Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2007-04-19

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
16:34 TimToady cmarcelo: I think Sage would be okay with it.  On the other hand, I'll take a Haskell translator over no translator any day
16:34 fglock cmarcelo: it's not haskell
16:34 TimToady and indeed, a Python irclogger is better than nothing... :/
16:35 moritz but don't expect it to be working very soon
16:36 moritz the logging part seems to work, basically
16:36 moritz now for the output...
16:37 cmarcelo fglock: ?? I was refering to a hs2p6 recently commited somewhere in the pugs tree..
16:37 fglock no, that wasn't me
16:37 [particle] cmarcelo: never heard of it
16:42 cmarcelo misc/pX/Common/hs2p6/, it's from pmurias
16:43 fglock back in 3h &
16:43 svnbot6 r16018 | fglock++ | kp6 - fixed bind; added test
16:46 devbot6 dev: Ticket #22 (task created): Perl 5 to Perl 6 translator <http://dev.pugscode.org/ticket/22>
16:48 kolibrie moritz: if you want to write it in p6, you can look at examples/network/evalbot/evalbot.pl for an example
16:49 moritz kolibrie: I decided against perl6 because atm there are too many problems with my pugs builds...
16:49 moritz kolibrie: and I dont know if DBI works in p6
16:50 kolibrie moritz: I haven't used DBI under p6, but the bindings are there to do it
16:50 moritz kolibrie: right, but I want fast results ;-)
16:50 TimToady if you want fast results why are you hanging around near Perl 6?  :)
16:50 kolibrie lol
16:51 offby1 We will sell no language .... << long pause; swirls wine in glass >>  ... before its time
16:51 TimToady we'll gladly sell futures in it though...
16:51 moritz TimToady: ;-)
16:52 moritz TimToady: because some day, hopefully, p6 will enable me to get results fast
16:55 jrockway heh, options.perl.org # buy options on the perl6 release date, benchmark results, etc.
16:56 moritz jrockway: good marketing idea ;-)
16:59 TimToady audreyt++ # [PATCH] Add Tolkien quote to mro.c [perl5-porters]
17:00 avar parrot needs more alice in wonderland quotes
17:04 jrockway the russian in main.hs (i think) crashed my emacs when i was first examining the code :)
17:20 moritz my very first attempt:
17:20 moritz http://moritz.faui2k3.org/irclog/out.pl?channel=perl6;
17:20 lambdabot Title: IRC log for #perl6
17:21 moritz currently I'm gathering the data on a different maschine than the server (because the server has no Net::IRC)...
17:21 moritz so it is not updated automatically
17:21 moritz but that will come soon
17:36 avar moritz: pocoirc is way better:)
17:38 moritz avar: do you have a link for me?
17:41 moritz ok, found the SVN repository
17:42 moritz "Note that using this module requires some familiarity with the details of the IRC protocol. I'd advise you to read up on the gory details of RFC 1459
17:43 moritz not too promising for a quick start, eh?
18:14 moritz test
18:14 moritz yay, basic functionality is there
18:23 OwlEye test: umlaute: ä ü ö
18:25 moritz OwlEye: they work for me
18:25 OwlEye and in your log bot? :P
18:25 moritz OwlEye: that's what I'm talking about
18:25 OwlEye oh, nice!
18:25 moritz OwlEye: see http://moritz.faui2k3.org/irclog/out.pl?channel=perl6;
18:25 lambdabot Title: IRC log for #perl6
18:26 moritz more unicode test: 수도쿠
18:26 OwlEye even formatted :)
18:26 svnbot6 r16019 | kudra++ | Another summary bites the dust
18:26 OwlEye and UTF-8 seems to work too
18:26 moritz OwlEye: just a simple table
18:27 OwlEye very well done
18:27 moritz non-utf-8 would be more interesting right now
18:27 moritz especially latin-1
18:27 OwlEye we just spoke about it and you already have results :D
18:27 moritz that's perl
18:27 OwlEye mine was non-UFTC-8
18:27 moritz cool
18:27 OwlEye i use ISO-8859-15
18:28 moritz which is latin1, except for a few characters (like € iirc)
18:28 OwlEye yes i think you are right
18:28 audreyt irc.pugscode.org now redirects there.
18:29 audreyt autosmoke.pugscode.org now redirects to http://perlcabal.org/smoke.html
18:29 lambdabot Title: TAP Matrix - Thu Apr 19 14:39:13 2007 GMT
18:30 moritz audreyt: you're a big fan of hostnames, right?
18:30 audreyt right.
18:30 rindolf Hi all.
18:30 moritz hi rindolf ;_)
18:31 TimToady audreyt: are you going to YAPC::NA?  or do you know if there's any hackathoning going on around it?
18:31 OwlEye audreyt, irc.pugscode.org does not let me in
18:31 TimToady trying to figure out planes...
18:31 audreyt TimToady: no, in OSDC.tw it turns out folks are skipping Houston for some reason
18:31 audreyt weather perhaps?
18:32 rindolf audreyt: can you point me to some good CPAN modules for doing some OO tricks? Here's what I have so far - http://www.shlomifish.org/perl-oop-systems-n6eWb7==LK/summary.txt
18:32 audreyt dunno about hackathon... I'm still massively backlogging on mail and stuff
18:32 TimToady I know the feeling
18:33 kolibrie I'm planning on YAPC::NA, but cannot stay this year for a hackathon :(
18:37 OwlEye thats in Houston, TX, aye?
18:37 kolibrie correct
18:37 audreyt yeah
18:38 OwlEye will any of you attend the DebConf by chance?
18:38 svnbot6 r16020 | kudra++ | The story so far
18:43 TimToady audreyt: how's ingy doin'?
18:44 audreyt TimToady: still pending two operations (one chest, one wrist).
18:44 TimToady you mentioned "spine" at first, how's that?
18:44 rindolf I'm going to kill Ask Bjorn Hansen!
18:45 audreyt it seems to be fixed, or at least stabilized, atm
18:45 rindolf He still hasn't uploaded XML::RSS 1.30!
18:45 audreyt rindolf: if you kill him, he won't upload it ever after.
18:45 rindolf audreyt: yes.
18:45 rindolf audreyt: I was joking.
18:45 rindolf audreyt: he also has been relatively unresponsive.
18:46 geoffb somehow death is not funny with ingy in surgery
18:46 rindolf geoffb: oh.
18:46 rindolf geoffb: it's an expression, and I didn't know.
18:46 rindolf What's wrong with ingy?
18:46 geoffb serious car accident
18:46 moritz rindolf: car accident
18:46 rindolf Wow!
18:47 rindolf Was he driving?
18:47 rindolf Or in a car?
18:47 audreyt was driving motorbike. more details later.
18:47 rindolf Oh.
18:47 rindolf Motorbikes are very dangerous.
19:58 pmurias testing out the new irc log ;)
20:01 fglock pmurias: what is the url?
20:01 moritz actions (with /me) are still not logged :(
20:01 moritz fglock: irc.pugscode.org or http://moritz.faui2k3.org/irclog/
20:01 lambdabot Title: IRC logs
20:02 fglock moritz: thanks
20:02 moritz fglock: np
20:07 moritz if somebody hides from the log, and nobody notices, does the tree really fall?
20:07 fglock hmm - the changes in kp6 containers broke OO
20:08 pmurias fglock: hi
20:08 fglock hi
20:10 pmurias does ghc have a test suit?
20:11 fglock http://hackage.haskell.org/trac/ghc/wiki/Building/RunningTests
20:11 lambdabot Title: Building/RunningTests - GHC - Trac
20:15 OwlEye lichtkind, yay!
20:15 lichtkind OwlEye: owl lange nich gesehen
20:15 OwlEye lichtkind, yeah! how have you been my dear friend?
20:25 lichtkind renormalist: steffen?
20:28 pmurias good night&
20:30 lichtkind goo nicght
20:33 lumi moritz: The logger had a hiding feature, actually
20:34 moritz lumi: lots of them ;-)
20:34 moritz lumi: currently only "plain messages" are logged :(
20:35 lumi moritz: So not actions or.. what.. notices?
20:35 moritz lumi: no actions, joins, leaves, kicks, topic changes etc.
20:35 lumi moritz: I meant the previous one, not the shiny new one
20:35 lumi Oh right, I mentally blank them
20:35 moritz lumi: ok
20:36 moritz lumi: was it something like "all lines starting with $marker are ignored"?
20:37 lumi moritz: Yes
20:37 lumi Something like [off]
20:37 lumi I think
20:38 moritz lumi: I can implement that, it's a one-liner
20:41 moritz and now without ignore...
20:51 moritz ?eval 1
20:51 evalbot_r16020 1
20:53 fglock would it be an oversimplification if containers and values were just a special kind of objects?  
20:53 fglock or they already are
20:55 fglock hmm - per S12, it seems containers have an underlying object
20:56 fglock which means I can integrate MOP and containers
20:56 audreyt they already are.
20:56 audreyt a Scalar container is just a variable bound to a normal Scalar object.
20:56 fglock audreyt: hey :)
20:56 audreyt you get to that object with VAR($x)
20:56 fglock yup
20:57 audreyt so, in short, yes.
20:57 fglock thanks
20:57 audreyt np :)
21:02 fglock &
21:12 masak S05 still contain regexes on the form s/.../.../; I thought these were all replaced by s[...] = ...; but maybe the old form is still valid?
21:16 moritz one might think that if $larry (or somebody else) would deprecate and remove such a feature, that he greps through the specs and removes all occurences...
21:16 moritz so probably it's still allowed
21:16 masak moritz: just making sure
21:16 masak what's fail do outside of a regex?
21:16 masak ?eval sub a { fail }; a
21:17 evalbot_r16020 *** Failed␤    at <eval> line 1, column 9-14
21:17 moritz masak: I'm not sure as well ;-)
21:18 moritz ?eval my $f = "abc"; $f ~~ s/^a.//; $f
21:18 evalbot_r16020 \"c"
21:18 masak throws an exception, it seems
21:18 moritz well, it... fails ;)
21:19 masak moritz: seems like reasonable semantics, given the name :)
21:20 Tene ?eval sub f { fail }; my $b = f(); "Goodnight World"
21:20 evalbot_r16020 "Goodnight World"
21:20 masak moritz: so I guess the regex itself can be considered a try block of sorts, since it catches this failure ordinarily and does nothing
21:27 moritz sounds reasonable
21:33 masak how is <ws> implemented? seems it must "go outside of itself" to gain enough information about what surrounds it
21:33 moritz masak: with <before.. > and <after...> assertions?
21:33 masak moritz: ah
21:33 masak yes.
21:34 moritz <before \w>\s+<after \w> | \s*
21:34 moritz or something of the like
21:34 moritz ok, that's wrong
21:34 moritz ;)
21:34 masak but the theory works
21:34 moritz right
21:35 PerlJam The trick is to stop thinking of <ws> as a regular expression.   It's code.  With full access to the string.
21:35 moritz <!before \w>\s*<!after \w> | \s+ perhaps
21:35 masak hm, yes
21:35 moritz PerlJam: well, it kinda _is_ a regex - but a regex can do more than just looking at the given string
21:36 masak PerlJam: I was just curious as to the implementation if one wanted to implement the regex engine without cramming too much extralinguistic magic into it
21:37 PerlJam masak: Check the innards of PGE.  It implements <ws>
21:37 masak <before.. > and <after...> seems to be the answer I was looking for
21:37 masak PerlJam: thx, I will
21:54 lichtkind gnight boysngirls
21:55 moritz lichtkind: 'night
21:55 lichtkind :)
21:59 TimToady masak: yes, s/foo/bar/ is still valid; it's only the bracketed form that uses pseudoassignment instead.
22:00 TimToady also, both this and <ws> have rules in src/perl6/Perl-6.0.0-STD.pm already showing the intent of the spec (I hope).
22:01 TimToady though of course <ws> for parsing Perl is not the same as <ws> in general, given we also have to recognize "unspace" and pod
22:02 TimToady and also insert any pending heredocs at the line boundary
22:02 PerlJam heh ... come to think of it, perl 6 put's python's semantic whitespace to shame ;>
22:02 masak TimToady: why was the bracketed form changed to use pseudoassignment?
22:04 TimToady just to make a mechanism to have any old expression on the right, and avoid brackets that look like some other construct that are just quotes
22:04 TimToady as a bonus it gets us += and such
22:05 masak TimToady: so s[...][...] still works as well?
22:05 TimToady ?eval my $_ = "0 1 2 42"; s:g[\d+] += 1; $_
22:05 evalbot_r16020 \"0 1 2 42"
22:05 TimToady hmm.
22:06 TimToady no, that form is no longer allowed
22:06 masak why not?
22:06 TimToady it was causing some kind of parsing problem, don't remember offhand
22:06 Tene ?eval my $_ = "1 2 3 4"; s:g[\d+] = 137; $_
22:06 evalbot_r16020 \"1 2 3 4"
22:07 TimToady seems a bit busted
22:07 Tene ?eval $s = "1 2 3"; $f ~~ s:g/\d/5/; $f
22:08 evalbot_r16020 \undef
22:08 Tene ?eval $s = "1 2 3"; $s ~~ s:g/\d/5/; $s
22:08 evalbot_r16020 \"1 2 3"
22:08 Tene ?eval $s = "foo"; $s ~~ s/foo/moose/; $s
22:08 evalbot_r16020 \"moose"
22:09 TimToady ?eval my $_ = "0 1 2 42"; s:g[<digit>+] += 1; $_
22:09 evalbot_r16020 *** Cannot parse regex: <digit>+␤*** Error: ␤\"0 1 2 42"
22:09 TimToady ?eval my $_ = "0 1 2 42"; s:g[<[0..9]>+] += 1; $_
22:09 evalbot_r16020 \"0 1 2 42"
22:10 TimToady integration with the regex engine has never been pugs' strong point...
22:10 Tene ?eval my $_ = "foo"; s[foo] = "duck"; $_
22:11 evalbot_r16020 \"duck"
22:11 Tene ?eval my $_ = "a b c"; s[\w] ~= "z"; $_
22:11 evalbot_r16020 \"za b c"
22:12 Tene ?eval my $_ = "a b c"; s:g[\w] ~= "z"; $_
22:12 evalbot_r16020 \"a b c"
22:12 TimToady very strange
22:12 Tene ?eval my $_ = "0 1 2 42"; s[\d+] += 1; $_
22:12 evalbot_r16020 Out of memory!
22:12 moritz wtf?
22:12 TimToady $_ = "foo"; s[fo*] = "duck"; $_
22:13 TimToady ?eval $_ = "foo"; s[fo*] = "duck"; $_
22:13 evalbot_r16020 \"duck"
22:13 TimToady not your first append append z before the match
22:13 TimToady *note
22:13 TimToady so I guess it's recursing on the ^ location
22:14 TimToady though how what you said differs from what I said earlier, I dunno...
22:14 TimToady oh, you left out the :g
22:14 TimToady hmm
22:14 Tene I dropped the... yeah
22:15 Tene because s[\w] works and s:g[\w] doesn't
22:15 TimToady except s[\w] didn't work right
22:15 TimToady shoulda put the z after the a
22:16 Tene Oh, right.
22:16 Tene ?eval my $_ = "0 1 2 42"; s[\d] += 1; $_
22:16 evalbot_r16020 \"10 1 2 42"
22:17 TimToady Match is true but it's getting the wrong location out of it, I guess.
22:17 moritz ?eval my $_ = "0 1 2 42"; s[\d] = 1 + ~ $/; $_
22:17 evalbot_r16020 \"10 1 2 42"
22:18 moritz ?eval my $_ = "0 1 2 42"; g:s[\d] = 1 + ~ $/; $_
22:18 evalbot_r16020 Error: No such subroutine: &g
22:18 moritz ?eval my $_ = "0 1 2 42"; s:g[\d] = 1 + ~ $/; $_
22:18 evalbot_r16020 \"0 1 2 42"
22:18 moritz ?eval my $_ = "0 1 2 42"; s:g[(\d)] = 1 + $0; $_ #my last try ;)
22:18 evalbot_r16020 \"0 1 2 42"
22:18 TimToady ?eval $_ = "foo"; s[\w] = (say $/);
22:19 evalbot_r16020 OUTPUT[␤] Bool::True
22:19 TimToady ?eval $_ = "foo"; s[\w] = (say $/.perl);
22:19 evalbot_r16020 OUTPUT[\Match.new(␤  ok => Bool::True, ␤  from => 0, ␤  to => 0, ␤  str => "", ␤  sub_pos => (), ␤  sub_named => {}␤)␤] Bool::True
22:19 TimToady yes, from and to are wrong.
22:20 geoffb Pugs is using PCR these days, isn't it?  Perhaps out-of-date linkage?
22:20 Tene ?eval my $_ = "duck"; s[(\w)] = (say $/.perl)
22:20 evalbot_r16020 OUTPUT[\Match.new(␤  ok => Bool::True, ␤  from => 0, ␤  to => 0, ␤  str => "", ␤  sub_pos =>␤    (Match.new(␤       ok => Bool::True, ␤       from => 0, ␤       to => 0, ␤       str => "", ␤       sub_pos => (), ␤       sub_named => {}␤     ),), ␤  sub_named => {}␤)␤] Bool::True
22:20 TimToady maybe it's time to run a smoke...
22:22 TimToady could also be svk damage on feather, I suppose
22:23 geoffb Did clkao ever figure out what caused feather's svk to be borken?
22:24 Tene We just need to find a way to get svk to damage things more creatively, causing tests to unexpectedly pass instead of fail. ;)
22:25 TimToady we need a room full of monkeys for that.
22:25 TimToady "To be or not to be, that is the gvrklsjbxx."  # a near miss
22:27 Tene Although... if svk has changed some code... does that count as submitting a patch?  And if so, should we send it a commit bit?
22:28 geoffb Why discriminate against non-carbon intelligence?  Silicon has rights, too!
22:29 TimToady and now that it can lase, it also has mights!
22:58 TimToady erg, my smoke appears to be looping on propcharset.  why am I not surprised...
22:59 TimToady anybody remember what the ulimit is set to for individual tests?
22:59 TimToady that test usually finishes in 5 min, and it's up to 12 already...
23:02 TimToady hmm, looks like not limited by default
23:03 TimToady and just as I was about to kill it, it went away on its own, after 15 min runtime
23:03 TimToady so something 3x slower, it would seem
23:04 TimToady s:2nd/<sp>/ is /
23:05 TimToady ah, no, it's my stupid computer running at reduced speed for some reason.
23:05 jrockway power saving
23:05 jrockway ?
23:05 TimToady must think it's on battery
23:06 TimToady 800Mhz kinda sucks
23:06 veritos larry: what is your pc anyway?
23:06 moritz is there a decent documentation for Irssi.pm anywhere?
23:06 TimToady an intel laptop; but it's probably just the Fedora confused about coming back from suspend
23:14 dmq pasted "new char class macros" (162 lines) at http://paste.husk.org/8159
23:15 dmq timtoady: looks like i will be able to do perl6 style \v\V \h\H after all
23:17 dmq and maybe make perl5 faster at handling \s at the same time.
23:30 moritz foo
23:31 TimToady bar
23:31 Tene goose
23:32 dmq you gotta admit those are pretty funky tests
23:32 TimToady gander
23:34 TimToady I sincerely hope they are automatically generated.
23:34 dmq naturally. :-)
23:35 dmq you made Perl real suitable for stuff like that.
23:36 dmq doesnt have to be an evil ternary either. ive been thinking maybe functions would be better.
23:37 dmq btw, that code really made me appreciate how logical operators work in Perl as compared to C.
23:38 dmq it would be nice to be able to change stuff like ((X) ? 2 :0) into (X) && 2
23:40 TimToady btw, s:g[\d+] += 1; works fine in my copy, so it must be feather having trouble.

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo