 13:52
mathw: Sounds like fun. :)
jnthn can't sing at all
:(
mathw it is fun :)
stifyn what are you singing? 13:53
mathw the fourth movement is "Arise, ye Russian People!" and encourages all of Russia to rise up and hack the invading enemies into tiny pieces
stifyn: Prokofiev, Alexander Nevsky
jnthn Awesome!
mathw The only thing I don't like about it is the orchestra's part, which I think is a bit detracting from the singing 13:54
jnthn: in all of Russia no foe shall live! We shall beat them to death! etc.
jnthn How many references to Vodka does it have? :)
mathw some of it sung very sweetly and angelic
stifyn sounds like not as many as to death :-)
mathw jnthn: none, according to our semi-literal translation
at the end it mentions a festival
but that's only after all the beating to death has been accomplished 13:55
jnthn Well, of course.
Need to celebrate such things. :)
mathw well they win
that's the story, anyway
jnthn \o/
mathw and even let some of the prisoners go home alive
jnthn Wow, that's generous. :)
mathw mmm 13:56
especially after all the stuff about how no foe shall live
sECuRE hm, when i do both calls in perl, i do get an UnManagedStruct by getaddrinfo() (that’s correct), but when i pass it to connect(), the code path which checks for address->vtable->base_type == enum_class_UnManagedStruct does not get run
jnthn Do you put it into a scalar? 13:58
sECuRE yes
jnthn If so, you're likely ending up passing the container containing the UnmanagedStruct.
sECuRE how would i directly pass it?
jnthn Look up descalarref
As in, search for usage of it in the codebase
13:58 mtk joined
jnthn May be some in IO.pm 13:58
sECuRE yep 13:59
yes, indeed, that helps 14:00
thanks
yay, i can get heise.de using IPv6 \o/
(in rakudo)
moritz_ \o/ 14:03
submit the patch, I'll apply it :-)
sECuRE cool. let me clean up stuff first and add some error handling 14:04
btw: the necessary rakudo changes are also merged already :)
moritz_ you mean parrot changes?
sECuRE err, yes
moritz_ yes, I follow the parrot-dev mailing list 14:05
and #parrot too :-)
sECuRE alright
hm, in the places where root_new is used i don’t see checks if it’s null or not. is there maybe a reason that root_new can never return null?
moritz_ just try it with a random name 14:06
I guess it'll thrown an error instead of returning new
14:06 drbean left
sECuRE yes, indeed 14:06
14:07 timbunce left 14:08 kaare_ left 14:10 GinoMan_ joined 14:11 GinoMan left, GinoMan_ left 14:21 skangas_ joined 14:22 skangas left, skangas_ is now known as skangas 14:23 cognominal left, nrr left, am0c left, cognominal joined 14:24 am0c joined 14:25 nrr joined, barika left 14:26 barika joined 14:27 cognominal left 14:29 ggoebel left 14:31 cafesofie joined, cognominal joined 14:34 ddima joined
ddima hi 14:34
14:35 ggoebel joined
takadonet ddima: hey 14:37
14:48 felliott left 14:51 kaare_ joined, kst` joined
sECuRE moritz_: done now. you can find my patch at t.zekjur.net/0001-Adapt-IO-Socket-I...do-A.patch 14:52
14:52 dukeleto left 14:53 cjk101010 left, kst left, rbuels left, dukeleto joined, bluescreen left 14:54 dukeleto left, dukeleto joined 14:55 bluescreen joined 14:57 stifyn left, stifynsemons left, bluescreen left 15:00 ggoebel left 15:02 Mowah joined 15:04 mtk left, mtk joined, ggoebel joined 15:07 kst` left 15:08 kst` joined, timbunce joined 15:12 spq1 joined 15:15 timbunce left, IllvilJa left 15:16 IllvilJa joined 15:19 wtw left 15:25 Psyche^ joined 15:26 Patterner left, Psyche^ is now known as Patterner 15:28 plobsing left 15:30 stifyn joined, stifynsemons joined 15:37 plobsing joined 15:47 justatheory joined 15:48 timbunce joined 16:01 MayDaniel left 16:08 icwiener joined
pmichaud good morning, #perl6 16:08
masak morning, pmichaud.
16:09 icwiener left, icwiener joined
pmichaud lots of joins/quits this morning.... do I feel a tremor in the intertubes? 16:10
masak could be the IPcalypse starting early. 16:11
16:14 uniejo joined
jnthn o/ pmichaud 16:14
16:14 icwiener_ joined
colomon \o 16:16
16:17 icwiener left 16:28 kst` left, icwiener joined 16:29 kst` joined 16:31 timbunce left 16:32 icwiener_ left, stifyn left, stifynsemons left 16:33 am0c left
jnthn decommute & 16:34
16:35 tzhs left, uniejo left 16:37 JimmyZ joined, timbunce joined
TimToady /e 16:39
16:43 icwiener_ joined, icwiener_ left, icwiener_ joined 16:46 icwiener left, pmurias joined
pmurias what's the difference between contextual variables and Perl 5 local style ones? 16:47
pmichaud "contextual variables" are now called "dynamic variables", iirc.
pmurias contextual variables are not shared between threads
16:47 uniejo joined
masak dynamic variables can be, if they're read-only. 16:48
pmichaud anyway, dynamic variables are located by looking at lexicals in the dynamic caller frames. I don't think there's any close equivalent in p5.
TimToady dynamic variables can be shared as long as there's a clear pecking order 16:49
even if rw
but yes, sharing ro variables is pretty easy :)
pmurias are dynamic attributes supported? 16:50
TimToady p5 locals are really just globals
masak 'local' variables in p5 are really just package gl... right.
TimToady we could have p5-style attributes if we ever implement temp
pmurias i'm reading about "Contextual Programming" and i'm trying to determin what the real difference between a dynamic variable and just setting and resetting a normal variable is 16:51
masak there is at least a little code smell inherent in p6 dynamic variables, since these are not namespaced at all. it'd be nice to see a solution for namespacing dynamic variables.
pmurias: url?
TimToady dynamic var is a lexical, not a global
16:52 cafesofie left
pmurias common-lisp.net/project/closer/contextl.html 16:53
masak thank you.
16:53 wooden left
pmurias i have to make a presentation about that tommorrow 16:54
TimToady masak: I don't think dynamic variables are useful enough to warrant the baggage of namespacing
dynamic hashes and common prefixes are probably good enough
masak TimToady: maybe.
TimToady: but I'd be wary every time I used both dynamicals and callbacks. 16:55
TimToady esp if we allow :: in a dynamic name :)
masak :D
then I'm fine :)
TimToady we can think about that
might also be useful to target the final default to a package other than GLOBAL 16:56
which makes it slightly less smelly, I suspect
std: my $*FOO::BAR = 2; 16:57
p6eval std 625303c: OUTPUT«ok 00:01 121m␤»
TimToady heh
flussence takadonet++ # looks like Text::Tabs is going to be helpful after all
masak rakudo: my $*FOO::BAR = 42; say $*FOO::BAR 16:58
p6eval rakudo 723986: OUTPUT«Cannot modify readonly value␤ in '&infix:<=>' at line 1␤ in main program body at line 22:/tmp/2PoIAPndlk␤»
16:58 envi left
masak submits rakudobug 16:58
TimToady not clear that it parsed it correctly though
s/it/std/
masak should it error out on FOO:: not being predeclared? 16:59
16:59 icwiener joined 17:00 JimmyZ left
TimToady there's definitely something wonky in the STD parse of that 17:00
pmurias where are dynamic variables specced?
TimToady nevertheless it can probably be made to work someday 17:01
17:01 uniejo left
TimToady package globals don't generally care about whether FOO:: exists yet for normal variables 17:01
masak TimToady: well, Rakudo parsed it too, since it got to runtime.
TimToady std: $FOO::bar = 42; 17:02
p6eval std 625303c: OUTPUT«ok 00:01 120m␤»
masak ok.
pmurias rakudo: my $*foo = 123;sub foo {my $*foo = 145;bar()};sub bar {say $*foo};foo()
p6eval rakudo 723986: OUTPUT«145␤»
TimToady but the tree from viv looks very strange to me
17:02 icwiener_ left
pmurias rakudo: my $*foo = 123;sub foo {my $*foo = 145;bar()};sub bar {say $*foo};foo();say $*foo; 17:02
p6eval rakudo 723986: OUTPUT«145␤123␤»
pmurias rakudo: {my $*foo = 123;sub foo {my $*foo = 145;bar()};}sub bar {say $*foo};foo();say $*foo;
p6eval rakudo 723986: OUTPUT«===SORRY!===␤Confused at line 22, near "{my $*foo "␤»
pmurias rakudo: {my $*foo = 123;sub foo {my $*foo = 145;bar()}};sub bar {say $*foo};foo();say $*foo; 17:03
p6eval rakudo 723986: OUTPUT«Could not find sub &foo␤ in main program body at line 22:/tmp/DKyZ45XZP1␤»
TimToady I see two ways to implement it
pmurias rakudo: sub bar {say $*foo};{my $*foo = 123;sub foo {my $*foo = 145;bar()};foo()}
p6eval rakudo 723986: OUTPUT«145␤»
pmurias is the last example correct?
TimToady one is by forcing $*FOO::BAR to parse "FOO::BAR" as a string
masak right. 17:04
TimToady the other is to turn $*FOO:: into a shadow package
masak hm.
TimToady I think viv is parsing consistent with the latter interpretation
17:04 icwiener_ joined
TimToady if we had shadowed hashes, it might just fall out of that 17:04
masak what would be the advantage of the latter approach?
shadowed hashes? 17:05
TimToady if it's not in this hash, fail over to that hash
17:05 icwiener left
TimToady so you get a chain of %*FOO:: symbol tables out to the FOO one 17:05
masak rakudo: module Foo {}; my $*Foo::bar = 42; say $*Foo::bar 17:06
p6eval rakudo 723986: OUTPUT«Cannot modify readonly value␤ in '&infix:<=>' at line 1␤ in main program body at line 22:/tmp/jTBUjIfID7␤»
TimToady one problem is that package symbol tables have a slightly different sigil policy, iirc
17:07 plobsing left
TimToady pmurias: yes, it's correct 17:07
17:08 hudnix joined 17:09 hudnix left, hudnix joined
TimToady I guess the basic problem is that my %*FOO doesn't know that it's supposed to shadow its ancestor, but maybe we can force it in the :: case 17:09
17:09 icwiener_ left 17:10 icwiener joined, felliott joined
TimToady er, $*FOO does lookup in %*FOO:: I mean, and that's what doesn't necessarily know to shadow %CALLER::FOO 17:10
%CALLER::FOO:: rather
CALLER<%*FOO::> is perhaps clearer 17:11
masak aye.
dalek kudo: 86f61b9 | (Michael Stapelberg)++ | src/core/IO/Socket/INET.pm:
Adapt IO::Socket::INET to the IPv6-supporting rakudo API

Signed-off-by: Moritz Lenz [email@hidden.address]		 17:12
masak moritz_++
flussence yay
moritz_ sECuRE++ # he did all the work 17:13
masak \o/
TimToady anyway, probably no for 6.0.0.0.0...
*t
masak agreed.
pmurias so a dynamic var is a lexical which also looks in the callers scopes if it can't find the var? 17:14
17:14 kensanata left
TimToady something like that 17:14
it's a lexical that is declared with a * twigil as part of its name, and mentioning the * twigil in an rvalue does the lookup for the funny name up the dynamic scope 17:15
masak rakudo: my $a = 1; say $a; my $a = 2; say $a
p6eval rakudo 723986: OUTPUT«===SORRY!===␤Redeclaration of symbol $a at line 22, near " = 2; say "␤»
masak rakudo: my $*a = 1; say $*a; my $*a = 2; say $*a
p6eval rakudo 723986: OUTPUT«===SORRY!===␤Redeclaration of symbol $*a at line 22, near " = 2; say "␤»
pmurias rakudo: my $*a =1;say $a 17:16
masak std: my $*a = 1; say $*a; my $*a = 2; say $*a
p6eval rakudo 723986: OUTPUT«===SORRY!===␤Symbol '$a' not predeclared in <anonymous> (/tmp/4aRnH9xNAp:22)␤»
std 625303c: OUTPUT«Potential difficulties:␤ Useless redeclaration of variable $*a (see line 1) at /tmp/ApN9d13hU4 line 1:␤------> my $*a = 1; say $*a; my $*a⏏ = 2; say $*a␤ok 00:01 121m␤»
pmurias rakudo: my $*a =1;say $*a
p6eval rakudo 723986: OUTPUT«1␤»
masak TimToady: should pmurias' 'say $a' find $*a?
TimToady no
masak phew. :) 17:17
TimToady Sanity is overrated, but that doesn't mean it's worthless. :)
pmurias is $*foo derived from the lisp tradition of calling dynamic vars *foo*
?
PerlJam TimToady: would looking for $*a find an $a in an outer scope? 17:18
TimToady no
only in the final global package would it find a $a
PerlJam excellent
TimToady it strips the * before the final lookup
17:19 masak left, impious joined
TimToady I guess that's what I was thinking of earlier, so $*FOO::bar shouldn't have a problem, if package FOO is in GLOBAL 17:19
17:19 cotto_work2 joined
TimToady (and if we do the failover on not finding it) 17:20
17:20 cotto_work left
TimToady deleting keys from a shadowed hash is a bit dicey though 17:20
might need a :deleteall
if :delete defaults to the first hash that contains the key
PerlJam :obliterate 17:21
TimToady shadow hashes are basically a user-view generalization of what lexical scopes already do 17:22
so maybe hashes can have two different default policies as add ons
1) use this default value 17:23
2) use this default hash
PerlJam except with lexical scopes, you can look up-scope only so many levels with appropriate syntax. How do you do that with shadow hashes?
(I'm not even sure of the utility of this, just trying to stretch the analogy a bit) 17:24
TimToady %hash.defhash.defhash<foo> 17:26
17:26 masak joined
PerlJam ah. Equally dehuffmanized too :) 17:27
though with less shouting :)
17:27 rbuels joined
sECuRE moritz_: thanks for merging! 17:27
moritz_ sECuRE: thanks for creating the patch 17:28
17:29 kst` left, rgrau joined, kst` joined, MayDaniel joined
sECuRE moritz_: you can now close #77992 :) 17:30
17:30 Tedd1 left
moritz_ sECuRE: I'm already on it 17:31
TimToady hmm, %hash<foo>:delete would automatically delete from the hash it found 'foo' in 17:32
since it's a subscript modifier, not a method on %hash
17:32 MayDaniel left
diakopter want: www.newegg.com/Product/Product.aspx...6820148349 17:32
TimToady %hash.alldefhashes»<foo>:delete would do the other thing
or just make people write the loop 17:33
pmichaud have: www.newegg.com/Product/Product.aspx...6820167031 :-P
PerlJam did you really mean %hash.defhash»<foo>:delete ?
TimToady that only finds one def hash, and I want the whole pedigree 17:34
PerlJam Hmm.
masak is slightly unnerved and thinks that shadowed hashes should be opt-in
TimToady they would be
at least explicit ones
masak consider me re-nerved, then. 17:35
TimToady but dynamic hashes with :: might be an opt-in
that's what I was saying earlier
masak *nod*
I got that.
TimToady just trying to make you namespacehappy
masak yes, I probably shouldn't complain while you're solving my problems. :P 17:36
diakopter just got a core-i7-2600 with sata3, and the disk is loud compared to the rest of it. so I want a sata3 and 256GB. eh, maybe the 128GB is enough.
TimToady and I do think shadow hashes are the right way to go, not lying to ourselves about ::
pmichaud diakopter: I find I get by just fine on 80gb
diakopter: I have a standard spinning-disk drive for storage
although I'm thinking of replacing that with an ultra-quiet drive
diakopter looks at the sata3 intels
17:37 mberends joined
diakopter just crucial 17:37
pmichaud so, I have my OS, home dir, and working file set on the 80gb ssd for speed. Then I keep lesser-used files and archive storage on the 320GB hdd
PerlJam TimToady: I tend to agree, but I'm not enamored of the names you've used today :)
TimToady method names? those are certainly negotiable 17:38
pmichaud (and I'm thinking of getting a 1TB caviar green drive)
afk for a bit
17:38 Tedd1 joined
PerlJam TimToady: unfortunately, I have nothing to negotiate with at the moment. Just my general dislike. 17:39
TimToady the spot where the names show up to the user will be in the declaration with "is", generally: my %foo is default(42) 17:40
that's probably the right name for the default value
for the other, I dunno, "is shadowing(%OUTER::FOO)" is a bit long 17:41
maybe it's huffmanly correct though 17:42
PerlJam that's what I think
pmichaud huffmanly correct +1
TimToady might be there's a better metaphor than "shadow" though
anyway, I should finish backlogging and then work on my talk slides for India 17:43
I'm gonna try to import into OOo, btw, since the .odf is just a zip of xmls, and I can reverse engineer the markup I want 17:44
and it'll act like a real presenter tool, hopefully 17:45
pmichaud so far I'm not giving many presentations this year, so my need for a presentation tool has lessened :=| 17:48
but I still want to write one :-)
masak in Perl 6?
pmichaud ideally, yes.
I'll use whatever tool works best... but I think it'd be good to push Perl 6 development a teeny bit with it 17:49
jnthn home :)
o/ pmichaud
#phasers in...70 mins?
mberends \o/ +1 17:51
jnthn 71 mins? ;)
17:51 pmurias left
mberends :) 17:51
pmichaud I may be late to #phasers 17:53
dalek odel: b3bdc96 | (Martin Berends)++ | java/ (13 files):
[java] augment FindMethod and Invoke with SpecialFindMethod and SpecialInvoke, still passing 45 NQP tests out of the 299 that dotnet/ does
 17:57
18:01 plobsing joined
jnthn mberends: yay :) 18:03
18:04 felliott left
mberends jnthn: your recent 6guts post is very encouraging for 6model/java as well :) 18:08
18:11 felliott joined
icwiener Hi, is there a Perl6 user channel? 18:14
PerlJam This is it.
icwiener Oh. :)
PerlJam (so far)
masak we decided to do everything in one channel.
TimToady s/user/abuser/
masak :P
we might decide otherwise as Perl 6 grows, but right now having everything in one place has its advantages.
icwiener Sounds reasonable. :) 18:15
TimToady well, we already have two places, if you count #phasers
masak also, if you're a Yapsi developer, you're both a Perl 6 user and a Perl 6 dev. :P 18:16
icwiener Just a small question. I saw pastebin.com/11K6KW4k on learningperl6.com and am a bit confused. Isn't the line written to the file unchanged and then scrambled afterwards? 18:17
18:18 dakkar left
masak first off, that's old Perl 6. 18:18
'for =$input -> $line' is now 'for $input.lines -> $line'. 18:19
icwiener Meh ... :) Times fly by these days.
masak second, the $line parameter as declared is read-only, so you can't just go s/// on it and expect it to work.
either that code was never tested against an implementation, or it ran against Pugs at some point. 18:20
if you do want to modify $line, you should do '-> $line is copy'.
but you'd still need to write it back to the/a file, if that's what you want. 18:21
icwiener: does that answer your question?
tadzik =$input looks like POD :)
masak long live the death of prefix:<=>! 18:22
flussence reminds me of <?=$var?> from somewhere... :)
icwiener masak: Fully. :) Thanks for the advice. I will try (rakudo just finished compiling) to use it.
masak icwiener: let us know how it goes. we're eager to fix issues and help with stuff.
icwiener masak: Are there more up-to-date introductions to the language features? Maybe github.com/perl6/book 18:23
masak: Sure. :)
PerlJam wonders if merlyn or bdf still have any interest in that book. 18:25
the proposal was written a little over 4 years ago
masak icwiener: yes, that one is up-to-date. if you're willing to take some serious technical lingo, the synopses at perlcabal.org/syn/ are very good as well.
18:26 woosley1 left
icwiener Thanks. :) 18:26
18:33 jarfhy joined
dalek odel: a6a9a49 | (Martin Berends)++ | java/ (2 files):
[java] add definedness and defaultValue arguments to Parameter constructor
 18:33
18:34 cafesofie joined, Mowah left
moritz_ foegot to bump PARROT_REVISION after the v6 patch 18:38
18:40 cjk101010 joined
moritz_ could somebody please bump it to RELEASE_3_0_0-235-g102e7e2? 18:43
18:44 kst` is now known as kst 18:47 impious left 18:49 timbunce left 18:51 rgrau left 18:52 rgrau joined 18:53 icwiener_ joined 18:55 icwiener left 18:56 timbunce joined
tadzik I can :) 18:56
moritz_ tadzik: I was faster 18:57
tadzik yeah, just pulled :)
moritz_ tadzik: handed the small off to her mother, and now I have my hands free again :-)
18:57 vmspb joined
moritz_ masak: did you read my blog post with comments on the p6cc solutions and reviews? 18:58
dalek kudo: 188755a | moritz++ | build/PARROT_REVISION:
bump parrot version to something that can deal with IPv6
dukeleto hopes to see ipv6 support in the next rakudo dev release 18:59
tadzik :)
moritz_ dukeleto: it's already in 19:00
dukeleto: the bump should have been included in the previous commit, I just forgot it
masak moritz_: no, I'm srsly backlogged on the blogosphere. url? 19:01
moritz_ planetsix :-)... I'll get the URL for you 19:02
masak: perlgeek.de/blog-en/perl-6/thoughts-on-p6cc.html
masak reads voraciously 19:03
moritz_: already on the first bullet point, it strikes me that I might want to incorporate some of these things (with attribution) into my review notes. would that be OK? 19:04
moritz_ masak: sure
masak \o/ 19:05
"Just strange that he then reinvented subtraction" -- I didn't even think of that!
well, it only goes to show that *conceptually*, colomon didn't think of the Complex numbers as number data types, but just as a container data structure. 19:06
could as well have used Pair.
colomon masak: actually (as I just commented on in moritz_'s blog post) it's more an artifact of how the code was written.
masak oh, Complex was a late addition? 19:07
moritz_ retrospect knows always best
colomon Switching the points to a Complex type was something like the third different representation I used for them in that script.
masak moritz_: "hindsight is 20-20"
19:13 icwiener joined
colomon masak: by the way, the quick summary of my version of p2 is that almost the entire thing is attempting to get the result of fox's ($x < ($b[0]-$a[0]) * ($y-$a[1]) / ($b[1]-$a[1]) + $a[0]) carefully. But I admit it's hard for me to explicitly justify my instinct that that expression is just begging for floating point issues. 19:16
19:16 icwiener_ left 19:20 stkowski joined
masak colomon: I readily confess to not being an expert in computational geometry, so I don't have much preference for either long or short solutions. but I'd be interested in a case where the two algorithms differed in some significant way. 19:23
colomon that's the thing. barring the boundary cases, I think they only differ in extreme cases. 19:24
If you don't care about the boundary and can live with occasional weird results given weird data, fox's solution is brilliant. 19:25
and crazily enough ( as you realized) you could make it better by golfing it a bit more!
icwiener Yay. :) I struggled a bit with the output to a file but after "say $out: $line" and "$out.say $line" I found "$out.say($line)" :) Are parens mandatory there because say is used as a method? 19:29
colomon icwiener: you can also say $out.say: $line 19:30
19:31 tylercurtis joined, icwiener_ joined
icwiener_ colomon: I think $out.say($line) comes more naturally to me for now. :) 19:32
icwiener_ hates his roommate who reconnects all the time for being able to download more files from some mega download website. :/ 19:33
Anyway, next stop: regexp :)
19:34 icwiener_ left, daxim left, icwiener left 19:35 jlaire left 19:37 jarfhy left, jevin joined 19:38 cafesofie left 19:54 V15170R joined 19:56 hans_ joined, klunky joined 20:11 MayDaniel joined, timbunce left 20:16 cdarroch joined, cdarroch left, cdarroch joined, MayDaniel left 20:18 timbunce joined 20:24 coldhead joined 20:25 shi joined 20:29 timbunce left 20:31 cjk101010 left 20:35 timbunce joined 20:44 plainhao left, _jaldhar left 21:01 timbunce left 21:03 kst left, kst joined 21:04 fhelmberger left 21:08 timbunce joined, hans_ left
masak quiet evening. 21:10
21:11 Kodi joined
[Coke] LALALALALALAICANTHEARYOU 21:11
tadzik sleepy evening
Kodi Locally, I've got ^^ and xor all fixed and spiffed up except for one snag: *-currying. The problem is that in order to turn Parrot Undefs into Perl 6 Falses, I wrap a subroutine call around the PAST node for the infix op, which hides the structure that Action.pm's &whatever_curry expects. But if I were to implement ^^ and xor myself as a subroutine call instead of using the "xor" pasttype, that would destroy short-circuiting. So, 21:15
anybody know of a good equivalent to the "call" pasttype that doesn't automatically evaluate every child node?
masak huh. wish I did. 21:17
21:18 pothos left 21:19 y3llow left 21:20 y3llow joined 21:21 pothos joined
Kodi I guess I could wrap every child node up in a block before passing them to my custom xor, but that sounds pretty kludgey. 21:21
masak hey, if it passes the tests... :) 21:23
21:23 molaf__ left
tylercurtis How are && and || implemented? 21:23
Kodi tylercurtis: With special, magical pasttypes.
tylercurtis: The one for xor is deficient only in that it returns a Parrot Undef when it has more than one true argument.
moritz_ Kodi: a PAST::Block isn't automatically called, afaict 21:24
so you'd have to promote all statements to blocks or "thunks"
masak Kodi: sounds like it's the xor pasttype that ought to be fixed, then.
moritz_ and then call them from infix:<^^>
moritz_ -> sleep
Kodi moritz_: Right, that's the idea, I think.
PerlJam Kodi: if nothing else, you'll make a pattern for short circuiting that other people can use :) 21:25
Kodi masak: Y'know, I'm thinking you're right. The xor pasttype should probably take an additional special argument meaning "language-specific false value to be returned when we get multiple true arguments". 21:26
21:26 plobsing left
masak Kodi: why is this a problem only for xor? 21:26
Kodi: as you yourself said, || also returns one of the operands when given an alternative. 21:27
Kodi masak: Whenever one of the other short-circuiting operators returns a value, that value was one of its arguments.
masak hm. right.
Kodi xor is special in that when it returns false, the false value comes from nowhere.
masak Kodi: if the additional special value is optional, I don't see the harm. 21:28
Kodi Unfortunately, I'm not a Parrot developer.
PerlJam Kodi: surely you can wrap xor in a sub that does: if xor == Undef, return False
or soemthing like that
Kodi PerlJam: That's exactly what I did, and which led to the Whatever-currying problem. 21:29
PerlJam oh.
21:30 felliott left 21:32 rindolf joined
rindolf Hi all. 21:32
masak rindolf, hi. 21:33
rindolf masak: what's up?
masak rindolf: I have a belly full of noms and roof over my head. not complaining :) 21:34
what about you?
rindolf masak: well, I was finally able to boot my laptop with kernel 2.6.37 .
masak: and I've been working on Lingua::IT::Ita2heb.
Which transliterates Italian to Hebrew.
A friend of mine created a project for it on github and I joined him. 21:35
It's not on CPAN yet.
I moosified it.
And defined my first two roles in the process.
roles seem nice.
masak they are. 21:36
PerlJam rindolf: you have to be careful about composition in Moose land sometimes, but generally they are very nice
masak PerlJam: isn't there some way to make Moose touchy about composition? 21:37
21:37 felliott joined
masak felliott++! \o/ 21:37
felliott masak++!! 21:38
masak "Warning. Karma addition cycle detected."
rindolf masak: why does felliott deserve a ++?
masak rindolf: he was the only one doing things in RT and in the test suite a couple of days back. 21:39
rindolf masak: ah, OK.
felliott++
masak it was nice to see the activity roll by, and I feel he deserves karma for it.
pro tip: cause activity in RT and/or spectest suite, and I will give you lots of karma. :) 21:40
probably others as well.
21:43 Chemtox joined 21:49 Mowah joined
tylercurtis PerlJam: ooc, what is it about composition in Moose that one must be careful about? 21:50
21:51 Chemtox left 21:59 Kodi left 22:07 tylercurtis left
dalek p-rx/nom: 227012d | jonathan++ | / (3 files):
Start clearing up the grammar migration. Remove the now-disused original protoregex bit.
 22:14
p-rx/nom: 96bf49d | jonathan++ | / (4 files):
Remove pre-port cursor code.
p-rx/nom: ca48590 | jonathan++ | src/Regex/Cursor.pm:
Toss initial experimental port of Cursor to NQP. See 6model repo for diakopter++'s work on this, which is blazing the Cursor-in-NQP trail.
p-rx/nom: 7ef57d5 | jonathan++ | src/Regex/Cursor2.pir:
Install the migrated Cursor, so far just installed as Regex::Cursor2, also as Regex::Cursor.
22:15 MayDaniel joined 22:18 c1sung_ left 22:19 vmspb left 22:20 bacek left 22:24 c1sung joined 22:27 shi left
dalek p-rx/nom: 5cb2f1f | jonathan++ | src/Regex/Cursor2 (2 files):
Eliminate Regex::Cursor2 namespace; now just the Cursor2 type object alias installation remains, plus some vestigal naming references.
 22:32
p-rx/nom: 4f57429 | jonathan++ | src/ (3 files):
Track down remaining uses of Cursor2 type object and eliminate them.
p-rx/nom: aa8293f | jonathan++ | src/Regex/Cursor2 (2 files):
Eliminate use of Cursor2 in sub IDs.
22:34 c1sung left, bacek joined, c1sung joined, george8 joined 22:35 george8 left 22:36 estrabd left 22:38 estrabd joined 22:39 noganex_ joined 22:40 shi joined, snearch joined 22:42 noganex left 22:43 c1sung left 22:44 kaare_ left, snearch left 22:45 MayDaniel left
dalek p-rx/nom: beeb40a | jonathan++ | src/stage0/ (4 files):
Update bootstrap to eliminate usage of Regex::Cursor2 type object alias, so it can finally be removed.
 22:45
p-rx/nom: 40c4ce9 | jonathan++ | src/Regex/Cursor2.pir:
Remove installation of Regex::Cursor2.
p-rx/nom: 5239c81 | jonathan++ | / (6 files):
Rename files that had Cursor2 in the name.
masak jnthn++ 22:46
jnthn Bootstraps. They're like walking on dominos carefully arranged in a circle.
masak you could say that again.
jnthn Bootstraps. They're like walking on dominos carefully arranged in a circle. 22:47
masak ...and you did. :)
jnthn Hey, you said I could. :P
dalek p-rx/nom: 619c67d | jonathan++ | src/stage0/ (4 files):
Final bootstrap update to completely eliminate Cursor2's legacy from the source tree.
 22:50
22:50 bacek left, timbunce left 22:51 felliott left
jnthn Now I can do something interesting... :) 22:53
22:54 felliott joined
masak std: my Int ($a, $b) 22:54
p6eval std 625303c: OUTPUT«ok 00:01 120m␤»
masak rakudo: my Int ($a, $b); $b = 3.4; say $b 22:55
p6eval rakudo 188755: OUTPUT«Type check failed for assignment␤ in '&infix:<=>' at line 1␤ in main program body at line 22:/tmp/b5eEvRPLWW␤»
masak rakudo: class A { has Int ($.a, $.b) }; say A.new(:b(3.4)).b
p6eval rakudo 188755: OUTPUT«3.4␤» 22:56
jnthn Somebody should fix objects. :P
masak jnthn: is it just that attr typechecking is off?
jnthn masak: ja
masak I'm pretty sure there's a ticket for that somewhere.
jnthn Me too.
23:03 kst left 23:04 kst joined 23:15 c1sung joined 23:19 bacek joined 23:21 c1sung left 23:25 dukeleto left, dukeleto joined 23:27 c1sung joined, cotto_work2 is now known as cotto_work
masak rakudo: my %h = 1..5 Z <a> xx *; say %h.perl 23:28
p6eval rakudo 188755: OUTPUT«{"1" => "a", "2" => "a", "3" => "a", "4" => "a", "5" => "a"}␤»
masak rakudo: my %h = 1..5 Z <a> xx *; %h.delete(<2 3 4>); say %h.perl
p6eval rakudo 188755: OUTPUT«{"1" => "a", "5" => "a"}␤»
masak \o/
23:32 Mowah left 23:34 c1sung left
dalek p-rx/nom: a9206dd | jonathan++ | / (4 files):
Add NQPNativeHOW meta-object for native types and native package declarator to go with it.
 23:37
p-rx/nom: 510f09f | jonathan++ | src/ (3 files):
Make it possible to give an attribute a type constraint. Currently only the P6opaque REPR pays attention to this.
p-rx/nom: 21c8f5b | jonathan++ | src/NQP/Actions.pm:
Cleanup some leftovers from the 6model migration.
p-rx/nom: 4683d55 | jonathan++ | src/stage0/ (3 files):
Push native into the bootstrap.
masak rakudo: sub foo(@a = [1, 2, 3], *@b) {}; foo() 23:39
p6eval rakudo 188755: ( no output )
masak rakudo: sub foo(@a = [1, 2, 3], *@b) {}; foo(42)
p6eval rakudo 188755: OUTPUT«Nominal type check failed for parameter '@a'; expected Positional but got Int instead␤ in 'foo' at line 1:/tmp/eJDCjCKYyX␤ in main program body at line 22:/tmp/eJDCjCKYyX␤»
masak right.
23:44 shi left 23:46 c1sung joined
masak rakudo: sub foo(@a is copy where { .elems == 0 | 4 }) { @a ||= -1, -1, +1, +1; say @a }; foo([1,2,3,4]); say "alive" 23:46
p6eval rakudo 188755: OUTPUT«alive␤»
masak submits rakudobug
jnthn masak: er.
You're expecting the constraint to get attached to the container? 23:47
masak rakudo: sub foo(@a is copy where { True }) { @a ||= -1, -1, +1, +1; say @a }; foo([1,2,3,4]); say "alive"
p6eval rakudo 188755: OUTPUT«alive␤»
masak jnthn: I'm expecting &foo to output something.
jnthn masak: ...oh. 23:48
er, yeah, me too. :/
23:48 Franek joined
masak rakudo: sub foo(@a is copy) { @a ||= -1, -1, +1, +1; say @a }; foo([1,2,3,4]); say "alive" 23:48
p6eval rakudo 188755: OUTPUT«alive␤»
masak huh. no where clause needed.
jnthn is it the is copy?
masak otherwise we can't assign.
this feels familiar... 23:49
masak czechs
I might have been remembering rt.perl.org/rt3/Ticket/Display.html?id=74430
jnthn oh 23:50
looks related.
masak yeah.
this goes in too, though. 23:51
jnthn Please add a ref to the other ticket.
masak sure thing :)
jnthn rakudo: sub foo(@a) { say PARROT(@a) }; sub bar(@a is copy) { say PARROT(@a) }; foo([1,2]); bar([1,2]) 23:52
p6eval rakudo 188755: OUTPUT«ObjectRef->ObjectRef->Array␤Array␤»
jnthn FAIL
masak rakudo: my @a; @a ||= -1, -1, +1, +1; say @a.perl
p6eval rakudo 188755: OUTPUT«[-1]␤»
masak not what I expected, but better than the other one :) 23:53
is ||= always item assignment?
dalek p-rx/nom: d5dc5d8 | jonathan++ | / (2 files):
Add int, num, str.
p-rx/nom: 1e3ab6a | jonathan++ | src/metamodel/how/NQPNativeHOW.pm:
Add int.^name et al.
jnthn Not sure off hand.
std: my @a; @a = my $a, my $b; 23:54
p6eval std 625303c: OUTPUT«ok 00:01 122m␤»
jnthn std: my @a; @a ||= my $a, my $b;
p6eval std 625303c: OUTPUT«ok 00:01 122m␤»
jnthn oh, hmm
jnthn wanted to make std spit out an unused warning or something :)
masak not sure STD does sink checks.
jnthn Yeah...it does *something* but not that. :)
23:56 c1sung left
jnthn masak: See if STD.pm6 around 3326 makes any sense to you :) 23:56
masak looks 23:57
jnthn There's some condition in there that determines whether it's list or item assignment.
masak makes a lot of sense. 23:58
jnthn Rakudo gets it RONG
token infix_postfix_meta_operator:sym<=> { '=' <O('%item_assignment')> } 23:59
Unconditionally item ass.
Not sure how hard it is to port the STD thingy.
23:59 kst left
jnthn Probably not too bad. 23:59
masak right. ||= should be item ass., though.

Powered by Cro and the Raku® Programming Language.

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

Thank you!