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.
Auzon pugs: my $foo = "foobar"; $foo ~~ s/bar/d/; 00:04
exp_evalbot RESULT[Bool::True]
Auzon pugs: my $foo = "foobar"; $foo ~~ s/bar/d/; $foo
exp_evalbot RESULT[\"food"] 00:05
Auzon Ah....
Auzon pugs: {v5; print /a/;} 00:18
exp_evalbot OUTPUT[*** No such subroutine: "&v5"␤ at /tmp/VhMUYjOkWQ line 1, column 2-4␤]
Auzon pugs: {use v5; print /a/;}
exp_evalbot RESULT[1.0]
Auzon pugs: {use v5; $_ = "qwerty"; print /a/;} 00:18
exp_evalbot RESULT[1.0]
Auzon pugs: {use v5; $_ = "qwerty"; $_ =~ /(\w)/;}
exp_evalbot RESULT[1.0]
Auzon pugs: {use v5; $_ = "qwerty"; my($capt) = $_ =~ /(\w)/; return $capt} 00:19
exp_evalbot RESULT["q"]
Auzon pugs: {$_ = "qwerty"; my($capt) = $_ =~ /(\w)/; return $capt}
exp_evalbot OUTPUT[*** ␤ Unexpected "/(\\"␤ There is no =~ operator in Perl 6 -- did you mean ~~ (match) or ~= (concat-assign)?␤ at /tmp/5sRzHUDAiG line 1, column 35␤]
Auzon Very nice :)
dduncan Following a make realclean, I started to rebuild the latest Pugs using the final release of Perl 5.10.0, which worked, then started a make smoke, which is in progress. Anyone else testing Pugs with the final 5.10.0? 01:54
Squern Where's the latest docs on Perl 6 subroutine and method parameter lists? 02:42
Tene Squern: you can start looking at spec.pugscode.org/ I think. 02:44
lambdabot Title: Official Perl 6 Documentation
Limbic_Region dev.perl.org/perl6/doc/design/syn/S06.html
lambdabot Title: Synopsis 6: Subroutines - perl6
Squern Oh good, up to date synopsis. Thanks. 02:45
Perhaps dev.perl.org should point to them? They all point to the originals from 4-6 years ago
Tene I've been disappointed a few times by dev.perl.org being so out-of-date 02:46
Auzon pugs: method prefix:<±> (Num $x --> Num) { return +$x | -$x } my $tolerance = ±7!; 03:02
exp_evalbot OUTPUT[decodeUTF8': bad data: '\177'␤decodeUTF8': bad data: '\177'␤*** ␤ Unexpected "$tolerance"␤ expecting "=", context, ":" or "("␤ at /tmp/ObweMOMkM4 line 1, column 60␤]
Auzon pugs: method prefix:<±> (Num $x --> Num) { return +$x | -$x } my $tolerance = ±7; 03:03
exp_evalbot OUTPUT[decodeUTF8': bad data: '\177'␤decodeUTF8': bad data: '\177'␤*** ␤ Unexpected "$tolerance"␤ expecting "=", context, ":" or "("␤ at /tmp/AGNtDV3n9K line 1, column 60␤]
Auzon pugs: method prefix:<+-> (Num $x --> Num) { return +$x | -$x } my $tolerance = +-7;
exp_evalbot RESULT[\-7]
Auzon pugs: method prefix:<pm> (Num $x --> Num) { return +$x | -$x } my $tolerance = pm7;
exp_evalbot OUTPUT[*** No such subroutine: "&pm7"␤ at /tmp/hW2qmcbCic line 1, column 69-77␤]
Auzon pugs: method prefix:<pm> (Num $x --> Num) { return +$x | -$x } my $tolerance = pm 7;
exp_evalbot OUTPUT[*** No such subroutine: "&pm"␤ at /tmp/ZVFq5Upx8L line 1, column 69-78␤] 03:04
Tene pugs: my $x = 1; -1 | +1
exp_evalbot RESULT[(-1 | 1)]
Tene pugs: multi method prefix:<±> (Num $x --> Num) { return +$x | -$x }; my $tolerance = ±7; $tolerance 03:05
exp_evalbot OUTPUT[decodeUTF8': bad data: '\177'␤decodeUTF8': bad data: '\177'␤*** ␤ Unexpected "$tolerance"␤ expecting "=", context, ":" or "("␤ at /tmp/dETqiTqehs line 1, column 67␤]
Tene pugs: "evalbot chokes on utf8"
exp_evalbot RESULT["evalbot chokes on utf8"]
Tene pugs: multi method prefix:<_> (Num $x --> Num) { return +$x | -$x }; my $tolerance = _7; $tolerance 03:06
exp_evalbot OUTPUT[*** No such subroutine: "&_7"␤ at /tmp/Os4Ma3eXzF line 1, column 75-82␤]
Tene Bah.
Auzon Indeed
Squern So is "foo { code() } bar" for real? 03:07
That is, interpolating {...} in strings as code? 03:08
Auzon pugs: "foo { 'For serious!' } bar"
exp_evalbot RESULT["foo For serious! bar"]
Squern I ask because that's going to mean an awful lot of backwacking.
Auzon Apparently
pugs: q~foo { 'For serious!' } bar~ 03:09
Squern {} is pretty common in strings. It's also going to make YAML a pain.
exp_evalbot RESULT["foo \{ \'For serious!\' } bar"]
Auzon Only for double quoted string
Squern Oh, we never use those
Schwern www.google.com/codesearch?hl=en&...0&sa=N 03:11
lambdabot Title: lang:perl \"\{ - Google Code Search, tinyurl.com/2br2q9
Schwern Google code estimates "about 11,500" hits for "{ 03:12
Although it seems to be 271 in reality. Heh. 03:13
Auzon To be honest, I'd rather it use an ampersand at the beginning or something like that
Schwern That's what I thought 03:16
Or ${} and @{} to mean scalar vs array context like we do with @{[]}
Schwern digs up the appropriate mailing list 03:17
Auzon Yeah. I've never heard of anyone having problems with accidentally putting @{[...]} in their string
pugs: "{use v5;}" 03:18
exp_evalbot RESULT[""]
Auzon pugs: "{...}"
exp_evalbot No output (you need to produce output to STDOUT)
Auzon pugs: qq:!c "{This should be the default, optionally allowing code.}" 03:21
exp_evalbot OUTPUT[*** ␤ Unexpected "\"{"␤ expecting operator␤ at /tmp/7bbSAhsfNQ line 1, column 7␤]
Auzon pugs: qq:c(0) "{This should be the default, optionally allowing code.}"
exp_evalbot OUTPUT[*** ␤ Unexpected "\"{"␤ expecting operator␤ at /tmp/kLpFSbOkWm line 1, column 9␤]
Auzon pugs: q:s "{This should be the default, optionally allowing code.}"
exp_evalbot RESULT["\{This should be the default, optionally allowing code.}"]
Auzon pugs: q:c "{This should be the default, optionally allowing code.}"
exp_evalbot OUTPUT[*** ␤ Unexpected "}\""␤ expecting ".", "\187", ">>", "=", "^", operator name, qualified identifier, variable name, "...", "--", "++", "i", array subscript, hash subscript or code subscript␤ at /tmp/A5XCvtqJKU line 1, column 60␤]
Auzon So turning it on works, but not off
pugs: gather for 1..4; 03:30
exp_evalbot OUTPUT[*** ␤ Unexpected ";"␤ expecting "_", fraction, exponent, term postfix, operator or block construct␤ at /tmp/zFqHsZLtrj line 1, column 16␤]
Auzon pugs: gather for 1..4 {.say};
exp_evalbot OUTPUT[1␤2␤3␤4␤]
Schwern Oh, ${...} won't work because ${foo} already means something. 03:47
Oh, but not in Perl 6! 03:48
chacha_chaudhry join #gnowsys-dev 10:59
*typo :P
masak is there an official p5-to-p6 translator? what's the state of that? 12:46
moritz_ masak: not sure... the infrastructure the the -DMAD option to Perl 5 Configure 12:47
masak ok, thx 12:47
moritz_ at least that's what parses the Perl 5
masak a year or so ago, I heard Larry talk about a translator of his that converted something like 99% of CPAN into Perl 6 12:49
it sounded like he was on the verge of releasing something
but since then, I haven't heard so much, and now his work seems to be focused on the Perl 6 grammar file
moritz_ I think that was the plan, not the status ;)
masak might have been :) 12:50
moritz_ builds a 5.10.0 with -DMAD now 12:52
wolverian sid is still on 5.8.8?!?! so slow ;) 13:00
moritz_ aye 13:09
Patterner I expected Sid to have 5.6.0... 13:45
or 4.036...
wolverian um, what? 13:49
moritz_ they're slow, but they don't go backwards 13:50
wolverian I wouldn't even call sid slow 13:52
it was sarcasm! or something.
stop confusing me, guys
jjore-m moritz_, just fyi, MAD acquired a bit of bitrot in the last year. Allegedly it does not fully pass tests and no one has appeared to want to spend tuits on it. 14:26
though it's still probably most loved as an eventual path from a running Perl 5 VM to something useful to a Perl 6 compiler. 14:27
Juerd_ And one of the people who knows much about MAD is spending his tuits on a Perl 5 fork :( 14:36
jjore-m Gerard? 14:45
It's not like MAD in Perl 5 is terribly important right now anyway.
tho I'm glad larry has already added in all the annotations to Perl 5 that presumably he thinks he wants for use later. 14:46
"all" 14:47
Juerd_ jjore-m: Yes 14:52
Probably not all, but quite a lot
fglock [particle]: ping 15:58
fglock we just had a discussion on #parrot about tracking variables at compile-time 16:35
the reason for this discussion is that the kp6 compiler seems to be doing extra work, keeping track of lexical scopes that it doesn't have to 16:37
this makes it more difficult to port to other backends, such as Parrot
one such case is: my $x; BEGIN { $x = 3 } # illegal in Perl 6? 16:39
a corner case is: our $x; BEGIN { $x = 3 } # possibly illegal 16:40
but this would be ok: BEGIN { our $x = 3 } 16:41
fglock another implication is that lexically scoped macros would be impossible, but 16:42
pmichaud said macros would likely instantiate a new parser, 16:43
and the scope would be syntactically delimited
moritz_ sounds reasonable
fglock note that these cases are legal in pugs and in kp6, but S04 and S02 seem to forbid it 16:50
fglock the BEGIN discussion have happened before: 17:19
groups.google.com/group/perl.perl6....9a6cce75fb
s/have/has/ 17:20
fglock the original discussion on kp6 BEGIN: 17:24
svn.openfoundry.org/pugs/v6/v6-Kind...-block.pod
(long)
lambdabot tinyurl.com/2d3hcu
[particle] sd times blames schwern! www.sdtimes.com/article/latestnews-...15-05.html 19:32
lambdabot Title: Com.org - Only the best links ... 19:33
Schwern particle: ...for something I did five years ago! 22:27
And poor Richard Clamp gets no mention. :(
Limbic_Region w 22:31
[particle] Schwern: really? ah, that's rich 22:50
agjal good night 22:52
Schwern particle: I sent a correction. Thanks for pointing it out. 23:31
Juerd_ Schwern: @ is used even more often than {, in strings, or so a quick look at my code base shows 23:46
Schwern: And @ has needed escaping in "" longer than I have used Perl. I don't even know which version started to require that. 23:47
Schwern Perl 5
But we already know about @
And $
Why add to the pile? 23:48
Juerd_ I will counter every why question with a why not. 23:48
Schwern Actually Perl 6 fixes the [email@hidden.address] problem
Juerd_ And vice versa
Schwern Juerd: More to escape is more to think about which is bad. 23:49
Juerd_ No, it makes it even worse, IMO, by working *most of the time*
Schwern ?
I've only played with it a little but it appears you have to try pretty hard and the result looks an awful lot like a programming construct.
Anyhow, that's stuff I haven't looked into. 23:50
Juerd_ [email@hidden.address] works [email@hidden.address] does not.
Schwern So, umm... what was the point again?
Juerd_ So it only makes it *harder* to realise you're doing something that might break.
Schwern I think that's stretching. 23:51
Also, getting $obj.method() working is soooo worth it
But again, it's not something I've really looked into. Just the {} thing struck me. 23:52
$obj.method() at least still follows the "sigil in a string" tripwire.
That is, I still only have to be dimly aware that putting sigils in a string requires a little thought. 23:53
Whereas {} opens up a whole new class for little gain.
${} would be no more mental load than what we have now 23:54
Juerd_ Just like the whole damn UTF8 mess in Perl 5 - if you do nothing, but work with UTF8 d(my terminal went away at this point)ata a lot, you think you have good code
UNTIL you hit that edge case
$obj.method() requiring parens requires look-ahead parsing, which sucks for parsability.
Schwern Fortunately, that's not a user problem. 23:55
Juerd_ Oh yes it is
User needs to parse too!
Schwern ?
With their eyes?
Juerd_ And brane, yes
Schwern We don't parse like computers.
Juerd_ Often we do 23:56
Schwern Picking out $obj.method(...) is easy for us.
Also, it's bloody rare in text
Juerd_ I'm happy with s:i/foo/bar/ for exactly that reason.
Schwern That's a whole different league
The problem with s/foo/bar/i isn't so much that the special cases are at the back, it's that there's *no indiciation* of a special case at the front. 23:57
Juerd_ In [email@hidden.address] the () at the end make the @ at the beginning special.
Schwern That it's a silly case is relevant.
Schwern When you interpolate, there's always going to be a way to trip over it. The question is, how likely is that? 23:57
Juerd_ Anyway, I'm okay with the [email@hidden.address] thing, mostly. Just pointing out that it doesn't fix the problem, only make it a bit worse when you do run into it, even if that is less likely.
Schwern It would be interesting to grep CPAN for instances of m{\.\w+\(} inside a string. 23:58
Juerd_ I'm also okay with { } -- I'd *almost* ask for a \w/\W distinction like the one that was made for regexes: every \w character is literal, every \W character is special or reserved for future use. 23:59
Schwern so... uppercase braces?
Juerd_ \\\w is special, \\(\W) is literal $1
Schwern You've lost me 23:59
Juerd_ \w is alphanums and underscore
\W is the opposite
Schwern Ok

Powered by Cro and the Raku® Programming Language.

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

Thank you!