pugscode.org/ | nopaste: sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by diakopter on 11 July 2007.
gaal moose 05:55
bsb gaal, moose 06:58
bsb & 07:06
spinclad good moosetime(), all 07:42
fglock kp6 works fine with perl5.10.0-RC2 :) 12:02
elmex hmm, whats up with perl6 at the moment? is there still lots of work going into PUGS? is there an efford for the final implementation ontop of parrot? 12:21
moritz_ elmex: currently the most effort is on kp6 and parrot 12:42
elmex kp6? oh 12:43
ah, ok, got it, thanks for the info 12:45
moritz_ nqp: say("hi"); 13:05
exp_evalbot OUTPUT[src/pmc_freeze.c:1254: failed assertion 'must_have_seen'␤Backtrace - Obtained 19 stack frames (max trace depth is 32).␤ Parrot_print_backtrace␤ Parrot_confess␤ (unknown)␤ (unknown)␤ (unknown)␤ (unknown)␤ Parrot_thaw␤
..PackFile_Constant_unpack_pmc␤ PackFile_Con...
moritz_ woops
pugs_svn r19007 | fglock++ | [kp6] - remove Data::Dump::Streamer dependency; 13:47
r19007 | fglock++ | - moved t/kp6/internals to t/kp6-internals - Coro.pm is now really optional
pugs_svn r19008 | fglock++ | [kp6-kp6] updated t-bootstrap/02-grammar.t 13:57
moritz_ fglock: should t-bootstrap/02-grammar.t pass? 14:01
fglock moritz_: not yet, I'm adding failing tests as the previous ones get fixed 14:02
moritz_ fglock: ok... only the first test passes for me 14:05
pugs_svn r19009 | fglock++ | [kp6-kp6] one more test passes, but then "Segmentation fault" 14:11
fglock moritz_: I'm now testing with both 5.8.8 and 5.10.0rc2
5.10 has better error handling: "EVAL without pos change exceeded limit in regex at compiled/perl5-kp6-kp6/lib/KindaPerl6/Grammar/Term.pm line 890." 14:12
pugs_svn r19010 | fglock++ | [kp6-perl5] use an alternate hack for shortcircuiting 16:08
pmurias fglock: is it easier for you to have the 02-grammar in p5 than in p6? 16:47
fglock pmurias: yes, because I need it to be executed by kp6-kp6, which can't compile yet 16:48
pmurias it needs to be executed by kp6 16:49
s/needs/can
s/executed/compiled/ 16:50
see t-bootstrap/02-grammar.p6
fglock hmm - that's right
pmurias by easier i meant more convenient (i'm considering porting the changes to 02-grammar.p6) 16:51
fglock trying out... 16:52
fglock pmurias: yes, that works :) 16:55
pmurias port the changes? 16:56
jwest-- is there any real use of asm anymore?
moritz_ jwest--: are you talking about parrot assembler? 16:57
I think it's mostly replaced by PIR these days
jwest-- not parrot 16:58
just asm in general
inline asm
moritz_ in what?
the linux kernel uses inline ASM 16:59
jwest-- a lot?
moritz_ jwest--: not really
fglock pmurias: sure 17:00
moritz_ a far friend of mine wrote a chess program, he used inline asm for some heavily used code
jwest-- whycouldnt he have done that in c 17:01
or lisp or something
pugs_svn r19011 | fglock++ | [kp6-perl5rx] minor fix in Match.perl
moritz_ jwest--: because asm can be faster, in some cases 17:02
the two reasons for inline assembler are 1) speed 2) low level access
pmurias one probably also uses assembler for writing shellcode 17:03
jwest-- i dont mind seeing a lot of assembly code 17:05
really
fglock pmurias: nevermind updating, I'll just use 02-grammar.p6 instead
jwest-- just that i dont see it
pmurias fglock: updated 17:09
pugs_svn r19012 | pmurias++ | [kp6] brought 02-grammar.p6 up to date with the p5 version
r19012 | pmurias++ | cosmetic change in Test
pmurias jwest--: you can always ask gcc to supply you with some/lots of with an appropriate compiler flag 17:10
jwest-- heh 17:11
pugs_svn r19013 | fglock++ | [kp6-kp6] added instructions to t-bootstrap/02-grammar.p6; removed Perl 5 version
dlocaus cries 17:13
my ubuntu boot disk crashed.
pmurias dlocaus: got to use windows now? 17:15
dlocaus I have 2 machines.
I'm on the windows machine right now, the linux box is still searching for "harddisks and other hardware"
I see the HD light go on once in a while though. 17:16
pmurias fglock: it might be good to remove the "eval '$var'" odity 17:26
fglock pmurias: I don't think it works without that
it interpolates too early
pmurias fglock: i don't think the eval can help anything 17:33
it might me able to hide errors
if an undeclared lexical variable is used
giving a Segfault here 17:35
fglock fixed :)
pugs_svn r19014 | fglock++ | [kp6-perl5rx] fixed a segfault (uninitialized sub-regex) 17:36
moritz_ are BEGIN blocks in kp6 really executed at compile time? 17:37
pmurias moritz_: yes
at parse time exactly 17:38
moritz_ so a program 'BEGIN { say "foo" }' prints something at compile time, and does nothing at execution time?
that's cool, actually ;)
pugs_svn r19015 | fglock++ | [kp6-kp6] t-bootstrap/02-grammar.p6 - fixed test #4 17:40
gaal BEGIN { $compilation_time = time }, yes. 17:41
pmurias fglock: i think the eval can only introduce bugs by hiding the undeclared variable error 17:42
pmurias s/introduce/hide/ 17:44
fglock pmurias: can you test without the eval? there are some old regex-recursion tests in misc/
not sure if they still run
pmurias fglock: p6 tests? 17:45
gaal or should I say, my $parsed_at = BEGIN { time };
pugs_svn r19016 | fglock++ | [kp6] warn() returns a Perl 6 value 17:46
fglock moritz_: $ perl script/kp6 -Cperl5 -e ' BEGIN { warn "foo" } ' > compiled.pl
moritz_ fglock: nice ;) 17:48
fglock pmurias: misc/token-test-recurse2.pl and others, it's assorted Perl 6 or Perl 5
pugs_svn r19017 | fglock++ | [kp6-kp6] one more test 17:53
r19018 | pmurias++ | [kp6] the eval makes a difference here dunno why 18:00
pmurias fglock: please check the file i commited 18:01
fglock k
pmurias eval '$var' and $var behave differently
fglock I don't know - how about asking in the p5p irc? 18:05
I should add more "why" docs in my code 18:07
moritz_ well, eval '$var' returns the content of $var, while eval $var evals it, and returns the result
or did I miss something? 18:08
pmurias moritz_: it's eval '$var' versus $var
moritz_ pmurias: uhm, ok
context?
maybe eval '$var' always forces scalar context
pmurias eval '$var' is undef the first time 18:11
[particle] pugs: say +"123.45"
exp_evalbot OUTPUT[123.45␤]
pmurias fglock: the #p5p channel? 18:12
fglock yes
pmurias i'm to tired now (long day of school)
visiting a friend&
fglock pmurias: np :)
[particle] pugs: @a = (1,2,3); say +@a; 18:13
exp_evalbot OUTPUT[*** ␤ Unexpected " ="␤ expecting "::"␤ Variable "@a" requires predeclaration or explicit package name␤ at /tmp/orYlEN22Wo line 1, column 3␤]
[particle] pugs: my @a = (1,2,3); say +@a; 18:14
exp_evalbot OUTPUT[3␤]
pugs_svn r19019 | fglock++ | [kp6-kp6] fixed Match lookup 18:25
pugs_svn r19020 | fglock++ | [kp6-kp6] t-bootstrap/02-grammar.p6 - one more test 18:27
pugs_svn r19021 | pmurias++ | [kp6] test descriptions 18:42
pugs_svn r19022 | fglock++ | [kp6-kp6] fixed Class name conflicts 18:54
pugs_svn r19023 | fglock++ | [kp6] t-bootstrap/02-grammar.p6 added failing test 19:14
fglock & 19:27
[particle] | 19:28
dlocaus my (linux) HD died, I'll be out for a while... /cries 20:21
dlocaus waves 20:22
mncharity pugs: my B::x = 3; 23:13
exp_evalbot OUTPUT[*** ␤ Unexpected "="␤ expecting ":" or "("␤ at /tmp/UFCQiZVg9H line 1, column 9␤] 23:14
mncharity I couldn't remember, and was randomly curious, whether one can lexically scope variable names in non-local packages. 23:15
pugs: package A {} my A::x = 3; 23:16
exp_evalbot OUTPUT[*** ␤ Unexpected "="␤ expecting ":" or "("␤ at /tmp/8TvG9zWneX line 1, column 22␤]
mncharity pugs: package A {}; my A::x = 3;
exp_evalbot OUTPUT[*** ␤ Unexpected "="␤ expecting ":" or "("␤ at /tmp/yXn4aIuKUc line 1, column 23␤]
moritz_ mncharity: it's "module" in p6, not "package" (iirc) 23:17
pugs: module a {}; my $A::x=3; say $A::x; 23:18
exp_evalbot OUTPUT[*** ␤ Unexpected "::x"␤ expecting word character, "?", "!", trait, "=", infix assignment, term postfix or operator␤ at /tmp/YTMLdVNNrD line 1, column 19␤]
moritz_ pugs: module a { my $x}; say $a::x;
exp_evalbot OUTPUT[␤]
moritz_ pugs: module a { my $x = 3 }; say $a::x;
exp_evalbot OUTPUT[␤]
TimToady pugs: module a { our $x = 3 }; say $a::x 23:19
exp_evalbot OUTPUT[␤]
mncharity I was thinking about namespaces, and wondered whether pugs could have an exposed api, where variable lookup is handled by a macro P6MOP::lookup(name) -> <whatever it is names are bound to>. Then users could lexically override name resolution.
moritz_ that would allow all sort of hacks ;) 23:20
mncharity indeed :D
TimToady see S10:84 23:21
mncharity looking...
is there a copy of S10 somewhere with line numbers? :) 23:22
moritz_ mncharity: take a look at irc.pugscode.org 23:24
mncharity: there the S10:84 directly links to the correct line ;)
thanks to agentzh++ who implemented the line anchors in the HTML files 23:25
mncharity ooo, shiny 23:26
[particle] pugs: say 7 +& +^1; 23:32
exp_evalbot OUTPUT[6␤]
moritz_ that syntax looks plain scary 23:35
probably because the operation is scary as well ;)
TimToady we're trying to punish bit twiddlers
moritz_ whip the bit twiddlers! 23:36
but I'm sure it'll allow great obfus
TimToady "I feel bit-slapped!"
mncharity what do we have for a p6 wiki? www.perlfoundation.org/perl6/index.cgi ? 23:37
lambdabot Title: Perl 6 / Perl 6
mncharity perl.net.au/wiki/Perl_6 perl6.cz/wiki/Perl_6_and_Parrot_links 23:38
lambdabot Title: Perl 6 - PerlNet
TimToady there's also dev.pugscode.org/wiki/ 23:40
lambdabot Title: Pugs - Trac
TimToady and www.perlfoundation.org/parrot/index.cgi 23:42
lambdabot Title: Parrot / Parrot
TimToady pugs: say 7 +& -2 23:46
exp_evalbot OUTPUT[6␤]
[particle] lovely
TimToady I suppose we could mention that Int assumes 2's complement arithmetic. 23:47
[particle] good idea. 23:51
mncharity I dropped the links as perl.net.au/wiki/Perl_6#Wikis (perl.net.au, unlike perfoundation.org, didn't require me to remember or create an account) 23:56
lambdabot Title: Perl 6 - PerlNet
mncharity the one liner seems to be "perl.net.au is active, perlfoundation is official, pugscode is for pugs, perlfoundation.org/parrot/ is for parrot, and perl6.cz has some nice links". 23:58

Powered by Cro and the Raku® Programming Language.

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

Thank you!