»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 wolfe.freenode.net on 30 October 2009.
colomon ash_: can you give an example I would understand of what's not working with BUILD? 00:00
ash_ t/spec/S12-construction/BUILD.t
try running that, its the issue i am talking about
ash_ its not passing paramters to BUILD, if you do Foo.new(1); BUILD should get the 1, and its currently not at the moment 00:03
the *right way* for me to fix List.new is to use BUILD so, i just found the problem 00:04
colomon is still trying to understand the issue.... 00:06
ash_ so, perlcabal.org/syn/S12.html#Construc...ialization says if you want to say initialize some variables right? like my class has an attribute $!values, i can use BUILD($values) { $!values = $values; } to do that 00:07
it also says if you do Foo.new(5); it should be passing the 5 along to BUILD
in ng, the paramaters passed to new aren't getting passed to BUILD like they are supposed to be 00:08
colomon The error I get is "Unexpected named parameter 'b' passed" -- is that what you're seeing? 00:11
ash_ yeah 00:12
zamolxes frettled: right :) it's 2 AM here as well, but we'll get in touch in a few hours 00:13
ash_ ng: class Foo { submethod BUILD($a) { say $a }; }; my $f = Foo.new(5); 00:14
p6eval ng b65331: Not enough positional parameters passed; got 1 but expected 2␤current instr.: 'perl6;Foo;BUILD' pc 266 (EVAL_1:115)␤
ash_ colomon: thats a simple example of what I mean
that should say '5'
colomon ash_: oh, I wonder if :a(5) works at all.... 00:17
ash_ ng: class Foo { submethod BUILD(:$a) { say $a }; }; my $f = Foo.new(:a(5));
p6eval ng b65331: sh: ./perl6: No such file or directory␤
ash_ hmm
that worked locally
colomon yeah, it looks like b => 5 gets the same result. 00:18
zamolxes frettled: dug this out of my bash history : tail -n 10000 /var/log/apache2/access.log | perl -nE '($url, $s, $d) = m{"[A-Z]+ (/[^/ ]*)[/ ][^"]+" \d+ (\d+) (\d+) "}; $urls{"$url"}+=$d;$dt+=$d; END {print for map { $d =int($urls{$_}/1000000);$p =int($urls{$_}*100/$dt); "$d $p% $_\n"} sort { $urls{$a}<=>$urls{$b} }keys %urls;}'
it aggregates resource consumption per first token from the URL 00:19
justatheory TimToady, chromatic: What, has Twitter replaced IRC? 01:04
PerlJam justatheory: I think this thing called "life" may have replaced IRC 01:08
diakopter PerlJam: I suspect justatheory is referring to chromatic_x's and TimToady's conversation on twitter this afternoon 01:09
justatheory corect 01:10
cognominal ha, ha, as often, @chromatic_x shoot before aiming 01:11
tideflat perl6: say 3 01:26
p6eval elf 29531, pugs, rakudo 3867ff: 3␤
tideflat perl6: say 1 + 1
p6eval elf 29531, pugs, rakudo 3867ff: 2␤
tideflat perl6:my $a = 10;say a
perl6:my $a = 10;say $a 01:27
perl6: my $a = 10;say $a
p6eval elf 29531, pugs, rakudo 3867ff: 10␤
tideflat perl6: for $i in 1..1 say $a 01:28
p6eval pugs: *** ␤ Unexpected " in"␤ expecting "::"␤ Variable "$i" requires predeclaration or explicit package name␤ at /tmp/gBzz33TGX6 line 1, column 7␤
..rakudo 3867ff: Confused at line 2, near "in 1..1 sa"␤in Main (file <unknown>, line <unknown>)␤
..elf 29531: Parse error in: /tmp/gaFeglxI4x␤panic at line 1 column 0 (pos 0): Can't understand next input--giving up␤WHERE: for $i in 1..1 say $a␤WHERE:/\<-- HERE␤ STD_red/prelude.rb:99:in `panic'␤ STD_red/std.rb:76:in `scan_unitstopper'␤ STD_red/std.rb:224:in `comp_unit'␤
..STD_red/std.rb:210:in `…
tideflat perl6: for $i in 1..1 say $i 01:29
p6eval pugs: *** ␤ Unexpected " in"␤ expecting "::"␤ Variable "$i" requires predeclaration or explicit package name␤ at /tmp/gvTTB8mbAm line 1, column 7␤
..rakudo 3867ff: Confused at line 2, near "in 1..1 sa"␤in Main (file <unknown>, line <unknown>)␤
..elf 29531: Parse error in: /tmp/xYm3D2KzlO␤panic at line 1 column 0 (pos 0): Can't understand next input--giving up␤WHERE: for $i in 1..1 say $i␤WHERE:/\<-- HERE␤ STD_red/prelude.rb:99:in `panic'␤ STD_red/std.rb:76:in `scan_unitstopper'␤ STD_red/std.rb:224:in `comp_unit'␤
..STD_red/std.rb:210:in `…
ash_ rakudo: .say for 1..3;
p6eval rakudo 3867ff: 1␤2␤3␤ 01:30
ash_ is that what you meant to do tideflat?
tideflat yes
Can i just play with perl6 here?
ash_ sure, or you can message p6eval 01:31
if you don't want everyone to see it
tideflat OK
rakudo: .say for 1..3;
p6eval rakudo 3867ff: 1␤2␤3␤
tideflat rakudo: .say for 1..10 * 2;
p6eval rakudo 3867ff: 1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤11␤12␤13␤14␤15␤16␤17␤18␤19␤20␤
tideflat rakudo: say $_ * 2 for 1..10; 01:33
p6eval rakudo 3867ff: 2␤4␤6␤8␤10␤12␤14␤16␤18␤20␤
tideflat rakudo: say $_ ** 2 for 1..10;
p6eval rakudo 3867ff: 1␤4␤9␤16␤25␤36␤49␤64␤81␤100␤
tideflat rakudo: say $_ ** -2 for 1..10;
p6eval rakudo 3867ff: 1␤0␤0␤0␤0␤0␤0␤0␤0␤0␤ 01:34
tideflat $_ * $_ for 1..10;
PerlJam rakudo: say 4 ** -2
p6eval rakudo 3867ff: 0␤
PerlJam rakudo: say 4.0 ** -2
p6eval rakudo 3867ff: 0.0625␤
PerlJam rakudo: say 4 ** -2.0
p6eval rakudo 3867ff: 0.0625␤
PerlJam That's got to be a bug 01:35
tideflat rakudo: say $_ * $_ for 1..10;
p6eval rakudo 3867ff: 1␤4␤9␤16␤25␤36␤49␤64␤81␤100␤
tideflat rakudo: say $_ ** $_ for 1..10;
p6eval rakudo 3867ff: 1␤4␤27␤256␤3125␤46656␤823543␤16777216␤387420489␤10000000000␤
tideflat rakudo: say "\$_ is $_" for 1..10; 01:36
p6eval rakudo 3867ff: $_ is 1␤$_ is 2␤$_ is 3␤$_ is 4␤$_ is 5␤$_ is 6␤$_ is 7␤$_ is 8␤$_ is 9␤$_ is 10␤
tideflat rakudo: say [+] 1..10; 01:39
p6eval rakudo 3867ff: 55␤
tideflat rakudo: say [+] 1..50; 01:40
p6eval rakudo 3867ff: 1275␤
tideflat rakudo: say [>] 1..50;
p6eval rakudo 3867ff: 0␤
tideflat rakudo: say [<] 1..50;
p6eval rakudo 3867ff: 1␤ 01:41
tideflat rakudo: say [=<] 1..50;
p6eval rakudo 3867ff: say requires an argument at line 2, near " [=<] 1..5"␤in Main (file src/gen_setting.pm, line 2593)␤
tideflat rakudo: say [<=] 1..50;
p6eval rakudo 3867ff: 1␤
tideflat rakudo: say [>=] 1..50;
p6eval rakudo 3867ff: 0␤
tideflat rakudo: say [*] 1..5; 01:42
p6eval rakudo 3867ff: 120␤
tideflat say 1!; 01:43
rakudo: say 1!
p6eval rakudo 3867ff: Confused at line 2, near "!"␤in Main (file <unknown>, line <unknown>)␤
tideflat rakudo: say (3,3,3) <<+>> (3,3,3) 01:44
p6eval rakudo 3867ff: 666␤
diakopter tideflat: you can msg p6eval directly
tideflat I know. do you want me to?
PerlJam tideflat: sub postfix:<!>($n) { [*] 1..$n }; say 5!; # for your factorial 01:45
diakopter well, just remember there are a bunch of folks who read the irclogs (in their client or on the web)
tideflat oh, i will msg p6eval directly from now on 01:46
colomon tideflat: if you find something that puzzles you, please feel free to send it to p6eval here in #perl6. :) 01:48
tideflat Ok 01:49
drbean doraku.asahi.com/ "An exclusive Asahi Newspaper entertainment site for the Beatles generation" 02:22
doraku = Do 楽 02:23
pugs_svn r29532 | diakopter++ | [sprixel] add action() combinator (inject arbitrary JS code into the parser function on node success).\ 02:44
r29532 | Add notes about plans for an analysis phase for Boyer-Moore & JS literal RegExp (fastest by far on \
r29532 | the shootout language game) substitution.
diakopter ENOPM 02:54
Infinoid ohnoes 02:57
tideflat /msg p6eval perl6: for ['cats','dogs','fish'] {say "$_ are the best amimal"} 03:05
diakopter pmi<tab>: do nqp-rx's grammars run the Actions during the parse (so potentially multiple times incl backtracking), or after the parse once? 03:06
or anyone else who might know.
tideflat perl6: my $a = 'abe';for 1..10 {say $a;$a ~= 'abe'} 03:28
p6eval pugs, rakudo 3867ff: abe␤abeabe␤abeabeabe␤abeabeabeabe␤abeabeabeabeabe␤abeabeabeabeabeabe␤abeabeabeabeabeabeabe␤abeabeabeabeabeabeabeabe␤abeabeabeabeabeabeabeabeabe␤abeabeabeabeabeabeabeabeabeabe␤
..elf 29532: Unknown rule: infix_postfix_meta_operator:=␤It needs to be added to ast_handlers.␤ at ./elf_h line 2850␤
draegtun Morning all. Has promised I posted my perl6 followup (last night): transfixedbutnotdead.com/2010/01/14...ogramming/ 08:59
Su-Shee draegtun: I already read it before breakfast!
draegtun Su-Shee: Hope if didn't upset the breakfast :-) 09:00
Su-Shee no, not at all. 09:02
though every time I read something like this I keep thinking how much a good book about contemporary oo mixed with meta programming NOT based on Java is needed... 09:03
draegtun Su-Shee: I agree. Perl needs a good meta programming book methinks 09:04
mathw draegtun: thanks for that, I am reading now...
draegtun mathw: Thanks for reading, hope you enjoy 09:05
Su-Shee draegtun: I went through the shelf with "OO" recently .. all java. annoying.
mathw draegtun: Mmm nice. I did find your consistent use of {} for interpolation interesting. 09:06
draegtun mathw: thanks. I think I was just keeping it consistent to original Ruby/Javascript examples 09:08
sjohnson hi matt! 09:09
mathw draegtun: It might lead a few people to think it's compulsory even for simple scalars, but maybe I'm just being twitchy and paranoid 09:11
sjohnson: hi
draegtun: They can always be informed otherwise
draegtun draegtun: I see... I thought {} were compulsory ;-) 09:13
mathw: I see.. I thought {} were compulsory (oops!)
mathw no, ordinary scalars interpolate as in Perl 5 09:15
draegtun mathw: cool. Obviously reading an old doc again :( 09:16
mathw {} is for more complex things like method calls and arrays
I don't think they were ever compulsory 09:17
Unless they were for a time in the period where I wasn't paying attention
draegtun mathw: Well everything on the web isn't always true :) 09:37
mathw: I've amended my gist accordingly: gist.github.com/276591
mathw: only place I needed the {} was on "{$.name}'s colour is $colour_name " (same reason as in Perl5)
mathw: I'll add a comment to my blog post referring to this gist
moritz_ good morning 09:45
mathw draegtun: Great stuff 09:57
moritz_ aye, draegtun++ 09:58
meneldor Hi all 10:11
Supaplex all is on vacation 10:13
pugs_svn r29533 | mberends++ | [vill] Code generator groundwork for 'my $foo="bar";'. Various parts are now awaiting a formal runtime metamodel...
takadonet morning all 13:34
colomon morning! 13:37
takadonet colomon: alive and well sir? 13:38
colomon not yet slain by this cold, anyway. :) 13:39
jnthn eww cold :-( 13:40
jnthn Get well soon. 13:40
jnthn 's is better today than yesterday.
colomon It's been a whole week now, so I think it's reached the point where I have to pretend it's better and just get on with life. 13:41
Is rakudo supposed to have a release next week? 13:43
jnthn I guess so. 13:45
colomon eeek! 13:46
ash_ rakudo: my $a = List.new(1, 2, 3); say $a.perl; $a = list(1, 2, 3); say $a.perl; 15:04
p6eval rakudo 3867ff: []␤[1, 2, 3]␤ 15:05
ash_ should those be the same?
ash_ rakudo: my $a = Array.new(1, 2, 3); say $a.perl; 15:29
p6eval rakudo 3867ff: []␤
ash_ i am begining to wonder what the "right" functionality of .new on List/Array is...
colomon ash_: I would think those should be the same. 15:30
I doubt it is spec'd one way or another...
ash_ yeah, digging in S32-Containers i haven't seen any specs on .new for List/Array
colomon This is where we get to use our judgement and influence the future. :)
ash_ rakudo: class Foo { submethod BUILD(*@args) { say @args.perl; } }; Foo.new(1, 2, 3); 15:32
slurpies in BUILD seem to function funny
ng: class Foo { submethod BUILD(*@args) { say @args.perl; } }; Foo.new(1, 2, 3)
p6eval rakudo 3867ff: ( no output )
ng b65331: Method 'perl' not found for invocant of class 'Undef'␤current instr.: 'perl6;Array;_block3582' pc 240746 (src/gen/core.pir:20091)␤
ash_ i guess that's the next thing, should a slupry in a BUILD work? 15:33
i haven't seen that in the specs either
jnthn ash_: No, you probably want to do something like (if you want List.new(1,2,3) to be like list(1,2,3) 15:34
method new(*@values) { self.bless(values => @values) }; submethod BUILD(:@values) { ... } 15:35
That is, transform the positionals into a single named arg that BUILD will receive.
ash_ hm, k 15:36
PerlJam but should List.new and Array.new be that way out of the box?
jnthn PerlJam: In what sense? 15:37
YOu mean, should they be spec'd to work like that?
PerlJam aye
ash_ ng: augment class List { method new(*@args) { say 'me'; } }; List.new(1, 2, 3) # any reason that's not dispatching to new(*@args)?
p6eval ng b65331: ( no output )
colomon ash_: should be multi method, yes? But I remember that not working yesterday for me... 15:38
jnthn ash_: No, that's...odd. 15:39
ash_ ng: augment class List { multi method new(*@args) { say 'me'; } }; List.new(1, 2, 3)
p6eval ng b65331: ( no output )
ash_ shrug *out of my knowledge of rakudo now*
colomon
.oO(hope jnthn has a clue, 'coz I'm stumped...)
 15:40
PerlJam Can you turn a multi into an only via augment? 15:41
(sure it'll cause breakage, but can it be done?)
ash_ ng: augment class List { method new(*%kwargs, *@args) { say 'me'; } }; List.new(1, 2, 3) # adding a named slurpy doesn't help, i tried it both as the first and second argument too 15:46
p6eval ng b65331: ( no output )
jnthn has to go out for meeting with freinds now...back later... 15:47
ash_ cya jnthn
commuting ^ 15:50
s/\^/&/
TimToady PerlJam: functions are not search for in packages (except maybe as a last resort); you may put an only in an inner scope that hids outer multies 16:44
*searched
(and by putting CORE as an outer lexical scope, this is how we guarantee that a user's function defs are future proof against new keywords) 16:46
TimToady *hides 17:45
diakopter eh
TimToady (correcting a typo from an hour ago) 17:46
takadonet ... 17:47
diakopter '*hides' was correcting the typo
TimToady ablutions & 17:48
pugs_svn r29534 | kyle++ | [t/spec] Test for RT 72084: sub name beginning with last- 17:57
r29535 | kyle++ | [t/spec] Test for RT 72048: !~~ of role on instance is always true
Tene what's the motivation for asking "Was I called as a sub or as a method"? I'm missing something. 18:04
PerlJam Tene: me too 18:24
(at least I don't understand it in a Perl 6 context) 18:25
TimToady he's using it to solve a P5 namespace collision that P6 doesn't have 18:26
(hence my remark yesterday about "namespace superstition")
PerlJam I guess I was puzzled by "...I worry that the method/sub invocation isomorphism will be a trap just as in Perl 5." 18:29
ruoso .oO( IRC is always more productive than twitter ) 18:30
petsepes hi 18:30
ruoso If someone really needs the distinction, he can install a different routine in the sub
PerlJam petsepes: greetings
TimToady it's a trap in P5 because P5 conflates the method and subroutine namespaces 18:32
TimToady P6 doesn't, so it's not a trap unless you write a custom dispatcher that confuses things 18:33
diakopter or a compiler/runtime that confuses things.
note the lack of closing </snark> tag 18:34
PerlJam TimToady: right, I was confused by the worry, because it seems you have to do it to yourself
(in perl 6)
TimToady the upside of the isomorphism is that a dispatcher never has to worry about what kind of thing it's calling, because dispatchers only ever call functions, even when that function is implementing a method 18:39
which makes it easy to write dispatcher that shoots yourself in the foot if you want to :)
ruoso and methods are not our by default anymore, right?
TimToady but also allows us to export methods without writing wrappers
correct
if you *do* declare it our, you may call it as a subroutine 18:40
(or my)
ruoso I think the issue is pretty much addressed...
PerlJam It's only a real problem if you're coming from a Perl 5 background and expect that you have do Perl 5ish things :)
TimToady the downside from the VM perspective is that if your VM compartmentalizes, you end up writing wrappers anyway 18:41
ruoso and then the question is not "was this called as a method or as a sub", but "which was the dispatcher that called this?" since we could eventually have some other type of dispatcher
TimToady or you just don't rely on the VM's dispatcher from the user level
since it probably doesn't implement nextsame anyway... 18:42
from the P6 perspective, it's a no-brainer to say that underneath, everything gets invoked as a function
ruoso but I think that someone asking "which dispatcher was used here" can deal with implementation-specific stuff
TimToady well, a VM is not guaranteed to be able to tell you which dispatcher was used either, for that matter 18:43
ruoso it needs to some extent if it supports nextsame 18:44
TimToady we want to support custom dispatchers in any case; use of a specialized VM dispatcher would be an optimization from the general case 18:45
in some cases a very good optimization 18:46
but not at the expense of broken semantics
(unless you predeclare broken semantics :)
PerlJam a custom dispatcher seems like an "all bets are off" kind of thing. If you're worried about shooting yourself in the foot, you've only got yourself to blame because the default dispatcher(s) protect you. 18:48
Is that an accurate statement?
(with some wiggle room on "protect" :)
TimToady seems about right 18:50
but people coming from the world of P5 programming have a peculiar set of phobias which will take time to cure
ash_ rakudo: my $a = List.new(1, 2, 3); say $a.perl; my $b = list(1, 2, 3); say $b.perl; say $a == $b; # should this be true? 18:51
p6eval rakudo 3867ff: []␤[1, 2, 3]␤0␤
TimToady a great many Perl 5 "Best Practices" are really workarounds to design flaws, but prescriptive grammarians will continue to try to apply those workarounds after they are no longer needed. 18:56
PerlJam That first clause seems like a really good answer to why Perl 6 is so different from Perl 5 18:57
diakopter who really needs vowel points anyway
.oO( writing emitters specialized to particular permutations of its variations should eventually be resolved back to writing one emitter with lots of checks per instruction to-be-emitted. longer compilation time, but much more composably optimizable. )Oo.
 19:03
tolkad perl6: print "test"; 19:16
p6eval elf 29535, pugs, rakudo 3867ff: test
tolkad !help
tell lambdabot @echo 19:19
@echo
ooh nvm
ash_ so, according the dispatch mechanism, if there are 2 methods, 1 with foo(*@args); and another with foo(*@args, *%kwargs); and you call foo(1, 2, 3); that should dispatch to the one only with the arg's slurpy, not kwargs, right? 19:24
TimToady it's arguable whether a (nilable) slurpy should be considered an extra constraint, but we really need an opinion from the implementors whether it's practical to make the distinction 19:30
the fact that theres another declaration effectively turns the * into a +, by that reckoning
which seems a bit bogus 19:31
why not just write the one method to handle the Nil case
ash_ so, 1 with foo ( ) # no params? you mean? 19:33
ash_ I guess i am trying to figure out why this isn't working: 19:36
ng: augment class List { multi method new(*@args) { say @args; } }; class Foo { multi method new(*@args) {say @args;}}; List.new(1, 2, 3); Foo.new(1, 2, 3);
p6eval ng b65331: 123␤
ash_ it should of printed 123 twice, not once 19:37
hmmm
might be a bug in the dispatcher, rather than an issue with the dispatching strategy
I thought that might of been related to the fact that Mu has a new(*@args, *%kwargs) but it doesn't look like thats the problem 19:39
PerlJam RaceCondition: greetings 20:06
RaceCondition PerlJam: hi again
quiet here 20:07
PerlJam RaceCondition: it's you imagination :)
(lots of out-of-#perl6-band communication happens too)
RaceCondition: What's your interest in Perl 6? Have you tried Rakudo yet? Or was it more just curiousity about the state of things? 20:08
RaceCondition PerlJam: curiosity, yeah... I don't have the need and time to learn perl 20:08
PerlJam What languages are you already familiar with? 20:09
RaceCondition well most of the mainstream ones, but I've peeked into many
ash_ i personally like multi-methods and roles in perl 6, not many languages have either of those, and even fewer have both 20:10
that and you can define your own operators, that ones kinda cool 20:11
rakudo: sub postfix:<!>(Int $x) { [*] 1..$x }; say 5!; 20:12
p6eval rakudo 3867ff: 120␤
ash_ moritz_++ for that one
KyleHa I like type constraints. 20:14
KyleHa rakudo: sub postfix:<!>(Int $x) { [*] 1..$x }; say "five"! 20:15
p6eval rakudo 3867ff: Nominal type check failed for parameter '$x'; expected Int but got Str instead␤in Main (file src/gen_setting.pm, line 324)␤
ash_ rakudo: multi sub postfix:<!>(Int $x) { [*] 1..$x }; mutli sub postfix:<!>(Str $x) { $x.uc }; say "five"!; say 5!; 20:16
p6eval rakudo 3867ff: Could not find non-existent sub mutli␤in Main (file src/gen_setting.pm, line 324)␤
ash_ oops
rakudo: multi sub postfix:<!>(Int $x) { [*] 1..$x }; multi sub postfix:<!>(Str $x) { $x.uc }; say "five"!; say 5!; 20:18
p6eval rakudo 3867ff: FIVE␤120␤
ash_ silly spelling
KyleHa How hard is it to get Perl 6 to recognize 'mutli' as 'multi'? 20:19
ash_ macro mutli { 'multi' } ? 20:20
KyleHa Macros make me think of TeX.
ash_ they make me think #define
lichtkind can anybody here answer to me some things about proto regexen? 20:24
ash_ sure, we can try 20:25
RaceCondition how does Perl 6 compare with Lisp? I see it has macros and multimethods 20:40
ash_ in what respects? would you like to compare them in? 20:46
i think perl 6 has less ( )
well most of the time :P 20:47
KyleHa Perl 6 has more funny characters for some definitions of 'funny'.
ash_ perl 6's syntax doesn't fit on a 2"x"4 note card 20:48
i think
well maybe if you had a really good printer
RaceCondition so why does Lisp's syntax fit?
TimToady doesn't have any :)
RaceCondition oh, in that sense 20:49
ash_ lol lisp's syntax is very simple
RaceCondition I thought you mean "lisp code doesn't fit on a ..."
TimToady to the first approximation, Perl 6 is Lisp with a decent syntax, and minus a lot of assumptions about how lists should be represented :) 20:50
but that's also a vast oversimplification... 20:51
KyleHa I thought LISP was a vast oversimplification. 20:52
TimToady nevertheless, despite all the trappings of OO, Perl 6 is fundamentally more in the FP camp
diakopter ENOPMSTILL 20:54
diakopter does anyone know when the Actions in an nqp-rx grammar are run? 20:54
TimToady I believe they are run at reduction time, but note that Perl 6 grammar isn't supposed to backtrack much. 20:56
KyleHa Is moritz_ around today? I'd like to reorganize (or disorganize) S03-operators/smartmatch.t, but I don't want to overcrud the spectests. 20:57
TimToady there are still only two spots that we rely heavily on backtracking, and neither of them is likely to prove a performance problem.
KyleHa @seen moritz_ 20:58
TimToady rakudo: say [+42].perl 20:59
p6eval rakudo 3867ff: [42]␤
TimToady there's one of 'em
rakudo: $_ = "foo"; say "$_.notamethod" 21:00
p6eval rakudo 3867ff: foo.notamethod␤
TimToady there's the other
ash_ in the second case, is that backtracking to find the most recent defintiion of $_ ? 21:01
TimToady no
it's distinguishing from this:
rakudo: $_ = "foo"; say "$_.notamethod()"
p6eval rakudo 3867ff: foo.notamethod()␤
TimToady well, it should be :)
ash_ oh, okay, so it has to string interpolation? 21:02
TimToady finding the end of the interpolation involves looking for trailing brackets of some kind or other
and not just the first brackets, but the last brackets 21:03
"$foo.bar().baz()"
or a more likely case, "$x[$y]{$z}" 21:04
sjohnson go to baktiz!!
frettled oh! 21:09
uniejo Is the interpolation length also depending on does Callable, Positional and does Associative ? 21:13
KyleHa Is dev.pugscode.org relocated or just out of the picture? 21:16
frettled KyleHa: if you're missing Trac, see the top post at howcaniexplainthis.blogspot.com :) 21:21
KyleHa frettled: Thanks, that's just what I was missing.
frettled Someone™ just needs to get enough tuits to fix the spam problem. 21:23
KyleHa The first thing that pops into my mind is to block Google via robots.txt. 21:32
The spammers are submitting bogus tickets, which I don't think anyone legitimately uses anyway, so lets amputate that. 21:34
PerlJam Starting with a new VM is probably the most important thing IMHO
frettled mm
PerlJam I'm a fan of clean slates where I can get them.
diakopter KyleHa: do you know anyone who was using dev.pugscode.org? (and if you were, what were you using on it?) 21:35
KyleHa diakopter: I looked at the timeline to see recent commits to the pugs repo. From there I'd look at diffs, etc. 21:36
diakopter oh
KyleHa That's how I kept up with spectest activity, back when I was keeping up. It's been a while since I looked.
diakopter KyleHa: may I interest you in this instead? perlcabal.org/svn/pugs/log
its rss feed has been <ahem> customized for use by pugs_svn 21:37
KyleHa: that SVN::Web installation is also on feather. 21:38
also, it's the one to which the irclog web ui constructs links 21:39
KyleHa Oh, very nice. Thank you, diakopter++ 21:41
ash_ rakudo: class List is also { method new(*@args) { say 'hi'} }; List.new(1, 2, 3); 21:42
p6eval rakudo 3867ff: ( no output )
diakopter fights cisco anyconnect
pugs_svn r29536 | kyle++ | [t/spec] split S03-operators/smartmatch.t into S03-smartmatch/*.t 21:44
KyleHa OK, I'm going for forgiveness over permission.
ash_ i think there is something wrong with how 'new' is getting dispatched 21:45
in both rakudo and ng
PerlJam rakudo: class Foo is List { method new(*@args) { say "hi" } }; Foo.new(1,2,3); 21:46
p6eval rakudo 3867ff: ( no output ) 21:47
PerlJam blame jnthn :)
ash_ maybe its augmenting/is also 's fault
PerlJam ash_: Well, I just tried inheritance and the result was the same. 21:48
although that worked fine on my local perl6 (not sure if it's ng or master since I haven't compiled it in a while) 21:50
dalek kudo/ng: a57ebba | (Kyle Hasselbacher)++ | t/spectest.data:
[spectest.data] update according to pugs r29536
 22:00
ash_ PerlJam: if you use is also its master
but also rakudo in here times out if you define a class 22:01
it seems like its okay if you don't define a new method in an augment or in an 'is also'
PerlJam ash_: I meant the inheritance example ... but that's a good litmus test to find out which version I have :)
ash_ i wonder why rakudo: still time's out on class definitinos 22:07
ash_ Mu == undef? right? 23:00
frettled not quite 23:02
But it's one of the undef-like values, depending on type.
TimToady arguably it's the most undefined value 23:14
insofar as all other undefined values derive from it
sjohnson TimToady: those two lines u wrote might end up as messages in an irc bottle 23:25
ash_ colomon___: ping? 23:31
colomon___ ash_: pong? 23:37
ash_ I fixed List.new, i have patch for it
colomon___ \o/ 23:38
ash_ i figured i could bug you instead of jnthn, also he's not here :P
colomon___ sure
ash_ gist.github.com/278521 is a copy of it, or i can email it to someone 23:39
colomon___ I will be slightly erratic for a few (reading to son)... 23:39
ash_ no worries
colomon___ Are there tests? 23:45
:)
ash_ not yet, i'll do those next
List.new(1, 2, 3) should ~~ list(1, 2, 3) 23:46
I'll work on some though
colomon___ m u 23:48
colomon___ yay, works for Array too! 23:58
ash_ should work for anything that inherits from List 23:59
so apparently augment and 'is also' can't add 'new' to a class
ng: class Foo { }; augment class Foo { method new(*@args) { say 'hi'} }; Foo.new(1, 2);
p6eval ng a57ebb: ( no output )

Powered by Cro and the Raku® Programming Language.

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

Thank you!