»ö« | perl6.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo: / pugs: / std: , or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by diakopter on 25 January 2010.
pmurias diakopter: is that what you want? 00:01
diakopter pmurias: I guess, but I don't see any spectests displaying that syntax 00:04
pmurias diakopter: it's specced in the synopses 00:05
my int16 @array[100];
sleep&
diakopter I don't see that anywhere in S02 00:07
diakopter besides, how would you enforce that push/pop/unshift/shift/splice don't operate on a fixed-length Array? 00:08
pugs: my Int @array[100]; say @array.elems; 00:09
p6eval pugs: OUTPUT«0␤»
diakopter rakudo: my Int @array[100]; say @array.elems; 00:10
p6eval rakudo 1d4928: OUTPUT«Confused at line 10, near "[100]; say"␤in Main (file <unknown>, line <unknown>)␤»
diakopter .sigh 00:10
Limbic_Region long time no chat diakopter - how's life? 00:11
diakopter Limbic_Region: ok 00:14
elf: my Int @array[100]; say @array.elems; 00:16
p6eval elf 29594: OUTPUT«syntax error at (eval 124) line 3, near ";->postcircumfix__91_32_93"␤ at ./elf_h line 5881␤»
diakopter mildew: my Int @array[100]; say @array.elems;
p6eval mildew: OUTPUT«Can't open perl script "mildew": No such file or directory␤»
diakopter mildew-js: my Int @array[100]; say @array.elems;
p6eval mildew-js: OUTPUT«Can't open perl script "mildew": No such file or directory␤»
diakopter ng: my Int @array[100]; say @array.elems;
p6eval ng b8201f: OUTPUT«0␤»
diakopter :| 00:17
diakopter pugs: my Int @array[-1]; say @array.elems; 00:26
p6eval pugs: OUTPUT«0␤»
diakopter pugs: my Int @array[-33]; say @array.elems;
p6eval pugs: OUTPUT«0␤»
diakopter pugs: my Int @array[-33,-66]; say @array.elems;
p6eval pugs: OUTPUT«0␤»
diakopter augh
I finally found S09. 00:28
diakopter TimToady: can the length of a fixed-size array be defined with a runtime value? my $a = rand(99); my int @vals[$a]; 00:34
std: my $a = 99.rand; my int @vals[$a]; 00:36
p6eval std 29594: OUTPUT«ok 00:01 107m␤»
diakopter std doesn't mind it
TimToady in a meeting, but you'd need a BEGIN in there somewhere 00:37
current spec assumes a compile-time $a, I think 00:39
would be possible to relax that to elaboration time, I suppose
diakopter wants compact arrays of runtime-defined fixed length 00:40
TimToady assuming the constructors support it, shouldn't be too hard to hook to the declarative syntax 00:42
diakopter seems like there should be a property (or attribute) on Array that specifies whether it's fixed-length
, too
or better, whether it's not fixed-length
diakopter std: my buf @array[@@array] 00:44
p6eval std 29594: OUTPUT«ok 00:01 106m␤»
lichtkind_ anybody alive here? 01:00
lichtkind anybody here? 01:12
wayland76 When's the next release? 01:33
diakopter wayland76: of? 01:34
rakudo? the january one was skipped.
wayland76 Ah, ok, so Februiary is the next Rakudo release (and yes, that was what I was asking; thanks for anticipating :) ) 01:38
Tene is augment lexically scoped in rakudo? 04:40
sjohnson damn abokgb 04:46
wasting my precious bandwidth
diakopter sigh 05:36
at least those don't make it to moritz_'s ilbot irclog 05:37
diakopter TimToady: may I add the mention of my int @fixed_length[$a]; to S09? 05:59
sjohnson fuckin versions!!! 10:04
*shakes first in air* 10:05
moritz_ good morning 10:16
can't we somehow disallow these CTCP VERSION requests?
moritz_ is not very familiar with IRC and channel modes
sjohnson i will look into it 10:20
from what i can tell from Freenode docs 10:21
it might not be doable
where is that friendly ircop angel who hangs around here...
kloeri_ or something
sjohnson moritz_: tomorrow, when Seven is migrated to (ircd) 11:28
+C will disable these annoying CTCP's for chans
moritz_ sjohnson: thanks 11:29
sjohnson no worries 11:34
pugs_svn r29595 | masak++ | typo 13:11
dalek ok: 7b9d5cb | moritz++ | src/builtins.pod:
list some Str builtins
 13:40
ok: 25b257d | moritz++ | src/ (4 files):
Merge branch 'master' of [email@hidden.address]
pmichaud good morning, #perl6 13:58
moritz_ \o/ 13:59
colomon \o
jnthn morning, pmichaud 14:04
pmichaud jnthn: I figured out the flatten flag late last night :)
colomon \o/
pmichaud jnthn++ # you were right about caching
except instead of caching the flag on the instances, we cache it in the class object :) 14:05
so now it's just $P0 = typeof value; ...check/cache flatten in $P0... 14:06
even nicer is that a lot of our value pmcs now don't end up with prophashes 14:07
jnthn pmichaud: :-D 14:08
pmichaud: Nice.
pmichaud: Was alongish the lines of what I was thinking. wfm. :-)
mathw yay! that sounds like efficiency
pmichaud yes, it should be fairly efficient, while also preserving the correct p6 semantics 14:09
so, we now know that Iterable objects flatten in list context.... do Iterators themselves flatten (unless masked in a scalar container)?
colomon Hmmm... we know we want to be able to create Arrays of Iterators, right? 14:15
pmichaud yes 14:16
but in that case, each of the iterators would be masked by a scalar container
pmichaud (the element of the array is a scalar container) 14:16
colomon ah.
so that doesn't provide any useful insight. 14:17
pmichaud I think I'm going to start out assuming that they do flatten... it simplifies list processing greatly if so 14:18
colomon I think my instinct is that iterators flatten, but I don't know that I can justify it.
colomon pmichaud++ # JFDI and see what happens 14:22
mathw As I haven't really looked into it, this is pure opinion, but I would expect iterators to act like the list they generate in appropriately listy situations 14:26
which means flattening, I guess
Su-Shee (I've just read "flattering".. time to call it a day...) 14:28
mathw hmm 14:48
bitwise OR in C++ now looks wrong to me 14:49
even though I've not written any perl 6 for a while, I keep thinking I'm making a junction
moritz_ :-) 14:49
mathw and it also feels wrong to use just one character for an operator which, most of the time, isn't actually used very often 14:50
colomon :)
mathw I should write some Perl 6
mathw My boss has suggested I run my Perl course again, but I don't think I could get away with upgrading it yet :( 14:51
corpogen hello guys im having a problems running somethig called celera software 15:03
moritz_ is that related to Perl 6 in any way?
corpogen this is the routine and the output scsys.co.uk:8002/38696.
moritz_ doesn't look like Perl 6 :/ 15:04
so you're in the wrong channel
corpogen no idea
moritz_ now you know 15:05
sjn corpogen: try asking in #perl and make sure you have all necessary info in that pastebin page (including the code that the error message refers to) 15:12
sjn corpogen: or, reimplement your tool in Perl 6, and hang around here ;-) 15:13
pugs_svn r29596 | colomon++ | [t/spec] More tests for RangeIterator. 15:21
PerlJam guten Morgen 15:52
jnthn PerlJam: Wie gehts? 15:53
jnthn ...or something... :-) 15:53
PerlJam um .... I've already exhausted my conversation German, sorry :) 15:54
*conversational
jnthn: But ... I'm fine, how are you? :)
colomon Bag(3,8,[2,Seq(9,3)],8) >>->> 1; # Bag(2,7,[1,Seq(8,2)],7) 15:57
How is that even supposed to begin to be implemented?
PerlJam colomon: magic, as usual :) 15:58
colomon When I was looking at that yesterday, I was assuming the end result was a list. 15:59
PerlJam colomon: is hyper really supposed to recurse like that? 16:01
jnthn PerlJam: I'm...strangely exhausted. 16:02
colomon PerlJam: that's start from the spec.
*straight
PerlJam The fun of not always being detail oriented is that there's always these new little gems in the spec that surprise you. 16:03
(of course, the spec being in flux increases the surprise factor too :) 16:04
pmichaud and the surprises have kept on coming, lately! :-) 16:06
colomon pmichaud: it's easy for me to see how to you can connect iterators and hyper operators. 16:09
but building up an identical structure afterward?!??
pmichaud I'm not sure about that "afterward" part, though. It could be builtup/maintained as it runs. 16:10
colomon either way, that's a lot more complicated than the current Iterator interface.
pmichaud sure, but that's not a function of iterator anyway 16:11
The Bag sees four elements from its iterator
jnthn colomon: It's not so hard maybe.
colomon: I mean, it may be like
pmichaud the third element is an Array, which is fed to its own hyperiterator invocation, which returns another array
jnthn sub hyper_helper(::T @items) { ... } and you can do T.new(@result) in there or so. 16:12
pmichaud that array itself has two elements, with the Seq being fed to yet another hyperoperator
jnthn There's ways to write generic routines. :-)
(I'm assuming a recursive approach there...I guess pm maybe is too.)
pmichaud yes
basically it ends up acting like Bag(3-1, 8-1, [2, Seq(9,3)] >>->> 1, 8-1) 16:13
which in turn
Bag(3-1, 8-1, [2-1, Seq(9,3) >>->> 1], 8-1)
and then
Bag(3-1, 8-1, [2-1, Seq(9-1, 3-1)], 8-1)
jnthn I think the current impl may know how to do that. 16:14
rakudo: say ([1, 2, [3, 4], 5] >>+>> 1).perl
pmichaud oh, I doubt that part. Anyway, the hyper metaop would just thread over the structures
p6eval rakudo 1d4928: OUTPUT«[2, 3, [4, 5], 6]␤»
pmichaud okay, looks like it does work :-) 16:15
jnthn pmichaud: Well, it knows to recurse.
pmichaud: Not about types though.
pmichaud adding types wouldn't be too difficult, I don't think
jnthn Me either.
I think hypers on hashes work too.
jnthn used to be, like, productive.
Hopefully Feb I'll me productive again too. :-) 16:16
*be
pmichaud I'm sure I will.
I'm being productive now, although only in terms of effort and not results :)
(although we're getting closer)
jnthn Spending time on design is fine. :-) 16:17
pmichaud more like "implementations that don't quite work"
jnthn "research" :-)
pmichaud I guess that makes me a "scout ant" for the time being. :)
although I will be very happy to eliminate the 'flatten' flag here in a bit :-) 16:18
lichtkind could anybody reset my pugs account? 16:20
pmichaud ooops 16:23
I just realized the release announcement for Jan 2010 didn't make it off my machine
Should I resent?
*resend?
colomon it seems like that ship has kind of sailed... 16:27
colomon jnthn: looking at your hyper_helper back there... does that imply we can do it without support from the class itself other than an appropriate .new being available? 16:30
jnthn colomon: Yes. 16:32
colomon sweet.
jnthn colomon: That's how the current impl works, iirc.
colomon now studying the master implementation... 16:41
lisppaste3 frettled pasted "Blog entry about Rakudo ng as new master (feedback wanted)" at paste.lisp.org/display/94097 17:00
frettled If I'm off the mark in any way, please whack me. 17:01
(I may have missed some goings-on in this channel due to some spamming, please bear with me.)
pmichaud frettled: I don't see anything wrong with the article, but it makes me think we ought to list what some of the improvements of ng are 17:06
diakopter lichtkind: still there?
pmichaud i.e., we're regressing, but we gain some very important stuff in return
frettled pmichaud: yes, very good point, that would help add a more positive spin
pmichaud for one, we'll have working laziness
we'll also have Parcel/List/Array/other types that fit the updated spec. Rakudo master was very out of date there. 17:07
pmichaud The ng version will support protoregexes and a better form of longest token matching 17:07
metaoperators are truly meta -- and they're dynamically generated on-demand 17:08
we have a base object metamodel that is far closer to the specification than what we had before. Also, major portions of the metamodel are written in Perl 6.
we should get some good speed improvements (many due to added laziness) 17:09
frettled excellent talking points
jnthn Also, from a Rakudo * point of view, many of the things Pm just mentioned directly address priority #1 ROADMAP items too.
pmichaud we don't have it yet, but shortly we'll have the ability to embed closures in regexes
frettled would it be correct to say that laziness will be "mostly working", or will it be complete? 17:10
pmichaud and variable interpolations
jnthn pmichaud: How about the auto-vivification things?
pmichaud "mostly working". The laziness spec itself is undergoing change.
Oh yes, array and hash elements will vivify properly.
jnthn \o/
pmichaud (they already do for arrays, although I'm still tweaking a bit there)
pmichaud lexical subs and variables work properly 17:11
operators have the correct names (i.e., with angles). subs have the correct sigils (i.e., with &'s) 17:12
we have working phasers 17:13
and a much improved phaser model
pmichaud anyway, those are a few changes that could not be easily added to master 17:15
(indeed, the reason master doesn't have many of them was because it was proving so difficult to do them... thus ng resolves a lot of those issues)
afk, lunch 17:16
frettled Brilliant, thanks a lot. 17:18
diakopter lichtkind: do you still want your pugscode password reset?
diakopter TimToady: may I add a mention of run-time-expression fixed-length array constructors to S09? 17:23
TimToady yes, question is whether the actual construction is driven by autoviv 17:24
diakopter oh
TimToady and what happens if you happen to be in the scope of the declaration but not have executed it
frettled There, blogified: howcaniexplainthis.blogspot.com/201...or-us.html
TimToady does it just use the current value of $a whatever it is then? 17:25
it's not just a walk in the park :)
frettled I really like grabbing the good talking points from #perl6 and blogging about it, I like to think that there are some readers who don't follow things very closely here ;)
colomon frettled++ 17:26
TimToady also, the type should probably be a container role instantiated with the particular shape
so that the type of two such arrays is officially different
diakopter I'd think it evaluates the expression at the time of initialization. 17:27
TimToady well, people are going to expect my @array[40]; to autoinit without an = 17:27
diakopter oh; I meant declaration. 17:28
TimToady while we don't assume .= new for scalar vars, maybe we can for autovivifiables
TimToady
.oO(Good design is a bitch, but I can hope it's my bitch.)
 17:31
anyhoo, it would be good if the lack of .= new could be made to fall out of autovivifiability without using an exception list of types 17:32
frettled :) 17:33
TimToady just as $x[3] = 4 creates an array in $x if it needs to, speaking of which, we need a mechanism for that too 17:34
rakudo: my $x; $x[3] = 4; say $x[3] 17:35
p6eval rakudo 1d4928: OUTPUT«Method 'postcircumfix:[ ]' not found for invocant of class 'Failure'␤in Main (file src/gen_setting.pm, line 324)␤»
diakopter I didn't know that
TimToady Perl 5 does it
diakopter pugs: my $x; $x[3] = 4; say $x[3]
p6eval pugs: OUTPUT«4␤»
TimToady basically, .[] on a Mu-filled scalar container should autoviv an array somehow 17:36
TimToady just as any element can autoviv a subarray 17:37
diakopter ok. since sprixel's first stage won't do autovivification, I can defer that..
colomon pmichaud: I do still have a question on iterators and hyper. What happens if you do something like >>-- to a hyper that works through iterators? Is .get supposed to return something which is rw?
diakopter I suppose that (my last msg) is a truism 17:38
TimToady you can say that again
that goes without saying
diakopter I mean, a tautology. 17:38
TimToady I prefer tauntaologies
s/ao/o/ 17:39
TimToady blames his battery, which is running low, because he forgot his power cord at work yesterday...
diakopter TimToady: maybe there's a different constructor that takes the expression 17:41
TimToady and does what with it, is the question 17:43
TimToady instantiates a new role with that length (unless cached), and the creates the class for that role, and the calls the "real" constructor on that? 17:45
TimToady do we want @a[2,2] to be a different type than @a[4]? Seems so 17:46
TimToady er, @a[2;2] rather 17:47
jnthn It sucks that the GNAA found a backdoor way to insert messages into your channel. :-/ 17:48
*our
jnthn backlogs recent discussion 17:49
TimToady will check back later & 17:50
diakopter I'd think the types would differ by dimensionality only 17:52
jnthn TimToady: I'd kinda seen us being able to do it as a role parameter too, fwiw.
TimToady: Since I'd got caching of roles based upon ===-style identify of the type parameters, I think it should work.
lichtkind diakopter: sorry had to go yes i still want the reset 17:56
diakopter lichtkind: well apparently I've forgotten my password (again), so I can't :) 18:05
but I'm sure someone else can
lichtkind i hope so 18:06
moritz_: you have pugs acount? 18:07
Khisanth jnthn: there is +R ...
jnthn Khisanth: +R ? :-)
jnthn goes to look it up
Khisanth silence unregistered
or the new ircd is supposed to have a mode that also stops the CTCP 18:08
jnthn ah, cool. 18:09
jnthn ain't so clued up on IRC
lichtkind jnthn: do you have pugs account? 18:10
jnthn lichtkind: Yes.
lichtkind jnthn: could you please reset my account?
jnthn lichtkind: Ah, I thought diakopter++ had already done it?
lichtkind: Hm. Is that a commitbit task... 18:11
lichtkind jnthn: he need also a pwd reset :)
jnthn ah, I guess it's reinvite
lichtkind: done 18:12
(hopefully) :-)
lichtkind :) 18:13
lichtkind jnthn: it warked thanks a lot 18:21
pugs_svn r29597 | lichtkind++ | included november wiki 18:23
lichtkind it does really work :)
colomon Hmmm. It seems like properly implementing the series operator, infix:<...>, requires an iterator internally, as the series may be large or infinite. 18:36
pmichaud back
colomon But what type should the series operator return?
pmichaud colomon: a list
colomon how do you wrap a list around an iterator? 18:37
pmichaud a list *is* an iterator :-)
colomon so you just return the iterator?
pmichaud essentially, yes 18:38
colomon what is the catch in that essentially? :)
pmichaud think of things in terms of gather/take... what does gather return?
it returns a list
the catch is that you have to return the iterator in a way that it will flatten 18:39
colomon ah.
pmichaud as opposed to in a scalar, where it won't flatten
colomon right.
pmichaud but if infix:<...> is written in terms of gather take, the gather handles all of that for you :-)
I should have a working gather/take soonish.
TimToady: we postulated earlier that iterators in list context flatten... does that sound rightish? (An iterator in a scalar container wouldn't flatten, however.) 18:40
colomon I guess this brings up the question of how often we should be going to be bare metal and directly coding an iterator, and how often we should just use gather/take.
pmichaud bare metal when speed is an issue, I'd think
or when it's pretty easy to do so
or when there are optimizations available 18:41
there's also the question of "how bare is the metal?"
:-)
colomon :) Sure enough, I prefer my bare metal still in Perl 6 rather than Parrot.
colomon I feel like iterators provide a target-rich environment for doing cool stuff, I'm just not sure where to tackle it. 18:51
pmichaud it'll become clearer, soon, I hope.
I'm fixing the failing 08-var-array.t test now (the error really doesn't have anything to do with arrays, though), next will be .map
that should get use pretty close back to the ng state 18:52
s/use/us/
Tene pmichaud: how is ng1 coming?
colomon I was just peering at STD.pm, trying to figure out how to bring the reduce meta operator into ng1, but I fear the grammar involved is still beyond me.
pmichaud slow
pmichaud but making good progress today. Yesterday posed a few blockers. 18:52
Tene Might have some time after work today. I don't think there's anything going on that needs me tonight. 18:55
pmichaud okay, great!
Tene Sorry I've been absent this week.
pmichaud actually, ng1 has some good stuff with Range and RangeIterator
Tene: np, absence is a not-uncommon trait for us all :)
sometimes real life and/or other priorities just show up 18:56
ng1 now passes "make test" 19:02
short break, then I'll do .map 19:03
colomon double \o/ 19:04
lichtkind what is vi? 20:23
i mean viv
lichtkind diakopter: you some minutes for an interview? 20:30
pmichaud ng: say 'alive' 20:38
p6eval ng b8201f: OUTPUT«alive␤»
_ash ah, cool, the eval bot is reading the chat logs again 20:43
ash_ rakudo: say 'hi' 20:44
p6eval rakudo 1d4928: OUTPUT«hi␤»
PerlJam pmichaud: how goes the ng/ng1 ?
pmichaud slow, but speeding up
I'm about to do (lazy) .map
almost have it ready, I think.
ash_ awesome, pmichaud++ 20:46
PerlJam When's the Rakudo release? 20:47
pmichaud already did it last week 20:48
PerlJam oh, I guess I missed it.
pmichaud apparently the announcement didn't leave my computer, though :-(
I've been wondering if I should go ahead and send now
ash_ pmichaud: i made a repository for my attempt at making a nqp-rx like language, based on the llvm, its only just started though, i am still learning flex/bision so it doesn't compile yet
pmichaud ash_++
PerlJam pmichaud: I just read the release announcement. I think the announcement itself is the most important part of the release :-) 20:52
pmichaud yeah, I'll send.
PerlJam (so, by all means, send it!)
pmichaud (re-send)
ash_ pmichaud github.com/ashgti/nq-nqp-rx if your at all interested, i am gonna keep on chucking on that 21:02
dukeleto how does one invoke/call a lambda in perl 6? { ... }.invoke ? 21:09
ash_ for a bare block? 21:18
pmichaud { ... }.() 21:19
ng: say { 3 + 4 }.();
p6eval ng b8201f: OUTPUT«sh: ./perl6: No such file or directory␤»
pmichaud bah
colomon The dot is needed?
ash_ no, but its more explicit 21:21
ash_ rakudo: -> $a { say $a; }.(5); 21:21
p6eval rakudo 1d4928: OUTPUT«5␤»
colomon rakudo: my $x = -> $a { say $a; }; $x.(5); 21:39
p6eval rakudo 1d4928: OUTPUT«5␤»
pmichaud jnthn: (just in case you're around).... Signatures are immutable, yes? 21:47
jnthn pmichaud: I dearly hope so. 21:48
pmichaud: You caught me just as I was about to log off and go to the pub. :-P
pmichaud aha!
currently .params computes a new list everytime it's requested
jnthn pmichaud: Yeah
pmichaud should we perhaps cache the result somewhere?
same for count and arity?
jnthn That was my intention, just never got around to it.
pmichaud okay
jnthn I completely intended to cache that result though. 21:49
But wanted to compute it lazily too.
Only got one out of two done. :)
pmichaud: Anything else before I go pivovat? 21:50
lichtkind jnthn: you mean pivovar :) 21:51
jnthn lichtkind: Not unless I'm going to brewery. :-) 21:52
lichtkind haha
jnthn lichtkind: "pivovat" is actually slang, not real word :-)
lichtkind it seemes to me more like drinkin 21:53
jnthn lichtkind: There's a bunch of Slovak verbs that end -ovat (not sure if in Czech too). :-)
Drink is pit
:-)
jnthn When I was in Prague the waiter called it a pivko. I laughed. :-) 21:53
lichtkind jnthn: there is to but it indicates a verb
jnthn aye. pivovat = to drink beer :-) 21:53
lichtkind pivko means little beer
jnthn Huh? I was drinking pints! :-P 21:54
lichtkind jnthn: good then i understood right
jnthn Maybe I shoulda asked for a velmi velke pivo :-)
lichtkind jnthn: normal beer = pivo , small = pivko , eben smaller = pivičko 21:55
even smaller
jnthn pivičko is a terrifying concept.
Ok, idem do krcmy. See you all later :-)
lichtkind have fun
pmichaud jnthn: no, nothing else -- just was noticing that &block.count was failing and then saw "hey! that's expensive!"
jnthn pmichaud: heh. Feel free to fix, or put on my task list. :-) 21:56
jnthn gone
colomon pmichaud: while you're messing around in there, have you noticed that grep (for instance) doesn't handle $_ properly in ng? 22:09
ash_ ng: grep { say $_; }, (1, 2); 22:28
p6eval ng b8201f: OUTPUT«Mu()␤Mu()␤»
ash_ hmm
i'd argue thats a bit off
colomon ash_: yes, exactly.
ash_ ng: grep { say @_; }, (1, 2); 22:29
p6eval ng b8201f: OUTPUT«␤␤»
dalek kudo/ng: 22b55c4 | moritz++ | t/spectest.data:
re-enable anon_block.t and class-and-instance.t
 22:30
TimToady ng: grep { say $^a; }, 1, 2; 22:56
p6eval ng 22b55c: OUTPUT«Null PMC access in get_string()␤current instr.: 'perl6;Perl6Role;perl' pc -1 ((unknown file):-1)␤»
TimToady ng: grep -> $a { say $a; }, 1, 2;
p6eval ng 22b55c: OUTPUT«Null PMC access in get_string()␤current instr.: 'perl6;Perl6Role;perl' pc -1 ((unknown file):-1)␤»
TimToady ng: grep -> $a { say $a; }, <a b>; 22:57
p6eval ng 22b55c: OUTPUT«a␤b␤»
TimToady ng: say 1 22:57
p6eval ng 22b55c: OUTPUT«1␤»
diakopter ng: grep { say $^a; }, 1;
TimToady ng: say 1, 2
p6eval ng 22b55c: OUTPUT«Null PMC access in get_string()␤current instr.: 'perl6;Perl6Role;perl' pc -1 ((unknown file):-1)␤»
ng 22b55c: OUTPUT«12␤»
pmichaud looks like parameter issue 22:58
pmichaud ng: { say $^a }.(1) 22:58
p6eval ng 22b55c: OUTPUT«1␤»
TimToady or an opaque container issue
pmichaud or a grep problem
TimToady ng: grep { say $^a.WHAT }, 1; 22:59
p6eval ng 22b55c: OUTPUT«Null PMC access in get_string()␤current instr.: 'perl6;Perl6Role;perl' pc -1 ((unknown file):-1)␤»
pmichaud > my $a = (2,3,4).map({ $_ * 3 }).iterator; say $a.get; say $a.get;
6
9
TimToady input side, not output
pmichaud yay!
we can has lazy map
pugs_svn r29598 | lwall++ | [perl6.org/index] demote Elf from the front page for lack of recent activity. 23:18

Powered by Cro and the Raku® Programming Language.

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

Thank you!