Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2007-07-08

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:08 diakopter_afk changed the nick to diakopter
00:10 Limbic_Region salutations diakopter
00:12 diakopter thanks :) u2
00:12 bonesss joined #perl6
00:14 * diakopter experiences a UTF8 bug in the bash/ssh/bash/screen/irssi combo
00:19 buetow joined #perl6
00:23 diakopter Limbic_Region: in what browser/OS were you seeing the 500s on cgiirc?
00:24 Limbic_Region WinXP home with FireFox
00:24 Limbic_Region WinXP profession with IE
00:25 Limbic_Region still getting them btw
00:28 diakopter can you PM me the url you're trying?  I'm not seeing the 500s
00:28 diakopter curious - are you using https?
00:29 diakopter6 joined #perl6
00:29 diakopter6 from cgiirc. hi
00:30 weinig joined #perl6
00:30 diakopter6 EWORKSFORME
00:30 Limbic_Region sorry
00:30 Limbic_Region was AFK
00:30 Limbic_Region https://feather.perl6.nl/irc/
00:30 diakopter ah :)
00:31 cognominal_ joined #perl6
00:31 diakopter well I guess it would help if we were using the same URLs. :)
00:31 Limbic_Region that IS the url that is advertised
00:31 diakopter heh; :)
00:31 Limbic_Region if you are using a different URL then the front page should be updated
00:31 Limbic_Region http://feather.perl6.nl/ -> see the link for IRC chat
00:31 lambdabot Title: Perl6 Community Development Server
00:31 diakopter k
00:32 diakopter I found the url by looking for irc.cgi on the filesystem and cetera
00:32 diakopter an url, I mean
00:34 Limbic_Region *shrug* - it did stop working all of a sudden on Friday
00:37 diakopter yeah
00:37 diakopter my fault I'm sure, since I scrapped the PAM auth
00:37 Limbic_Region when I first reported the problem to Juerd some time ago, he told me there were two issues, one with HTTPS and the other with PAM
00:38 Limbic_Region one was fixed relatively quickly (the HTTPS)
00:38 * avar found a use for p5 subs.pm
00:38 Limbic_Region the second was fixed in a day
00:38 Limbic_Region and it worked for a few days
00:38 Limbic_Region then on Friday, it just stopped working
00:39 diakopter yeah, cuz while it appeared to be working it was still throwing lots of errors
00:43 jedai joined #perl6
00:43 Limbic_Region well I don't know how many folks here use it other than me so it might not be much of a big deal
00:49 diakopter please try https://feather.perl6.nl/cgi-bin/cgiirc/irc.cgi
00:49 diakopter just to confirm it works there for you as well
00:50 Limbic_Region in a sec
00:51 LimbicRegion joined #perl6
00:51 LimbicRegion changed the nick to Limbic_Region
00:52 Limbic_Region daughter turned off the wireless card
00:52 Limbic_Region but yes, it works
01:05 LimbicRegion joined #perl6
01:11 _iShock joined #perl6
01:26 vel joined #perl6
01:30 daxim_ joined #perl6
01:57 ProperNoun joined #perl6
01:59 rlb3 joined #perl6
02:05 Limbic_Region joined #perl6
02:24 dduncan so my smoke all worked with the clean pugs checkout plus clean perl 5.9.5 ... and the change I made re smoker.yml caused no problems that I'm aware of
02:26 diakopter changed the nick to diakopter_afk
02:35 fglock joined #perl6
03:08 teratoma left #perl6
03:23 agentzh joined #perl6
03:23 agentzh audreyt++ # did what i wanted to do :)
03:23 lambdabot agentzh: You have 1 new message. '/msg lambdabot @messages' to read it.
03:25 agentzh avar said "tell @agentzh " yesterday and it certainly didn't work ;)
03:25 agentzh @messages
03:25 lambdabot avar said 5h 55s ago: You need to check in PCR inc/Test/* stuff into svn, or all the inc/ stuff
03:25 agentzh avar: already checked in as pugs' inc/
03:26 avar that's moved into pcr's?
03:26 agentzh avar: i'm guessing you only checked out PCR
03:26 agentzh without checking out the whole pugs tree
03:27 agentzh avar: it's something like "use Module::Install '../../inc'" in PCR's Makefile.PL
03:28 agentzh avar: re compile_p6grammar.pl
03:28 agentzh it currently treats regex {...} and rule {...} as token {...} and i'm going to change that.
03:30 agentzh sigh...slept too late last night...still quite sleepy...
03:30 pugs_svnbot r16893 | Aankhen++ | trunk:
03:30 pugs_svnbot : * [irclog] couple more abbreviations.
03:30 pugs_svnbot diff: http://dev.pugscode.org/changeset/16893
03:30 lambdabot Title: Changeset 16893 - Pugs - Trac
03:31 agentzh dduncan: so you have changed .smoker.yml to smoker.yml? okay, i'll update it on feather for auto-smoke...
03:31 dduncan yes
03:32 agentzh k
03:32 agentzh personally i quite like this change
03:32 agentzh since "ls" hides .smoker.yml...
03:32 pifuruan joined #perl6
03:33 dduncan assuming your Pugs is up to date, all the files save that one wouldn't need changing
03:33 dduncan my own smoke worked after the committed changes, and that 1 file rename
03:33 pifuruan hi
03:33 agentzh dduncan: my pugs tree on feather is aways up-to-date ;)
03:33 dduncan atta
03:33 agentzh pifuruan: hi
03:34 pifuruan what is pubs?
03:35 pifuruan pugs
03:35 lidden http://pugscode.org/
03:35 lambdabot Title: Pugs - pugscode
03:35 agentzh dduncan: smoker.yml can be confused by smoke.yml, btw
03:36 agentzh dduncan: we're already using smoke.yml for storing smoke results
03:36 dduncan I considered that ... perhaps it should be renamed to smoker_config.yml then?
03:36 agentzh dduncan: maybe smoker.conf.yml?
03:36 agentzh heh
03:36 dduncan I considered that earlier, but for now went for the smaller delta
03:36 dduncan question is, is smoker_config or config_smoker better?
03:36 dduncan the former would sort next to config.yml
03:37 dduncan alternately, can't those files be merged?
03:37 dduncan config.yml already has smoke config in it
03:37 agentzh reasonable
03:37 dduncan smoke_concurrent
03:37 agentzh yeah, why not?
03:37 dduncan that would probably work best ... but it would be more onerous to get right, so I don't want to be the one to attempt that
03:37 agentzh dduncan: commits welcome :)
03:38 pifuruan Pugs is an implementation of Perl 6, written in Haskell. It aims to implement the full Perl6 specification, as detailed in the Synopses. For more information, please see the overview and this interview.
03:38 agentzh pifuruan: right
03:38 dduncan that is, I can rename the smoke file again, but doing the merge I would have concerns about
03:38 agentzh dduncan: understood
03:38 pifuruan so where can i get "the full specification
03:38 pifuruan of perl6
03:38 agentzh pifuruan: http://perlcabal.org/syn
03:38 lambdabot Title: Official Perl 6 Documentation
03:38 pifuruan tks
03:38 agentzh pifuruan: you're welcome
03:39 pifuruan thank u very much,man,
03:39 agentzh pifuruan: let us know if you do like a commit bit
03:39 agentzh dduncan: smoker.yml is not that terribly bad. i'll stay with it for now
03:40 pugs_svnbot r16894 | Aankhen++ | trunk:
03:40 pugs_svnbot : * [irclog] took a shot at preserving whitespace.
03:40 pugs_svnbot diff: http://dev.pugscode.org/changeset/16894
03:40 lambdabot Title: Changeset 16894 - Pugs - Trac
03:40 dduncan okay
03:40 pifuruan i am so sorry, i am chinese ,forgive my poor English,so what is commit bit
03:40 agentzh pifuruan: are you? me too :)
03:40 pifuruan �һ�
03:41 agentzh pifuruan: the commit bit allows you to update the pugs subversion repository
03:41 pifuruan i like it
03:41 agentzh ( http://svn.pugscode.org/pugs )
03:41 lambdabot Title: Revision 16894: /
03:42 agentzh pifuruan: /msg me your email address, and i'll try to send one to you
03:42 pifuruan hanson122@gmail.com
03:42 agentzh k, a sec
03:42 pifuruan thks
03:42 dduncan pifuruan, try setting your irc client to utf8 text encoding if you can, for better compatibility when sending non-ascii characters
03:42 dduncan for that matter, all Pugs repo files are utf8
03:43 dduncan the text ones anyway
03:43 pifuruan ok,i will try
03:43 agentzh pifuruan: what user name would you like for your commit bit?
03:44 agentzh pifuruan: hanson?
03:44 pifuruan ok
03:44 agentzh k
03:44 agentzh sending
03:44 pifuruan i will check now
03:45 agentzh commitbit is not so fast though...
03:45 agentzh my page is still loading...sigh...
03:46 agentzh pifuruan: please let me know if you already get the mail
03:46 pifuruan i am afraid not yet
03:48 agentzh commitbit--
03:48 pifuruan i havn't got ur mail
03:48 agentzh could anyone here send one to him?
03:49 * agentzh gives up.
03:49 TimToady I will try.
03:49 agentzh wow
03:49 pifuruan I love all guys here
03:49 agentzh TimToady++
03:50 pifuruan hanson122@gmail.com
03:54 TimToady it thinks it has sent you an invitation
03:54 pifuruan ok,i received mail from audrey
03:56 agentzh pifuruan: actually you've received a mail from lwall
03:57 agentzh or his bot
03:57 agentzh ;)
03:57 TimToady 欢迎!
03:57 dduncan pifuruan, so it is customary to add your name to the AUTHORS file as your first commit ... and it is a utf8 file
03:57 dduncan I saw that encoded correctly!
03:57 * agentzh 2
03:57 pifuruan hh
03:58 TimToady assuming simplified is correct here, and not 歡迎
03:59 dduncan well, I just mean it didn't look like garbage
03:59 dduncan plus 'incompatible encoding' notice
03:59 agentzh TimToady++ # a Chinese guy
03:59 TimToady I don't suppose "huan1 ying2" would do you any better...
04:00 agentzh PinYin is very hard to read for even native CN speakers.
04:00 agentzh like "huan1 ying2"
04:00 TimToady but dduncan has a slightly better chance of knowing how to pronounce it.
04:01 agentzh TimToady: *nod*
04:01 TimToady though you'd have to know the tones to really do it right
04:01 agentzh exactly
04:01 TimToady dduncan: tone 1 is high, then drop a bit on tone 2
04:01 agentzh lol
04:01 TimToady I mean, other way
04:01 dduncan I don't actually know much about this
04:02 dduncan if anything, TimToady should know a lot more about it than dduncan does
04:03 agentzh tone 2 is like "huh?" while tone 1 is like "yayyyyyyyyyy"
04:03 TimToady mostly don't know much yet except some of the characters; mostly coming at it from the direction of japanese kanji
04:04 agentzh TimToady: it's already good enough :)
04:05 TimToady 好说。
04:05 agentzh ......
04:06 * pifuruan slaps agentzh around a bit with a large trout
04:06 * pifuruan slaps agentzh around a bit with a large trout
04:06 TimToady or maybe 不敢当 is more appropriate
04:06 agentzh heh
04:06 agentzh it's amazing...
04:07 TimToady yep, grep is pretty amazing... :)
04:07 pifuruan timTOady, where are u now
04:07 TimToady Mountain View, California.
04:07 pifuruan o
04:07 clkao TimToady: do you know "牛人" ?
04:08 TimToady cow person?
04:08 clkao ya
04:08 pifuruan ����
04:08 TimToady not in my dictionary, alas
04:08 clkao i shall let the chinese folks explain, the phrase is getting popular in taiwan too
04:08 clkao and i should go to bed
04:09 agentzh TimToady is a 牛人
04:09 clkao agree
04:09 agentzh clkao: go to bed *now*?
04:09 clkao it's 5am here
04:09 agentzh clkao: okay, you're not in .tw...
04:09 TimToady yow, and thought I was late for supper...
04:10 agentzh 牛人 == a smart person with amazing skills
04:11 TimToady so 牛逼 also works
04:11 TimToady (when grep fails, there's google...)
04:11 agentzh LOL
04:12 TimToady that character doesn't occur much in japanese.  "walk" and "fulljar" in my radical naming system.
04:12 TimToady actually, Xwalk, since it only has one tick on top
04:13 agentzh TimToady: is that system open-sourced?
04:14 TimToady vaguely, but I'm still refining it, and need to refactor the whole thing
04:14 agentzh i heard it was originally written in Perl 6 :)
04:15 agentzh (iirc)
04:15 TimToady I tried translating to p6 once a long time ago, but unicode was really slow back then
04:15 TimToady been kinda waiting for things to speed up before I try again
04:16 agentzh hopefully fglock's kp6 will get that point before you decide to try again ;)
04:16 TimToady also need to get a font that supports all the extra character in plane 1 and annotate them like I did all the CJK area...
04:17 TimToady all 20900 characters...
04:17 agentzh oh, so coooooool
04:17 TimToady that...took a while...
04:18 TimToady turns out there's about 3000 different radicalish things, if you count anything that's either repeated more than 5 times or is used as a partial character but is a complete character elsewhere
04:19 agentzh that's interesting :)
04:25 * agentzh decides to make a CPAN release for PCR.
04:29 avar agentzh: what namespace do you think generated grammar stuff should be in?
04:29 avar Pugs::Grammar::*?
04:29 agentzh no
04:29 avar like just some random parser..
04:30 agentzh grammar Foo; just generates package Foo
04:30 agentzh each package is derived from Pugs::Grammar::Base
04:30 avar I mean practically, as a upload-to-cpan thing
04:30 agentzh but it's not necessary if the user doesn't want <ident> or some such
04:31 agentzh avar: upload to cpan for other cpan modules
04:31 agentzh ?
04:31 agentzh avar: or a component in PCR itself?
04:31 avar I mean uploading modules generated with PCR
04:31 agentzh avar: i think it's arbitrary
04:32 agentzh avar: in fact i'm going to regenerate my Makefile::Document::Gmake with compiler_p6grammar.pl
04:32 agentzh no namespace standard is needed here, i think.
04:32 agentzh Pugs::Grammar::* are for PCR only
04:33 agentzh lunch &
04:33 avar aw forget it
04:34 rashakil left #perl6
04:58 pugs_svnbot r16895 | Darren_Duncan++ | trunk:
04:58 pugs_svnbot : ext/Muldis-DB/ : fleshed out Grammar.pod, updated Core.pod, other minor edits
04:58 pugs_svnbot diff: http://dev.pugscode.org/changeset/16895
04:58 lambdabot Title: Changeset 16895 - Pugs - Trac
05:04 pugs_svnbot r16896 | Darren_Duncan++ | trunk:
05:04 pugs_svnbot : ext/Muldis-DB/ : minor bug fix in Grammar.pod
05:04 pugs_svnbot diff: http://dev.pugscode.org/changeset/16896
05:04 lambdabot Title: Changeset 16896 - Pugs - Trac
05:10 diotalevi joined #perl6
05:12 hakobe changed the nick to hkb_aw
05:13 diotalevi so this is neat. If you teach File::Next that things readable by Archive::Extract are directories, you can use ack on CPAN.
05:13 diotalevi changed the nick to jjore
05:13 justatheory joined #perl6
05:18 dduncan izzatso?
05:20 dduncan the joys of abstraction
05:50 pasteling "agentzh" at 222.45.13.33 pasted "Why this p6 grammar is looping?" (16 lines, 321B) at http://sial.org/pbot/26105
05:53 agentzh avar, fglock: any clue?
05:54 weinig joined #perl6
05:54 weinig joined #perl6
05:56 agentzh sigh. it takes forever to commit...
05:57 agentzh &
05:57 agentzh left #perl6
06:01 xinming joined #perl6
06:03 agentzh joined #perl6
06:09 TimToady agentzh: why do you have curlies around the "our"?
06:10 agentzh TimToady: because...because...it's a syntax
06:11 agentzh i just want to make the grammar parser simpler to implement
06:11 agentzh those blocks are considered "global actions"
06:11 agentzh or "global code blocks that are directly passed to the generated file"
06:12 TimToady but $Count won't be visible as a lexical name
06:12 TimToady if it's inside {...}
06:12 TimToady if it works at all, it's because it's a package var and strict is off
06:12 agentzh TimToady: the curlies are removed by the emitter :)
06:12 agentzh TimToady: i know that is a big confusing...
06:13 agentzh *bit
06:13 TimToady well, it's just not Perl 6, is all...
06:13 agentzh TimToady: it's a mix of Perl 6 and Perl 5
06:13 agentzh just an odd grammar syntax designed specific for PCR
06:14 agentzh *specifically
06:14 agentzh ;)
06:14 TimToady yeah, but you called a "p6 grammar" when you nopasted it, which led me to believe in was Perl 6... :/
06:14 TimToady *it
06:14 agentzh sorry for the confusion
06:15 agentzh i should have called it "pcr grammar" or something like that :)
06:16 agentzh hmm...can't commit any more today...it seems...
06:16 TimToady maybe you could make it a P5( our ... ) macro to avoid confusing other people later...
06:16 agentzh TimToady: great idea!
06:21 agentz1 joined #perl6
06:21 agentz1 changed the nick to agentzh_
06:21 agentzh_ or simply borrow yacc's syntax: "%{ ... %}"
06:23 TimToady that would work, especially since p6 doesn't use %{
06:24 agentzh_ :)
06:25 Alias_ joined #perl6
06:25 pugs_svnbot r16897 | agentz++ | trunk:
06:25 pugs_svnbot : [PCR]
06:25 pugs_svnbot : * added token "digit" to P::G::Rule2.pm and updated Rule.pmc accordingly
06:25 pugs_svnbot :   (one more test is passing now :))
06:25 pugs_svnbot : * Pugs::Emitter::Grammar::Perl5 - appended "1;" to each generated
06:25 pugs_svnbot :   package
06:25 pugs_svnbot diff: http://dev.pugscode.org/changeset/16897
06:25 agentzh_ i'd expect compile_p6grammar.pl be much faster than v6.pm for compiling p6 grammars (or PCR grammars)
06:25 agentzh_ since it doesn't bother parsing perl 6 in actions (or closures)
06:26 agentzh_ after chaning my IPs 3 or 4 times, svn ci finally works :)
06:26 agentzh_ *changing
06:27 agentzh_ yay
06:27 Jameswf joined #perl6
06:29 agentzh_ the sadest thing for me is that i'm prevented from committing :)
06:29 Jameswf I was committed once
06:29 agentzh_ Jameswf: commits welcome!!!
06:30 agentzh_ *saddest
06:31 Jameswf sometimea I code quick and dirty, I will do something and my IT guy is like WTF. cleans it up and breaks it. I say it wa oogily but it worked
06:32 Jameswf i think my s is messed up
06:32 pugs_svnbot r16898 | agentz++ | trunk:
06:32 pugs_svnbot : PCR - added examples/digits.grammar which doesn't work yet; it's looping...sigh
06:32 pugs_svnbot diff: http://dev.pugscode.org/changeset/16898
06:32 lambdabot Title: Changeset 16898 - Pugs - Trac
06:32 agentzh_ will change the syntax to %{ ... %} ...
06:33 agentzh_ Jameswf: i like to code quickly too (but not the dirty part...)
06:55 agentzh_ changed the nick to agentzh
06:56 agentzh changing {...} to %{...%} breaks the parser...will do later...
06:57 toshinor_ joined #perl6
07:02 agentzh heh, it finally works: token verbatim { '%{' ( [ <!before '%}'> . ]* ) '%}' }
07:04 hkb_aw changed the nick to hakobe
07:05 agentzh is anyone working on the parsec emitter for pcr?
07:05 agentzh if not, i'll skip t/11-parsec.t's tests by default
07:05 agentzh it's annoying to see the errors everytime running 'make test'
07:21 pugs_svnbot r16899 | Aankhen++ | trunk:
07:21 pugs_svnbot : * [irclog] I am a lean, mean, abbreviation expanding machine.  Or something.
07:21 pugs_svnbot diff: http://dev.pugscode.org/changeset/16899
07:21 lambdabot Title: Changeset 16899 - Pugs - Trac
07:24 devogon joined #perl6
07:26 iblechbot joined #perl6
07:29 agentzh Aankhen++
07:36 dduncan left #perl6
07:43 pugs_svnbot r16900 | agentz++ | trunk:
07:43 pugs_svnbot : [PCR]
07:43 pugs_svnbot : * skipped the failing tests in t/11-parsec.t, t/14-mini_perl6.t,
07:43 pugs_svnbot : and t/12-plain_regex.t
07:43 pugs_svnbot : * added examples/digits.grammar which is looping...sigh...
07:43 pugs_svnbot : * Rule2.pm: changed the global code blocks to %{ ... %} per TimToady++
07:43 pugs_svnbot diff: http://dev.pugscode.org/changeset/16900
07:46 pugs_svnbot r16901 | agentz++ | trunk:
07:46 pugs_svnbot : PCR - added t/16-examples.t to test examples/*.grammar
07:46 pugs_svnbot diff: http://dev.pugscode.org/changeset/16901
07:46 lambdabot Title: Changeset 16901 - Pugs - Trac
08:06 ashelyb_ joined #perl6
08:23 rho joined #perl6
08:28 elmex joined #perl6
08:30 elmex joined #perl6
08:31 Jameswf I wande if pizza_party can be addapted to papajohns
08:35 ashleyb joined #perl6
08:36 monomorph joined #perl6
08:37 elmex joined #perl6
08:46 elmex joined #perl6
09:11 iblechbot joined #perl6
09:11 lixinso joined #perl6
09:24 chris2 joined #perl6
09:25 jedai joined #perl6
09:28 toshinori joined #perl6
09:35 toshinori joined #perl6
09:43 pugs_svnbot r16902 | agentz++ | trunk:
09:43 pugs_svnbot : [PCR]
09:43 pugs_svnbot : - t/16-examples.t: generated .pm files within tmp/.
09:43 pugs_svnbot : - added $VERSION to front-end modules like Pugs::Compiler::Regex,
09:43 pugs_svnbot :   and Pugs::Compiler::Grammar and raised it to 0.24.
09:43 pugs_svnbot : - renamed ChangeLog to Changes.
09:43 pugs_svnbot : - now developers should invoke 'Makefile.PL' as follows:
09:43 pugs_svnbot :     perl Makefile.PL releng
09:43 pugs_svnbot :   in releng mode, pugs' inc/ will not be reused.
09:43 pugs_svnbot : - updated Changes to reflect recent changes.
09:44 pugs_svnbot : - Makeile.PL - removed tmp/ and *.pm in 'make clean'.
09:44 pugs_svnbot : - updated README with my 'releng' script.
09:44 pugs_svnbot diff: http://dev.pugscode.org/changeset/16902
09:44 lambdabot Title: Changeset 16902 - Pugs - Trac
09:46 pasteling "evalbot_r16901" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26108
09:46 agentzh Pugs::Compiler::Rule 0.24 has been uploaded to CPAN :)
09:52 unobe joined #perl6
09:54 agentzh @tell fglock the old failing tests in t/11-parsec.t, t/12-plain_regex.t, and t/14-mini_perl6.t are skipped and 0.24 are released to CPAN :)
09:54 lambdabot Consider it noted.
09:55 meppl good morning
09:55 agentzh good afternoon!
09:56 meppl good afternoon agentzh
09:56 agentzh :)
09:58 unobe good night :-)
09:59 agentzh supper &
09:59 pugs_svnbot r16903 | agentz++ | trunk:
09:59 pugs_svnbot : PCR - CPAN release 0.24
09:59 pugs_svnbot diff: http://dev.pugscode.org/changeset/16903
09:59 lambdabot Title: Changeset 16903 - Pugs - Trac
10:11 unobe changed the nick to unobe_away
10:21 devbot6 dev: WikiStart edited by aharoni <http://dev.pugscode.org/wiki/WikiStart>
10:28 buetow joined #perl6
11:06 pugs_svnbot r16904 | agentz++ | trunk:
11:06 pugs_svnbot : PCR - added examples/langs.grammar (this file contains two grammars and the latter calls the subrules of the former and it works\!)
11:06 pugs_svnbot diff: http://dev.pugscode.org/changeset/16904
11:06 lambdabot Title: Changeset 16904 - Pugs - Trac
11:09 pasteling "evalbot_r16903" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26109
11:18 pugs_svnbot r16905 | agentz++ | trunk:
11:18 pugs_svnbot : PCR - prepends 'use strict; use warnings' to the .pm file generated
11:18 pugs_svnbot diff: http://dev.pugscode.org/changeset/16905
11:18 lambdabot Title: Changeset 16905 - Pugs - Trac
11:20 franck__ joined #perl6
11:28 integral joined #perl6
11:33 pugs_svnbot r16906 | agentz++ | trunk:
11:33 pugs_svnbot : PCR - updates version number to 0.25
11:33 pugs_svnbot diff: http://dev.pugscode.org/changeset/16906
11:33 lambdabot Title: Changeset 16906 - Pugs - Trac
11:36 masak joined #perl6
11:54 pugs_svnbot r16907 | masak++ | trunk:
11:54 pugs_svnbot : [t/syntax/numbers/misc.t]
11:54 pugs_svnbot : * corrected copy/paste bug
11:54 pugs_svnbot : [t/builtins/arrays/slice.t]
11:54 pugs_svnbot : * unbroke mid-testfile crash with a try block
11:54 pugs_svnbot diff: http://dev.pugscode.org/changeset/16907
11:54 lambdabot Title: Changeset 16907 - Pugs - Trac
11:59 masak I must say I was positively surprised by the last line in t/syntax/numbers/hexadecimal.t
11:59 masak see if somebody sees what I mean
12:00 masak it says "is :16<DEAD_BEEF>, 0xDEADBEEF"
12:00 masak I thought the old 0x syntax was dead, bereft of all life, pushing up the daisies
12:00 masak if so, what's it doing in a test file?
12:01 masak if not, why isn't it exercised more in the test file? :)
12:05 rafl joined #perl6
12:09 agentzh yay! "rule foo { ... }" works now
12:09 masak agentzh++
12:09 agentzh :)
12:10 masak how far are we really from 6.28.0?
12:10 * masak longs for christmas :)
12:12 Alias_ Thursday
12:12 Alias_ Of course, I'm not saying WHICH Thursday
12:12 Alias_ Really, I have no idea at all
12:13 masak seems 6.28's been around the corner for ages
12:16 agentzh for 6.28.x, we need audreyt ;)
12:16 masak yes
12:16 Alias_ And where's she lately?
12:16 pugs_svnbot r16908 | agentz++ | trunk:
12:16 pugs_svnbot : [PCR]
12:16 pugs_svnbot : - Pugs::Emitter::Grammar::Perl5 : set sigspace for "rule XXX {...}"
12:16 pugs_svnbot :   and calls the backtracking rule emitter for "regex XXX {...}".
12:16 pugs_svnbot : - added tests to t/04-compiler-grammar.t for this.
12:16 pugs_svnbot : - the ratchet rule emitter : saved the emitter params as special
12:16 pugs_svnbot :   comments in the generated code to help debugging.
12:16 pugs_svnbot : - added some remaining issues to the TODO file.
12:16 pugs_svnbot diff: http://dev.pugscode.org/changeset/16908
12:16 lambdabot Title: Changeset 16908 - Pugs - Trac
12:16 masak Alias_: exactly...
12:16 agentzh she's been resting at home mostly.
12:16 Alias_ burned out?
12:16 agentzh Alias_: sort of
12:17 Alias_ I did real something about some sort of health thing
12:17 masak yes
12:17 masak hepatitis
12:17 Alias_ ouch
12:17 Alias_ which one?
12:18 masak don't remember
12:18 masak prolly not fun either way
12:18 Alias_ ya, but I think it might matter
12:18 masak might have been A
12:18 masak ?eval my @a = ^5; say @a[*-1];
12:19 evalbot_r16866 changed the nick to evalbot_r16908
12:19 evalbot_r16908 OUTPUT[0␤] Bool::True
12:19 masak evalbot_r16908: nope
12:19 * masak needs to stop the bad habit of saying 'say' to evalbot
12:19 Alias_ I really wish I had gotten cvsmonitor working on svn
12:19 Alias_ Would have made this whole "who's doing what" thing so much easier
12:20 masak Alias_: what does cvsmonitor do?
12:20 Alias_ Let me see if I can find a working one
12:20 Alias_ In short though, it monitors cvs repositories
12:20 masak ya :)
12:21 Alias_ http://ali.as/cgi-bin/demo/stable/cv​smonitor/cvsmonitor.pl?cmd=viewBrows​eModule&amp;module=Alias.OpenSource
12:21 Alias_ There's a good example
12:21 lambdabot Title: CVS Monitor - Browsing Alias.OpenSource, http://tinyurl.com/258vby
12:21 Alias_ My open source coding, until I migrated it to svn
12:21 Alias_ Actually, that's boring
12:21 Alias_ Let me find you an interesting one
12:22 Alias_ http://ali.as/cgi-bin/demo/stable/cvsmonitor/cvs​monitor.pl?cmd=viewBrowseModule&amp;module=E.e17
12:22 masak nifty
12:22 Alias_ Enlightenment :)
12:22 lambdabot Title: CVS Monitor - Browsing E.e17, http://tinyurl.com/chbtn
12:22 Alias_ That's a better idea of the interesting stuff that results
12:22 masak I wish I'd had that for my projects
12:22 masak you were saying it doesn't work for svn?
12:22 Alias_ Nah, I never ported it
12:23 masak that's a shame
12:23 Alias_ It's a priorities thing
12:23 masak 'course
12:23 Alias_ Someone else could, fairly likely
12:23 Alias_ I ended up working on PPI and PITA and CPAN upgrades and a 100 bazillion modules instead
12:23 Alias_ Neatest part though
12:23 Alias_ It worked as a CVS client
12:23 Alias_ So it doesn't have to run on the server, you can point it at anything you have anonymous read rights to
12:23 masak cool
12:24 masak sounds like the Right Thing
12:24 masak should be able to do the same thing with svn
12:24 Alias_ That second URL is "So WTF even happened to the Enlightenment Window Manager Anyways?"
12:24 Alias_ I already broken the repository back-end code up into a driver API
12:24 Alias_ I've
12:24 Alias_ So it's mostly a case of someone filling out the svn stubs
12:25 Alias_ And negotiating my horrible old-style CGI code
12:25 Alias_ I was asked to write a Perforce backend at one point
12:25 Alias_ But they never came through with the money
12:25 Alias_ Otherwise we could point it at the Perl 5 source :)
12:26 masak & # afk
12:26 agentzh Alias_: neat
12:27 Alias_ Someone later wrote a commercial clone in Java that I believe supports SVN
12:27 Alias_ Dunno if it does the "point it at other people's code" thing though
12:27 agentzh heh
12:27 agentzh pointing it at pugs' code...
12:27 Alias_ which was always the killer feature for me :)
12:27 Alias_ pugs is in CVS?
12:27 agentzh svn
12:27 Alias_ meh
12:27 Alias_ If there's a svn2cvs around you can spit it into, I can do it though
12:28 agentzh GNU make's repos is in CVS though
12:28 Alias_ Where's that?
12:28 Alias_ I'll add it now :)
12:28 agentzh @google GNU make
12:28 lambdabot http://www.gnu.org/software/make/
12:28 lambdabot Title: GNU Make - GNU Project - Free Software Foundation (FSF)
12:29 Alias_ hmm
12:29 Alias_ ok, adding it :)
12:29 agentzh Alias++
12:29 Alias_ Man, it's old, but I love CVSMonitor still
12:29 Alias_ One of my favourite orphans
12:30 agentzh understood :)
12:30 Alias_ heh, I still remember the admin password too
12:30 agentzh Alias_: your server is kinda slow on my side...
12:31 Alias_ CVSMonitor is kind of a resource hog
12:31 obvio171 joined #perl6
12:31 Alias_ REALLY
12:31 Alias_ Like 200meg for those enlightenment pages
12:31 agentzh wow
12:31 Alias_ (ram)
12:31 Alias_ It's my test case for farging enormous repositories
12:32 Alias_ Takes over an hour to pull all the metadata
12:32 agentzh *impressed*
12:32 Alias_ <-- kicks ass when he finds a good idea
12:32 Alias_ ok, repository self-tests ok and added
12:32 Alias_ Let me find out what module names they use
12:32 agentzh Alias_: i've also something with pugs's repos using svk
12:33 agentzh Alias_: i'll show you the page...
12:33 Alias_ hang a sec, adding gnu :)
12:33 Alias_ BTW, because it's such a heavy app, look at the bottom of the cvsmon page :)
12:33 Alias_ I actually had to add that warning
12:33 agentzh Alias_: http://perlcabal.org/agent/stats/report.html
12:33 lambdabot Title: report - 数据挖掘报告
12:33 Alias_ My server got melted at one point
12:34 agentzh Alias_: you can ignore the chinese chars in it :)
12:34 agentzh Alias_: it's a static page rather than a cgi
12:34 Alias_ ya
12:34 Alias_ Mine is sorta split
12:35 agentzh Alias_: "Please do not SlashDot this application" ?
12:35 Alias_ There's an admin backgrounded process that does metadata pulling and analysis
12:35 Alias_ And it generates a giant Storable structure
12:35 Alias_ And saves it
12:35 agentzh Alias_: *nod*
12:35 agentzh it indeed costs a lot of resource...
12:35 Alias_ Storable is 95% of my process time :)
12:35 Alias_ Or something rediculous
12:36 agentzh Alias_: why not use RMDB?
12:36 Alias_ RMDB?
12:36 agentzh sorry, i mean relational data base
12:36 Alias_ ah
12:36 Alias_ Because 1) Probably slower
12:36 Alias_ 2) SQLite didn't exist at the time
12:37 Alias_ 3) Circular and complex data structures
12:37 agentzh fair enough
12:37 Alias_ It just turned out the only easy way to do it was generate the huge object tree and storable it
12:37 Alias_ hmm
12:37 Alias_ the source browser is dead
12:38 Alias_ So I don't know what modules exist in the make cvs
12:38 agentzh ah
12:38 Alias_ That was always one of the problems with cvs
12:38 agentzh modules?
12:39 Alias_ The fact you have to know the name of the modules in the repository
12:39 agentzh it's odd
12:39 Alias_ There's not really an easy way to see them
12:39 Alias_ It's a legacy of it's design by accretion
12:39 agentzh hehe
12:39 agentzh is svn suffering from the same problem?
12:40 Alias_ nah
12:40 agentzh good
12:40 Alias_ svn fixes most of cvs' problems
12:40 agentzh svn++
12:40 Alias_ Adds a few new problems, but fixes most of them
12:40 Alias_ Well
12:40 Alias_ not really :)
12:40 Alias_ There's no such thing as a branch or a tag in svn
12:40 Alias_ Which complicates things a lot
12:40 Alias_ Means writing svn port for cvsmonitor is supremely hard
12:41 agentzh we have to add branches explicitly in svn?
12:41 Alias_ branches are implemented as policy
12:41 Alias_ They only exist in wetware
12:41 Alias_ The software has no idea what a "branch" is
12:41 Alias_ It just sees copies
12:41 agentzh aye
12:41 Alias_ Making a Tag: svn copy foo bar
12:41 Alias_ Making a Branch: svn copy foo bar
12:41 Alias_ Making a Copy: svn copy foo bar
12:41 Alias_ :)
12:42 agentzh *nod*
12:42 agentzh that's what i've been doing ;)
12:42 Alias_ CVS Monitor has a bunch of smarts to be able to tell added and removed lines
12:42 Alias_ It's what the karma scores are derived from
12:42 Alias_ And it's branch-aware, mostly
12:42 agentzh hmm
12:42 Alias_ Making a branch should be considered adding lines :)
12:42 agentzh that's cool
12:43 Alias_ yeah, it was :)
12:43 Alias_ er
12:43 Alias_ making a branch should NOT be considered adding lines
12:43 toshinori joined #perl6
12:43 Alias_ I'm going to guess there's a module called "src" in the make repository
12:44 agentzh heh
12:44 agentzh Alias_: btw, for a separate issue...PPI has problems parsing heredocs like <<\EOF
12:44 Alias_ WTF is <<\EOF?
12:44 Alias_ Man, _I_ have a problem parsing that
12:44 Alias_ No wonder PPI does
12:44 agentzh Alias_: an old p4 syntax
12:44 agentzh still supported by p5
12:44 Alias_ ugh
12:44 agentzh like <<'EOF'
12:45 agentzh make's test suite makes heavy use of this odd syntax.
12:45 Alias_ hmm
12:45 Alias_ You are welcome to add support :)
12:45 agentzh i was using PPI to convert make's test suite (in p4) to Test::Base and found the problem...
12:45 agentzh heh
12:45 Alias_ The heredoc area of PPI is _relatively_ self-contained
12:45 toshinori joined #perl6
12:46 agentzh Alias_: and heredoc gives problem to regen p5 code from PDOM trees
12:46 Alias_ hmm?
12:46 Alias_ It should be mostly ok
12:46 Alias_ If you are using ->serialize
12:46 agentzh Alias_: i can't use ->content directly
12:46 Alias_ No, you can't
12:46 agentzh is there a serialize method?
12:46 agentzh sorry, i'm not aware of that.
12:46 Alias_ Because ->content ignores "rolled up" lines
12:46 agentzh Alias_: yes, i know
12:46 Alias_ heredoc content is not considered part of the content
12:47 Alias_ It's sort of "rolled up"
12:47 Alias_ Because it's line-oriented
12:47 Alias_ So it exists in a special location in the tree
12:47 Alias_ And when you write to a file it gets "unrolled" to the line stream
12:47 agentzh yes, i've read this from the PPI docs :)
12:47 Alias_ Sp ->serialize is the method you use to serialize the Perl code :)
12:47 Alias_ ->content is mostly a convenience
12:48 Alias_ So
12:48 agentzh serialize must be new method?
12:48 Alias_ No, been there since 1.00
12:48 Alias_ "rollups" are also where stuff inside of "use v6" sections will go
12:48 agentzh weird...which class?
12:48 Alias_ PPI::Document
12:48 agentzh thanks
12:49 Alias_ Because you can't serialize fragments of perl
12:49 Alias_ The concept is illogical
12:49 baest joined #perl6
12:49 Alias_ Well, not arbitrary trees anyways
12:50 agentzh k
12:51 agentzh Alias_: PPI makes a lot of apps possible :)
12:51 Alias_ blame heredocs :)
12:51 Alias_ heh
12:51 Alias_ Yes, yes it does
12:51 agentzh yup
12:51 Alias_ The supreme irony is that, having written it, I do almost nothing with it :)
12:51 Alias_ ppi_version is my main usage
12:51 Alias_ Well, perlver too I guess
12:51 Alias_ So I use it here and there
12:52 Alias_ But not to the extent I probably could be
12:52 agentzh lol
12:52 Alias_ I don't trust it yet
12:52 Alias_ For some reason there's a half dozen files in CPAN that explode it's memory usage
12:52 agentzh for example?
12:53 Alias_ When you are running all of CPAN, knowing which one is exploding your memory usage is hard
12:53 Alias_ Because it takes 6 hours of processing to get to that point :)
12:53 Alias_ I've never actually worked out which one it is
12:53 agentzh okay
12:53 Alias_ When I get back to working on CPAN::Metric I'll work it out
12:53 agentzh i wonder if you have a $day_job ;)
12:53 Alias_ ya
12:53 Alias_ I do
12:54 agentzh you seem to have done so much work
12:54 Alias_ Well, my general policy is to make the two align
12:54 Alias_ If I need something for work, I stall a bit and do it as a CPAN module
12:54 agentzh good point
12:54 Alias_ And I've been careful to insist on the right to do that when I get jobs
12:54 * agentzh will do this too!
12:54 Alias_ And really, it's not THAT much code
12:54 Yaakov Hello, Alias_.
12:55 Alias_ sloccount says my entire repository for CPAN is only 66,000 lines of code
12:55 Alias_ I reuse a lot of code
12:55 Alias_ And most of my CPAN stuff is small
12:55 * agentzh reuses Alias's code
12:55 agentzh my UML::Class::Simple is also small
12:55 agentzh because your Class::Inspector does most of the stuff for me :)
12:55 Alias_ :)
12:56 * Alias_ was really proud to see the PPI tree model get stolen btw
12:56 agentzh it's a reuse chain...hehe
12:56 Alias_ Was that you?
12:56 Alias_ For JavaScript or something...
12:56 agentzh Alias_: yup
12:56 Alias_ ya
12:56 Alias_ Was really excited to see it :)
12:56 agentzh Alias_: 50% of the GNU make test suite is passing for my makefile library
12:56 Alias_ ah right
12:56 Alias_ Make syntax, that's right
12:57 agentzh which has a PPI-like tree
12:57 Alias_ Now I remember
12:57 Alias_ Wasn't exactly sure which language
12:57 agentzh i haven't got cycles to upload the stuff to cpan
12:57 Alias_ See, that's another reason I get a lot done
12:57 Alias_ ruthless release automation
12:57 agentzh ah
12:57 agentzh a script?
12:57 Alias_ perl build_cpan.pl --install --commit Module::Name
12:58 agentzh awesome
12:58 agentzh i'll write one for myself ;)
12:58 Alias_ Runs QA on it, adds needed files automagically, adds author tests, cleans up newlines and stuff, runs the tests, checks it, reruns them, make tardists, adds the dist to the svn repository, commits, then gives me the URL for the location of the tarball in the repository
12:59 Alias_ And I have a firefox bookmark to upload a URI to pause
12:59 agentzh omg
12:59 Alias_ So "copy" from putty, and in firefox "release (paste URI)"
12:59 Alias_ And PAUSE pulls the distribution tarball straight out of SVN :)
12:59 agentzh that's so cool
13:00 moritz Alias_: you're a geek ;-)
13:00 Alias_ http://svn.ali.as/cpan/tools/build_cpan.pl
13:00 Alias_ moritz: It's entirely practical, I swear. :)
13:00 Alias_ Do you have any idea how many releases I've done :)
13:00 Alias_ Shiteloads
13:00 moritz Alias_: that's no contradiction
13:00 Alias_ http://svn.ali.as/cpan/releases/
13:01 lambdabot Title: Revision 1902: /releases
13:01 Alias_ That's just since I moved to svn
13:01 Alias_ So yeah, without the slowly evolving automation, I could never manage it
13:01 Alias_ I'm working on my own equivalent of Insurrection or CommitBit so I can add maintainers more easily :)
13:02 Alias_ http://svn.ali.as/why.jpg
13:02 wolverian does it work well with opera mini? :)
13:02 Alias_ Probably
13:02 wolverian (opera mini)++ # much better than the horrible browser your mobile comes with
13:02 wolverian (opera mini)-- # goodbye privacy
13:03 Alias_ wolverian: I also tested svn.ali.as today with the EVE Online in-game browser :)
13:03 wolverian heh
13:03 Alias_ So I can both fly space ships and add maintainers :)
13:03 Alias_ I'll add a screen shot later :)
13:03 wolverian I assume the wii browser works too
13:03 Alias_ I want to start a collection
13:03 mncharity joined #perl6
13:03 Alias_ Probably :)
13:03 Alias_ Got a wii and a camera?
13:03 wolverian no camera :/
13:05 * agentzh goes to work on a p5 version of Pugs::Grammar::Rule2.pm.
13:09 Alias_ agentzh, anyways, can you please submit a bug report on PPI for that heredoc feature you need?
13:09 Alias_ They generally get done
13:09 agentzh Alias_: will do later :)
13:09 Alias_ ok
13:09 Alias_ Cause I have a PPI timeslice coming up soonish
13:09 agentzh Alias_: i was just a bit lazy but will do, anyway :)
13:10 Alias_ And if you miss it, it will likely be a while till the next one
13:10 agentzh cute
13:10 Alias_ Audrey terminology
13:10 agentzh doing now
13:12 fglock joined #perl6
13:13 agentzh Alias_: done :)
13:13 Alias_ thanks
13:14 agentzh Alias_: thank *you* :)
13:14 Alias_ nah, no problem
13:14 Alias_ small syntax bugs are relatively easy
13:14 Alias_ And it's ironing out the nigglies that make it so much more useful day by day
13:14 Alias_ And keep people excited, when it handles even the stuff they think isn't possible
13:15 Alias_ Gotta maintain the illusion :)
13:15 agentzh yeah
13:15 agentzh exactly
13:16 Alias_ when I went to OSCON to release it, I made especially sure I had added support for all the cases I knew merlyn was gunna throw at me as things that perl parsers couldn't possibly handle :)
13:16 Alias_ And I was right :)
13:16 Alias_ I did a demo at pdx.pm and he started giving me examples, all the ones I expected :)
13:17 Alias_ First time I ever wiped the smirk right off him :)
13:17 agentzh *g*
13:17 Alias_ Actually, one of the highlights of my Perl career I think
13:17 Alias_ in retrospect
13:18 wolverian I hope he appreciated it as much as you did :)
13:18 Alias_ oh sure
13:18 Alias_ Spent a great night getting sloshed with him at a martini bar in Melbourne last YAPC::AU
13:18 Alias_ And he's one of the few people that actually understand why PPI works at the zen level :)
13:18 rindolf joined #perl6
13:19 Alias_ I think it's like maybe 6 people
13:19 Alias_ me, merlyn, audrey, and a few others
13:19 wolverian there are bars just for martinis?
13:19 Alias_ wolverian: sure is :)
13:19 Alias_ In Melbourne at least
13:19 wolverian insane world
13:19 Alias_ hidden down an alley somewhere
13:19 Alias_ amazing place, really funky
13:19 Alias_ like out of another time
13:20 wolverian and then you google 'em and find out they're really popular and have a really web2.0 site
13:20 Alias_ oh sure
13:20 Alias_ But it still felt like another world
13:21 moritz why should a bar have a web2.0ish site?
13:21 wolverian meh, now my summer feels so dull :)
13:22 Alias_ heh
13:22 Alias_ moritz: Because that's the the generic design company sold them, using Frontpage.NET default Web 2.0 automation
13:22 Alias_ or something in that general situation
13:22 xdemondogx joined #perl6
13:24 moritz omfg
13:24 xdemondogx ?
13:24 cerridwen joined #perl6
13:25 kanru joined #perl6
13:25 moritz !
13:26 perlMa joined #perl6
13:27 Alias_ ?
13:27 xdemondogx ?
13:27 moritz !!?
13:27 devbot6 moritz: Error: "!?" is not a valid command.
13:28 xdemondogx ??
13:28 moritz devbot6: and ! is?
13:28 devbot6 moritz: Error: "and" is not a valid command.
13:28 moritz devbot6: help
13:28 devbot6 moritz: (help [<plugin>] [<command>]) -- This command gives a useful description of what <command> does. <plugin> is only necessary if the command is in more than one plugin.
13:29 moritz wow, that was helpful
13:29 wolverian devbot6, list
13:29 devbot6 wolverian: Admin, Channel, Config, Misc, Owner, RSS, Services, TracBot, User, and Web
13:29 svnbot6 joined #perl6
13:29 svnbot6 joined #perl6
13:29 moritz devbot6: help User
13:29 devbot6 moritz: Error: There is no command "user".
13:30 moritz that would have been too easy
13:30 wolverian devbot6, list User
13:30 devbot6 wolverian: capabilities, changename, hostmask, hostmask add, hostmask list, hostmask remove, identify, list, register, set password, set secure, stats, unidentify, unregister, username, and whoami
13:30 moritz !RSS
13:30 devbot6 moritz: (rss <url> [<number of headlines>]) -- Gets the title components of the given RSS feed. If <number of headlines> is given, return only that many headlines.
13:30 wolverian see, it's logical, if you know the logic
13:30 wolverian (;
13:34 moritz devbot6: help help
13:34 devbot6 moritz: (help [<plugin>] [<command>]) -- This command gives a useful description of what <command> does. <plugin> is only necessary if the command is in more than one plugin.
13:38 thorat joined #perl6
13:40 svnbot6 r16909 | agentz++ | trunk:
13:40 svnbot6 : [PCR]
13:40 svnbot6 : - added examples/langs2.grammar to demonstrate the use of
13:40 svnbot6 :   the "rule foo { ... }" syntax.
13:40 svnbot6 : - added a test to 16-examples.t for langs2.grammar.
13:40 svnbot6 : - fixed emitter/00-ratchet.t
13:40 pugs_svnbot r16909 | agentz++ | trunk:
13:40 pugs_svnbot : [PCR]
13:40 pugs_svnbot : - added examples/langs2.grammar to demonstrate the use of
13:40 pugs_svnbot :   the "rule foo { ... }" syntax.
13:40 pugs_svnbot : - added a test to 16-examples.t for langs2.grammar.
13:40 pugs_svnbot : - fixed emitter/00-ratchet.t
13:40 pugs_svnbot diff: http://dev.pugscode.org/changeset/16909
13:40 lambdabot Title: Changeset 16909 - Pugs - Trac
13:40 agentzh weird...
13:40 agentzh two bots are reporting the same commit?
13:41 agentzh well, anyway, my karma has doubled :)
13:42 fglock agentzh: I found the bug in your paste; there is an error in the parameter order for calling subrules
13:42 lambdabot fglock: You have 1 new message. '/msg lambdabot @messages' to read it.
13:43 agentzh fglock: you mean the adder sample?
13:43 fglock agentzh: re release: congrats :)
13:43 agentzh fglock: thanks :)
13:44 drupek12 joined #perl6
13:44 fglock the digit counter example; I'll try to fix it
13:44 agentzh fglock: yay!
13:45 fglock I wonder why I didn't notice that before
13:45 agentzh fglock: because there's not enough tests! haha!
13:46 moritz maybe audreyt restartet svnbot6 assuming it was down?
13:47 agentzh audreyt++
13:48 moritz diakopter++ # managing bots, cleaning up irc logs
13:49 agentzh fglock: will try to do another CPAN release for PCR tonight :)
13:49 moritz 1771 lines are marked as spam atm
13:50 agentzh moritz: maybe it's your cute log site attracts those spammers?
13:50 agentzh *g*
13:50 agentzh *that attracts
13:53 agentzh fglock: the counter example is examples/digits.grammar
13:53 fglock yes
13:54 agentzh fglock: and the corresponding test is in t/16-examples.t
13:54 agentzh fglock: currently it's skipped.
13:54 agentzh but it's looping...
13:54 agentzh hehe
13:54 agentzh *because
13:56 agentzh fglock: currently PCR's regex parser doesn't do any error handling. i'll look into v6.pm's parser since it seems to have some diagnostics for invalid inputs
13:57 Alias_ You are lost in a twisty series of grammars, all of them alike
13:57 moritz where do you want to turn (NWSE)?
13:58 * Alias_ read NSFW for a sec there
13:59 iblechbot joined #perl6
14:01 rindolf Hi Alias_
14:01 Alias_ heya
14:01 Alias_ How's things
14:01 rindolf Alias_: what's up?
14:01 rindolf Alias_: I'm fine.
14:01 rindolf Alias_: worked on XML::SemanticDiff.
14:01 Alias_ Hey, were you looking for maint on something? Did it ever get done?
14:02 rindolf Alias_: and then Kip Hamton (its author) re-surfaced.
14:02 Alias_ So did he set maint for you ok?
14:02 Alias_ All happy?
14:02 rindolf Alias_: you mean co-maint for X-SD?
14:02 rindolf Alias_: well, actually, someone else did.
14:02 rindolf Alias_: from the CPAN Cabal.
14:02 Alias_ I am the CPAN Cabal :)
14:03 Alias_ well, not really
14:03 Alias_ But if there's nothing I need to do, that's fine
14:04 svnbot6 r16910 | fglock++ | trunk:
14:04 svnbot6 : [PCR] fixed a bug in the parameter ordering - agentzh++
14:04 pugs_svnbot r16910 | fglock++ | trunk:
14:04 pugs_svnbot : [PCR] fixed a bug in the parameter ordering - agentzh++
14:04 pugs_svnbot diff: http://dev.pugscode.org/changeset/16910
14:04 lambdabot Title: Changeset 16910 - Pugs - Trac
14:04 agentzh fglock++
14:04 rindolf Alias_: what time is it there?
14:04 agentzh wow, bots are racing..
14:05 Alias_ rindolf: midnight'ish
14:05 rindolf Alias_: I see.
14:08 pasteling "evalbot_r16909" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26113
14:19 fglock agentzh: did it work? I'm switching back to kp6... :)
14:19 agentzh fglock: it works :)
14:20 agentzh fglock: could you document the subrule calling protocol somewhere when you have spare time?
14:20 agentzh fglock: such as parameter list and return values...
14:20 agentzh fglock: i'm having problems in understanding it
14:25 svnbot6 r16911 | agentz++ | trunk:
14:25 svnbot6 : [PCR]
14:25 svnbot6 : - updated examples/digits since fglock++ has already fixed it.
14:25 svnbot6 : - unSKIP'd the test in 16-examples.t for examples/digits.
14:25 svnbot6 : - updated emitter/00-ratchet.t since the emitter template has changed.
14:25 pugs_svnbot r16911 | agentz++ | trunk:
14:25 pugs_svnbot : [PCR]
14:25 pugs_svnbot : - updated examples/digits since fglock++ has already fixed it.
14:25 pugs_svnbot : - unSKIP'd the test in 16-examples.t for examples/digits.
14:25 pugs_svnbot : - updated emitter/00-ratchet.t since the emitter template has changed.
14:25 pugs_svnbot diff: http://dev.pugscode.org/changeset/16911
14:25 lambdabot Title: Changeset 16911 - Pugs - Trac
14:41 fglock agentzh: the calling protocol changed 2 times already - at Compiler::Regex near lines 143 and 150 there are some backward-compatibility hacks and a little documentation
14:41 fglock it needs a cleanup
14:42 agentzh fglock: *nod*
14:47 masak joined #perl6
14:47 masak ok, correct me if I'm wrong, but...
14:47 masak t/data_types/array.t, test 53, "negative index [-1]"
14:47 masak that's got to be out-of-date, no?
14:49 masak it does this
14:50 masak ?eval my @a = <a b c e>; @a[-1]
14:50 evalbot_r16908 changed the nick to evalbot_r16911
14:50 evalbot_r16911 \"e"
14:50 masak ...which I'm pretty sure is written '@a[ *-1 ]' these days
14:50 moritz ...which is pretty ugly ;)
14:51 b_jonas joined #perl6
14:51 masak moritz: but still
14:51 masak it's correct perl6, as the changing standard goes
14:51 moritz masak: aye
14:51 masak what's in the test isn't
14:51 moritz masak: then change it
14:51 masak ok :)
14:51 masak that's all I needed to hear
14:56 TimToady should likely also have a test that @a[-1] fails properly
14:57 masak TimToady: right. I'll work that in now
14:57 svnbot6 r16912 | masak++ | trunk:
14:57 svnbot6 : [t/data_types/array.t]
14:57 svnbot6 : * "negative index" -1 etc are now "end index" [*-1] etc
14:57 pugs_svnbot r16912 | masak++ | trunk:
14:57 pugs_svnbot : [t/data_types/array.t]
14:57 pugs_svnbot : * "negative index" -1 etc are now "end index" [*-1] etc
14:57 pugs_svnbot diff: http://dev.pugscode.org/changeset/16912
14:57 lambdabot Title: Changeset 16912 - Pugs - Trac
14:57 masak two bots! svnbot6++ pugs_svnbot++ :)
14:59 masak TimToady: so even reading of @a[-1] fails? or just writing?
14:59 pasteling "evalbot_r16911" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26114
15:00 masak oops. was that me?
15:00 Limbic_Region joined #perl6
15:00 agentzh masak: not likely
15:00 masak I only ever changed a .t file...
15:01 masak agentzh: good. carrying on :)
15:01 agentzh masak: we've been seeing this for months
15:01 masak agentzh: and no explanation?
15:01 agentzh masak: a messed-up pugs tree on feather maybe
15:01 masak agentzh: could be
15:02 masak agentzh: does no-one have privs to go look?
15:02 agentzh masak: many people have tried but with no luck so far
15:03 agentzh masak: or maybe you can work out the puzzle?
15:03 masak agentzh: oki
15:03 agentzh hehe
15:03 masak agentzh: not likely. for starters, I don't have a feather account
15:03 agentzh k
15:04 masak TimToady: I just went ahead and assumed now that even reading from [-1] on a normal array is fatal
15:04 svnbot6 r16913 | masak++ | trunk:
15:04 svnbot6 : added back a few failing -1 tests into t/data_types/array.t
15:04 pugs_svnbot r16913 | masak++ | trunk:
15:04 pugs_svnbot : added back a few failing -1 tests into t/data_types/array.t
15:04 pugs_svnbot diff: http://dev.pugscode.org/changeset/16913
15:04 lambdabot Title: Changeset 16913 - Pugs - Trac
15:05 masak maybe that was to harsh; in that case I'll be happy to change it to something less harsh, like a normal undefined value
15:05 masak s/to/too/
15:07 masak actually, I could argue either way there
15:08 masak on one hand, accessing past the end of the array in the other direction (i.e. going above @a.end) only returns undef
15:08 TimToady it returns an unthrown exception, which is a variety of undef
15:08 TimToady but if you try to use the value, it should throw an exception
15:09 masak ok, so dies_ok is correct in the .t file for that?
15:09 TimToady as long as you do something like @a[-1] + 0 inside
15:09 masak ah -- no, right now only @a[-1]
15:09 masak I'll go change it
15:10 TimToady the direct result would merely test as undefined
15:10 TimToady biab &
15:10 masak oki
15:10 masak adding another test for that
15:13 masak hehe, now the test file hangs instead
15:13 svnbot6 r16914 | masak++ | trunk:
15:13 svnbot6 : changed t/data_types/array.t slightly after clarification
15:13 svnbot6 : from TimToady++
15:13 pugs_svnbot r16914 | masak++ | trunk:
15:13 pugs_svnbot : changed t/data_types/array.t slightly after clarification
15:13 pugs_svnbot : from TimToady++
15:13 pugs_svnbot diff: http://dev.pugscode.org/changeset/16914
15:13 lambdabot Title: Changeset 16914 - Pugs - Trac
15:14 masak it probably hangs because @a[*-4 .. *-2] is treated as something to do with infinity...
15:15 pasteling "evalbot_r16913" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26115
15:15 masak hm, sometimes I find it tedious to keep the plan figure at the top of test files updated
15:15 masak wouldn't it be possible to write a simple script to count the tests?
15:16 masak one that took into account if blocks and for loops
15:16 masak I think such a script could avoid the Halting Problem issue on account of .t files never going above a certain complexity
15:17 masak specifically, they never contain infinite loops with test method calls
15:17 nik joined #perl6
15:17 pugs_svnbot r16915 | agentz++ | trunk:
15:17 pugs_svnbot : PCR - compile_p6grammar.pl - adds header comments to the generated p5 files
15:17 pugs_svnbot diff: http://dev.pugscode.org/changeset/16915
15:17 lambdabot Title: Changeset 16915 - Pugs - Trac
15:19 svnbot6 r16915 | agentz++ | trunk:
15:19 svnbot6 : PCR - compile_p6grammar.pl - adds header comments to the generated p5 files
15:19 masak agentzh: there are a couple of unrecognized files in the perl5/Pugs-Compiler-Perl6/t directory -- is it ok if I svn:ignore those?
15:19 agentzh really? looking
15:20 masak it's 01-sanity, builtins, Test.pm, 02-test-pm and data_types/array_extending.t
15:20 masak at least on my box, but maybe I haven't cleaned lately
15:20 LimbicRegion joined #perl6
15:20 agentzh ah, you mean pugs's t/
15:20 masak yes
15:21 agentzh i thought it was PCR's
15:21 masak no, wait, it is
15:21 agentzh okay, it's v6.pm's
15:21 agentzh i was just reading too fast
15:21 agentzh sorry
15:21 masak they're not there when I remove the directory structure and svn up
15:22 masak so they're probably generated by make
15:22 agentzh masak: they're copied from pugs's t/
15:22 agentzh clkao worked that out iirc
15:22 agentzh they're mostly passing tests for v6.pm in the pugs test suite
15:23 masak agentzh: but I could just prop them as ignored so they don't show up when one does `svn st`?
15:23 agentzh masak: use svn st -q
15:23 dwave joined #perl6
15:23 masak agentzh: why? I know what it does, but why should I have to?
15:24 agentzh masak: setting them as ignored is fine
15:24 masak great. I'll do that
15:28 svnbot6 r16916 | masak++ | trunk:
15:28 svnbot6 : props
15:28 pugs_svnbot r16916 | masak++ | trunk:
15:28 pugs_svnbot : props
15:28 pugs_svnbot diff: http://dev.pugscode.org/changeset/16916
15:28 lambdabot Title: Changeset 16916 - Pugs - Trac
15:30 pasteling "evalbot_r16915" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26117
15:36 svnbot6 r16917 | masak++ | trunk:
15:36 svnbot6 : hiding autogenerated perl5/Pugs-Grammar-MiniPerl6/​lib/Pugs/Grammar/MiniPerl6.pm too
15:36 pugs_svnbot r16917 | masak++ | trunk:
15:36 pugs_svnbot : hiding autogenerated perl5/Pugs-Grammar-MiniPerl6/​lib/Pugs/Grammar/MiniPerl6.pm too
15:36 pugs_svnbot diff: http://dev.pugscode.org/changeset/16917
15:36 lambdabot Title: Changeset 16917 - Pugs - Trac
15:37 masak I'm peeking inside Pugs internals now
15:37 masak does anyone have a pointer to what the difference between IArray and VArray is?
15:38 LimbicRegion changed the nick to Limbic_Region
15:39 Jameswf 13 char
15:40 masak ?eval "V".ord - "I".ord
15:40 evalbot_r16911 changed the nick to evalbot_r16917
15:40 evalbot_r16917 13
15:40 masak Jameswf: I get it. Not quite what I was looking for, though
15:40 masak :)
15:41 masak if I know audreyt right, it's something more substantial than that
15:43 masak here's another 'cute' way to do it:
15:43 masak ?eval [-] map :{ .ord } <v i>
15:43 evalbot_r16917 13
15:44 Jameswf IArray $sweeet[1] #index VArray $sweeet[juice] #value
15:44 Jameswf my guess
15:44 masak Jameswf: please explain more verbosely than that
15:45 masak what's the second one? it doesn't parse in my brain
15:45 masak an index is a value too, yes?
15:46 Jameswf you can call an arrays contents my its indes 0-sizeof() or by value if you have set that so an array gas_prices[newyork] can also be gas_prices[0]
15:46 Jameswf its how you make the call
15:47 masak Jameswf: you sure you're not thinking of PHP now?
15:47 masak that doesn't sound like Perl 6 to me
15:49 masak because (1) it's not a call, (2) you can't use a bareword inside the index [] like that, unless it's a defined method, and (3) using $ before the array is Perl5 (and PHP), not Perl 6
15:49 Jameswf it all sort of blends php perl c its like 40 different dialects of spanish, Some times you speak mexican to a puortorican
15:49 masak :)
15:49 masak to me they're very distinct
15:50 masak maybe not syntactically, but spiritually
15:50 masak when I use PHP, I feel dead; when I use Perl, I feel alive :)
15:50 Jameswf I was going for context more than syntaf thats why there were no ;
15:50 Jameswf *syntax
15:50 masak Jameswf: anyway, I don't think that's it
15:51 masak this is internals we're talking about, so it's two different types of array being used inside Pugs
15:51 masak how you index a particular array should not come into the question
15:53 vel joined #perl6
15:54 Jameswf I cant find the terms on google except used as array names
15:54 masak I'm in src/Pugs/AST/Internals.hs now
15:55 masak there's a lot of types defined here starting with I
15:55 masak and many starting with V
15:55 masak trying to figure out what sets them apart
16:01 spinclad 'when 6.28 comes, every day will be Thursday'...
16:02 masak spinclad: :P
16:02 masak that's a great slogan
16:04 masak especially along with Arthur Dent's "This must be Thursday. I could never get the hang of Thursdays."
16:05 spinclad 'when i catch up with the backlog, every time will be 08:10'
16:06 spinclad (12:10 UTC)
16:06 masak I also find a lot of SV and V suffixes in the names in the Pugs internals
16:07 spinclad yeah, or Belgium
16:07 masak they remind me of something from Perl 5 internals
16:11 masak & # dinner
16:12 araujo joined #perl6
16:24 offby1 joined #perl6
16:30 vel joined #perl6
16:38 BooK__ changed the nick to BooK
16:47 justatheory joined #perl6
16:50 fglock joined #perl6
16:53 weinig joined #perl6
16:53 svnbot6 r16918 | fglock++ | trunk:
16:53 svnbot6 : [kp6] p5regex - implemented positional captures
16:53 pugs_svnbot r16918 | fglock++ | trunk:
16:53 pugs_svnbot : [kp6] p5regex - implemented positional captures
16:53 pugs_svnbot diff: http://dev.pugscode.org/changeset/16918
16:53 lambdabot Title: Changeset 16918 - Pugs - Trac
16:57 pasteling "evalbot_r16917" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26118
17:08 rashakil joined #perl6
17:10 Schwern joined #perl6
17:14 svnbot6 r16919 | fglock++ | trunk:
17:14 svnbot6 : [kp6] p5regex - quantified positionals capture to an array
17:14 pugs_svnbot r16919 | fglock++ | trunk:
17:14 pugs_svnbot : [kp6] p5regex - quantified positionals capture to an array
17:14 pugs_svnbot diff: http://dev.pugscode.org/changeset/16919
17:14 lambdabot Title: Changeset 16919 - Pugs - Trac
17:19 TimToady ?eval [-] map :{ .ord } <v i>
17:19 evalbot_r16917 changed the nick to evalbot_r16919
17:19 evalbot_r16919 13
17:19 TimToady that should not parse
17:19 TimToady ?eval [-] map { .ord }, <v i>
17:19 evalbot_r16919 13
17:20 vel joined #perl6
17:20 TimToady ?eval [-] map <v i>, :{ .ord }
17:20 evalbot_r16919 13
17:20 TimToady but a :{} where a term is expected should be expecting an infix after it.
17:20 TimToady so it should have taken < as a less than
17:21 TimToady so probably parsec being overly backtracky...
17:22 TimToady @tell masak I think probably @a[-1] with a literal negative should be a compilation failure
17:22 lambdabot Consider it noted.
17:24 TimToady @tell masak in order to catch a frequent P5Think error sooner than runtime.  But @a[$negative] is still a runtime failure
17:24 lambdabot Consider it noted.
17:40 vel joined #perl6
17:41 Psyche^ joined #perl6
17:55 Psyche^ changed the nick to Patterner
17:56 fglock just checking - named assertions don't capture, right? like <before ...>
17:57 fglock or do I need <?before ...> for that
17:57 vel6608 joined #perl6
17:59 fglock are captures-inside-assertions discarded? as in <before (...) >
18:00 vel joined #perl6
18:03 fglock nevermind, I found the answer to my question
18:04 masak joined #perl6
18:26 Belaf joined #perl6
18:31 [particle] joined #perl6
18:38 Squern joined #perl6
18:46 vadrer joined #perl6
18:48 vadrer ?eval say 'hello';
18:48 evalbot_r16919 OUTPUT[hello␤] Bool::True
18:49 masak ?eval 'hello'
18:49 evalbot_r16919 "hello"
18:49 smg ?eval print for <1 2 3 4 5>
18:49 evalbot_r16919 OUTPUT[] undef
18:49 smg ?eval map { print } <1 2 3 4 5>
18:49 evalbot_r16919 Error: ␤Unexpected "2"␤expecting operator
18:49 masak ?eval [~] <1 2 3 4 5>
18:50 evalbot_r16919 "12345"
18:50 smg is this lisp or perl?
18:50 masak smg: perl 6
18:50 lambdabot masak: You have 2 new messages. '/msg lambdabot @messages' to read them.
18:50 smg what is [~] ?
18:50 masak wee, messages!
18:50 smg masak: tell what's [~] pls
18:50 masak @messages
18:50 lambdabot TimToady said 1h 28m 4s ago: I think probably @a[-1] with a literal negative should be a compilation failure
18:50 lambdabot TimToady said 1h 26m 36s ago: in order to catch a frequent P5Think error sooner than runtime. But @a[$negative] is still a runtime failure
18:50 xwrn joined #perl6
18:50 masak smg: ~ is string concatenation
18:51 masak like the old . in perl 5
18:51 smg lol
18:51 smg but what does ~ concatenate then?
18:51 masak [~] takes a whole list and concatenates all the items
18:51 smg k
18:51 smg like join "" ?
18:51 masak smg: ~ concatenates strings
18:51 masak smg: yes
18:51 smg damn
18:52 smg is there a page for all changes from perl 5 => perl 6?
18:52 masak @tell TimToady re @a[-1]: ok. that's half of what I thought too
18:52 lambdabot Consider it noted.
18:52 masak @tell TimToady I will change the test to reflect this
18:52 lambdabot Consider it noted.
18:53 masak smg: yes. let me find it for you
18:53 smg thx
18:53 smg i'm too lazy atm
18:53 smg http://feather.perl6.nl/syn/Differences.html <-this?
18:53 lambdabot Title: Perl6::Perl5::Differences
18:53 smg lambdabot: true
18:53 masak not the one I was thinking about
18:54 smg okay then find the better one from your memory pls for me
18:54 smg :)
18:54 masak here you go: http://svn.pugscode.org/pug​s/docs/other/porting_howto
18:54 masak but even better is to download pugs and look at all the tests
18:55 masak they tell the whole story about the syntax and semantics
18:55 smg all the tests
18:55 smg where can i find them?
18:56 masak in the pugs repository
18:56 smg mh
18:56 masak which you can download in a number of ways
18:56 masak check out pugscode.org
18:56 jjore smg, the [ ] around ~ turned it into a reduce operation.
18:56 smg lol
18:56 smg that's really evil
18:56 smg it's from haskell?
18:56 jjore it accepts any function btw.
18:56 jjore op, rather.
18:56 smg isn't [ ] anonymous array?
18:57 unobe_away changed the nick to unobe
18:57 jjore no, not at this place.
18:57 smg k
18:57 smg context?
18:57 jjore dunno. probably?
18:57 smg k
18:57 smg :)
18:57 smg i love perl 5
18:57 jjore I love perl 6 harder.
18:57 smg but i can't stand with perl 6 atm
18:57 kane_ joined #perl6
19:00 Tene smg: why?
19:00 Gothmog_ it takes some getting used to, but that was just as true for perl 5, and its worth it
19:00 Juerd it's
19:00 smg i dunno
19:00 smg :)
19:01 masak smg: I agree with Gothmog_. Perl 6 is still in the spirit of Perl.
19:01 smg it is
19:01 masak except it has more neat stuff to play with
19:01 Juerd I guess that depends on the definition of Perl's spirit :)
19:01 smg i will see
19:01 smg for the moment i dislike some major changes for perl 6
19:01 smg but maybe i get used to them
19:01 masak Juerd: yes, it kind of has to depend on that :)
19:01 Juerd I personally think some parts of Perl 6 are too complicated, and Perl 5 always felt simple.
19:02 jjore Perl 6 is just tons of great stuff that's way too hard to implement in Perl 5 as add-ons. *You* try writing macros in Perl 5.
19:02 jjore Blech.
19:02 masak smg: most of the changes have good reasons behind them
19:02 avar ha! perl5 felt simple?:)
19:02 Juerd masak: Hm... And which changes do not? :)
19:02 smg masak: i know, because i know the perl6 developers are good
19:02 Juerd avar: Yes. Simple, not easy.
19:02 masak Juerd: I think I agree on some points, junctions for example
19:02 jjore junctions are what you've been using in SQL for ages now!
19:02 smg masak: but in the first place they look uncomfortable, but i understand :)
19:02 avar Perl [56] are a lot of things, but I wouldn't say simple is one of them, at least not on the language level
19:03 Juerd masak: I think junctions are relatively simple. Multidimensional anything, however, argh. I just don't get that.
19:03 avar Maybe on the "there's lots of code written for you via cpan" level
19:03 masak Juerd: hm, which changes do not have good reasons behind them? maybe none. have to thing about that
19:03 Juerd avar: Not simple like PHP :)
19:03 Juerd avar: But Perl 5 is much simpler than Perl 6
19:03 masak smg: give it some time
19:03 b_jonas joined #perl6
19:03 Juerd avar: I think references are simpler than captures, for example
19:04 Juerd avar: And arrays of arrays, simpler than multidimensional arrays.
19:04 avar Juerd: No, the other way around. You can easily explain some things on p6 which would take whole days to explain in p5:)
19:04 avar Yeah maybe, but syntax and some other semantics are simpler
19:04 Juerd avar: I doubt it.
19:04 avar and some or not, *shrug*
19:04 b_jonas and lambdas simpler then p6 functions with prototypes and stuff, but that's what we get
19:04 Alias_ jjore: I for one don't think the new constructor stuff is something that's way too hard to implement in Perl 5
19:04 masak interesting discussion
19:04 Juerd Well, yes, some things. And those are probably the changes that I like :)
19:05 Alias_ I'm not a far
19:05 Alias_ fan
19:05 Juerd The other changes, I'll just take because I long for the good stuff :)
19:05 b_jonas multidimensional array can be a performance thing in some cases
19:05 Gothmog_ I think arrays of arrays are rather clumsy if you want something multidimensional.
19:05 jjore Alias_: ok, but the new constructor syntax is implemented in perl 5 as source filters.
19:05 jjore that sucks.
19:05 Juerd b_jonas: I still think that such things are nice to have, but as an extension rather than in the main language.
19:05 b_jonas especially if you have numeric arrays
19:05 Juerd b_jonas: Same for junctions, perhaps.
19:06 Alias_ jjore: Sure, but I don't like having it imposed on eme
19:06 Alias_ me
19:06 agentzh question: how to avoid warnings like "Name XXX used only once: possible typo"?
19:06 jjore I'd much prefer to have ASTs or similar to work with.
19:06 Alias_ ... because it's someone's favourite constructor stule
19:06 Alias_ style
19:06 b_jonas well, I don't like p6 junctions in code either, yep
19:06 Juerd agentzh: Use it twice or disable warnings :)
19:06 b_jonas agentzh: our can work in some cases
19:06 Alias_ agentzh: $foo = $foo :(
19:06 b_jonas our $foo that is
19:06 masak sometimes I'm afraid that junctions will not be introspectable enough
19:06 b_jonas but not always
19:06 agentzh Alias_: there's tons of such things...because it's generated code...
19:07 agentzh Juerd: no warnings may be a good idea
19:07 Juerd agentzh: That's what it's for.
19:08 diakopter_afk changed the nick to diakopter
19:08 b_jonas you can no warnings just specific kinds of warnings
19:09 amnesiac joined #perl6
19:12 agentzh b_jonas: what's the warning catagory for the "used only once" thing?
19:13 b_jonas agentzh: try "once"
19:13 agentzh k
19:13 agentzh b_jonas: it works, thanks :)
19:14 masak any tips on how to familiarize oneself with the internals of pugs?
19:15 masak I started today, but quickly ran into a lot of questions with no-one to ask
19:16 agentzh masak: it's a big project ;)
19:16 masak agentzh: so I gather.
19:16 agentzh masak: but i'm glad to see you've started :)
19:16 masak still, it's nice to know a little about what's going on under the hood
19:16 agentzh masak: one tip
19:16 masak my plan is to learn enough so that I can apply simple patches when I have the time
19:17 agentzh grep the irc logs over the years for audreyt, gaal, and other hs people's remarks
19:17 masak 'aight
19:17 agentzh it may helps
19:17 agentzh *help
19:17 masak maybe follow along with the svn logs/diffs as well...
19:18 agentzh masak: *nod*
19:18 * masak takes deep breath
19:18 * agentzh grins.
19:18 agentzh masak: maybe you'll become a second audreyt here and start hacking on src/
19:19 masak who knows? anything might happen
19:19 agentzh masak: yup
19:19 masak though at this point I doubt I could replace audreyt
19:19 masak I would have to quit my day job :)
19:19 agentzh lol
19:20 masak and even that would probably not be enough
19:20 agentzh fglock: caching is anti-testing...
19:20 agentzh masak: agreed
19:20 agentzh masak: you need huge hours...
19:20 masak I'm not worshipping audreyt, I'm just un deep awe
19:21 masak I hope she's soon well enough to carry on with pugs
19:21 agentzh fglock: yay! compile_p6grammar.pl generates a working Pugs::Grammar::Rule.pmc and all tests are passing :D
19:22 agentzh masak: me too
19:22 masak agentzh++ # cool
19:22 agentzh masak: thanks :)
19:22 agentzh it's much much faster...
19:22 agentzh let me time it...
19:23 agentzh $ time util/update-rule-pmc-new update
19:23 agentzh real    0m1.503s
19:23 agentzh user    0m1.416s
19:23 agentzh sys     0m0.080s
19:23 * diakopter will return later tonight to fix 1. the default vhost on feather (which I broke yesterday with some incomplete changes before someone/cron restarted apache2), 2. fix the auto-builds of pugs, 3. implement the super-improved dalek I planned this morning (details later).  bbl
19:23 agentzh while the traditional v6.pm approach takes much longer:
19:23 diakopter changed the nick to diakopter_afk
19:24 agentzh $ time util/update-rule-pmc updateGenerating lib/Pug/Grammar/Rule2.pmc...
19:24 agentzh real    0m7.176s
19:24 agentzh user    0m7.072s
19:24 agentzh sys     0m0.088s
19:24 masak about a 4-fold improvement, not bad
19:24 agentzh indeed!
19:25 masak what's the sikkrit?
19:25 agentzh sikkrit?
19:25 offby1 he means "Sekrit"
19:25 offby1 obviously can't spell :-)
19:25 offby1 as in, "sekrit password"
19:25 masak offby1: :)
19:25 agentzh ah
19:26 masak offby1: I honestly didn't know it had a correct inkurikt spelling
19:26 agentzh compile_p6grammar doesn't parse p6 code, it only parse p6 regexes
19:26 agentzh *parses
19:27 masak offby1: that's what you get for not being a native english speaker, I guess
19:27 masak agentzh: and that makes it faster?
19:27 agentzh masak: yes :)
19:27 agentzh compiling full perl 6 is slow...
19:27 agentzh while compiling just perl 6 regexes can be rather efficient.
19:28 masak ok
19:28 masak sounds like interesting code to look at...
19:28 agentzh the code in actions are treated as Perl 5
19:28 agentzh or source-filtered Perl 5 :)
19:28 agentzh i'll commit the stuff...
19:28 masak do so
19:30 agentzh a sec
19:30 agentzh it's a huge commit
19:32 svnbot6 r16920 | fglock++ | trunk:
19:32 svnbot6 : [kp6] p5regex - fixed quantified captures
19:32 svnbot6 : - added v6/docs/p6regex-on-p5regex.txt
19:32 pugs_svnbot r16920 | fglock++ | trunk:
19:32 pugs_svnbot : [kp6] p5regex - fixed quantified captures
19:32 pugs_svnbot : - added v6/docs/p6regex-on-p5regex.txt
19:32 pugs_svnbot diff: http://dev.pugscode.org/changeset/16920
19:32 lambdabot Title: Changeset 16920 - Pugs - Trac
19:33 masak it's always nice to see big commits...
19:34 agentzh still working on the log message...
19:35 pasteling "evalbot_r16919" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26120
19:36 weinig joined #perl6
19:37 fglock joined #perl6
19:37 svnbot6 r16921 | masak++ | trunk:
19:37 svnbot6 : [t/data_types/array.t]
19:37 svnbot6 : * changes based on input from TimToady++
19:37 pugs_svnbot r16921 | masak++ | trunk:
19:37 pugs_svnbot : [t/data_types/array.t]
19:37 pugs_svnbot : * changes based on input from TimToady++
19:37 pugs_svnbot diff: http://dev.pugscode.org/changeset/16921
19:37 lambdabot Title: Changeset 16921 - Pugs - Trac
19:37 unobe changed the nick to unobe_away
19:37 masak that was just a tiny commit :)
19:37 agentzh committed
19:38 pugs_svnbot r16922 | agentz++ | trunk:
19:38 pugs_svnbot : [PCR]
19:38 pugs_svnbot : - added examples/Grammar.grammar which is the grammar spec
19:38 pugs_svnbot :   for Perl 6 regexes themselves.
19:38 pugs_svnbot :   this grammar file can be compiled by util/compile_p6grammar.pl
19:38 pugs_svnbot :   down to a .pm file and replaces lib/Pugs/Grammar/Rule.pmc.
19:38 svnbot6 r16922 | agentz++ | trunk:
19:38 svnbot6 : [PCR]
19:38 svnbot6 : - added examples/Grammar.grammar which is the grammar spec
19:38 svnbot6 :   for Perl 6 regexes themselves.
19:38 svnbot6 :   this grammar file can be compiled by util/compile_p6grammar.pl
19:38 pugs_svnbot : - added util/update-rule-pmc-new to automate this process.
19:38 svnbot6 :   down to a .pm file and replaces lib/Pugs/Grammar/Rule.pmc.
19:38 svnbot6 : - added util/update-rule-pmc-new to automate this process.
19:38 pugs_svnbot : - updated Rule.pmc using this new approach. It's about 4 times
19:38 svnbot6 : - updated Rule.pmc using this new approach. It's about 4 times
19:38 pugs_svnbot :   faster than the traditional v6.pm approach!
19:38 pugs_svnbot : - set $NoCache to 1 by default in P::C::Regex to help testing
19:38 svnbot6 :   faster than the traditional v6.pm approach!
19:38 pugs_svnbot :   (will reset to 0 before release though).
19:38 svnbot6 : - set $NoCache to 1 by default in P::C::Regex to help testing
19:38 pugs_svnbot : (no regressions)
19:38 svnbot6 :   (will reset to 0 before release though).
19:38 pugs_svnbot diff: http://dev.pugscode.org/changeset/16922
19:38 svnbot6 : (no regressions)
19:38 masak whoa
19:38 masak competing bots
19:38 agentzh masak, fglock, avar: Please see http://svn.pugscode.org/pugs/perl5/Pugs​-Compiler-Rule/examples/Grammar.grammar
19:38 lambdabot http://tinyurl.com/2ez2j3
19:39 masak agentzh: oki
19:39 agentzh PCR now uses compile_p6grammar.pl to compile this grammar file down to its own regex parser (i.e. Pugs::Grammar::Rule.pmc)
19:39 CardinalNumber joined #perl6
19:39 masak agentzh: nice
19:39 agentzh in other words, this part is self-hosting.
19:40 masak agentzh: not bad
19:40 agentzh (in fact, it's a grammar parser)
19:40 agentzh a superset of Perl 6 regexes.
19:41 SamB joined #perl6
19:41 masak I still have the feeling I'd be even more impressed if I had a better overview of Perl 6 work
19:41 masak but a self-hosting grammar does sound important
19:41 avar agentzh: ah, %{}% is perl5 code?
19:41 agentzh avar: yup
19:41 agentzh avar: so is the actions
19:42 agentzh (to be more accurate, the action code goes through a source-filter which de-sugars things like $/ and $<foo>)
19:42 avar well, they look sort-of-p6
19:42 avar ah, yeah
19:42 masak agentzh: why p5?
19:43 agentzh source-filters may give troubles but they're pretty handy.
19:43 agentzh masak: because p5 is fast!
19:43 masak agentzh: oh yeah, that's why
19:43 agentzh and more complete, more flexibale and more stable
19:43 fglock agentzh: looking
19:43 agentzh *flexible
19:43 avar ..and p6 isn't going to be here until next**{inf} christmas:)
19:44 agentzh right
19:44 Tene So, can we do source filters in p6 still if we want to?
19:44 masak avar: don't say that :( you're making me cry
19:45 rindolf Hi all! Can anyone help with this Prolog problem http://sial.org/pbot/26119 ?
19:45 lambdabot Title: Paste #26119 from "rindolf" at 82.81.175.57
19:45 masak in a way, p6 is here already
19:45 agentzh fglock: some parts of Grammar.grammar look ugly because PCR has inconsistent calling protocol between named subrules and anonymous subrules :(
19:46 agentzh Tene: there's macros in Perl 6.
19:47 agentzh Tene: they're more stable source-filters :)
19:47 avar and you can override the grammar lexically, presumably
19:47 Tene agentzh: yes, but how difficult would it be to do a traditional source filter? ;)
19:47 agentzh fglock: it has taken me many hours to get it right ;)
19:48 agentzh Tene: not sure...
19:48 jjore rindolf: solving for any one word is sufficient. You asked for word(...) OR word(...) OR ...
19:48 avar Tene: as in override the IO layer when it's reading in source files?
19:49 rindolf jjore: ah, I see.
19:49 rho joined #perl6
19:49 fglock agentzh: you should take a look at token-test.pl - p6rx-on-p5rx is close to usable :)
19:50 agentzh fglock: yes, i will!
19:50 rindolf jjore: thanks.
19:50 jjore np.
19:50 avar fglock: as in via kp6?
19:50 rindolf jjore: I indeed mixed between ";" (OR) and "," (AND).
19:50 fglock avar: yes; but it is portable to PCR
19:51 rindolf jjore: also had a few variable problems.
19:51 jjore sure.
19:52 avar fglock: If you get it working well enough kp6 might replace pcr:)
19:53 agentzh avar: the parser can be written in kp6 at least
19:53 fglock that's the idea
19:55 agentzh fglock: i'll try to find some time to work on a p6 regex tracer with a web UI tomorrow...
19:55 agentzh fglock: we need a good tool to debug grammars, it seems
19:56 agentzh fglock: not sure how good it will be, but i have some interesting ideas in my head aleady
19:56 avar something like -Mre=debug ?
19:56 agentzh avar: more descriptive
19:56 avar how so?
19:57 agentzh avar: it has a GUI.
19:57 agentzh avar: it's easier to explain once i can give you a live demo :)
19:58 agentzh avar: but i can't promise anything but i'd like to try :)
19:58 agentzh too late to make a PCR release tonight...4am already...
19:58 agentzh will crash soon...
19:59 agentzh see you guys "tomorrow"... &
19:59 agentzh left #perl6
19:59 masak agentzh: 'night
20:00 fglock joined #perl6
20:03 fglock does / (.) $0 := (.) / overwrite $0, or does it capture to an array?
20:06 masak fglock: good question. is that covered in the synopses?
20:07 fglock kind of - it's a corner case
20:07 masak definitely
20:08 masak I don't understand your second alternative fully, it seems magical to me
20:08 masak so I would go for 'overwrite'
20:09 fglock masak: / $<x> := (.) $<x> := (.) / captures to @( $<x> )
20:10 masak oh -- didn't know that
20:10 fglock I wonder if positional captures do the same, and to what extent I should care in compile-time
20:11 masak it's not compatible with my current understanding of binding :=
20:11 masak seemingly they work differently in regexes
20:11 fglock yes, it's regex named-capture binding
20:11 masak :)
20:12 masak and it still seems magical, but now intentionally so
20:12 masak now the latter alternative seems more plausible to me
20:16 masak I seem to remember there was something wonky with regex array captures
20:16 masak wonder if it ever got fixed...
20:18 masak audreyt wrote an email about it months ago, with an opinion that sounded like sanity compared to the (then) semantics of array capture
20:18 masak I don't know if it has been addressed
20:20 masak here's the thread in question: http://groups.google.com/group/perl.perl6.la​nguage/browse_thread/thread/c2b106d848ac8019
20:20 lambdabot Title: Motivation for /&lt;alpha&gt;+/ set Array not Match? - perl.perl6.language | Goo ..., http://tinyurl.com/yub8y7
20:21 masak anyone know what happened to that issue? last I heard from TimToady was that he needed to discuss it with TheDamian
20:23 fglock masak: yes, that was my idea, but it was not approved
20:23 masak fglock: oh, so things are still the way they were before the proposal?
20:24 masak is there a discussion recorded somewhere explaining why it wasn't approved?
20:24 fglock I'll take a look
20:24 masak upon rereading the thread, it still seems clear to me that the then semantics are strange and surprising
20:29 vel_ joined #perl6
20:30 masak it's stuff like this that makes me glad p6 is still quite fluid
20:30 * rho is back
20:30 masak there's still room for opinions and improvement
20:43 weinig joined #perl6
20:55 lixinso left #perl6
21:07 dduncan joined #perl6
21:08 weinig joined #perl6
21:13 unobe_away changed the nick to unobe
21:14 justatheory joined #perl6
21:17 jisom joined #perl6
21:18 vel joined #perl6
21:19 iblechbot joined #perl6
21:39 vel joined #perl6
21:42 svnbot6 r16923 | fglock++ | trunk:
21:42 svnbot6 : [kp6] p5regex - implemented / $<named> := (...) /
21:42 pugs_svnbot r16923 | fglock++ | trunk:
21:42 pugs_svnbot : [kp6] p5regex - implemented / $<named> := (...) /
21:42 pugs_svnbot diff: http://dev.pugscode.org/changeset/16923
21:42 lambdabot Title: Changeset 16923 - Pugs - Trac
21:45 Alias__ joined #perl6
21:46 pasteling "evalbot_r16922" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26122
21:57 avar joined #perl6
22:00 svnbot6 r16924 | fglock++ | trunk:
22:00 svnbot6 : [kp6] p5regex - the match-result-interpreter now uses a dispatch table
22:00 pugs_svnbot r16924 | fglock++ | trunk:
22:00 pugs_svnbot : [kp6] p5regex - the match-result-interpreter now uses a dispatch table
22:00 pugs_svnbot diff: http://dev.pugscode.org/changeset/16924
22:00 lambdabot Title: Changeset 16924 - Pugs - Trac
22:14 rlb3 joined #perl6
22:18 vel joined #perl6
22:20 hakobe changed the nick to hkb_zz
22:20 justatheory joined #perl6
22:20 fglock joined #perl6
22:39 vel joined #perl6
22:42 svnbot6 r16925 | fglock++ | trunk:
22:42 svnbot6 : [kp6] p5regex - named captures can capture to array
22:42 pugs_svnbot r16925 | fglock++ | trunk:
22:42 pugs_svnbot : [kp6] p5regex - named captures can capture to array
22:42 pugs_svnbot diff: http://dev.pugscode.org/changeset/16925
22:42 lambdabot Title: Changeset 16925 - Pugs - Trac
22:43 pasteling "evalbot_r16924" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26124
22:50 vel_ joined #perl6
22:55 clkao seen xdg
22:58 fglock clkao: hi
22:59 clkao hola
23:00 fglock clkao: would you update Sub-Multi in cpan? I've committed a small fix
23:00 clkao oh sure
23:00 clkao do you want to comaint it?
23:01 fglock sure, but please make the upload this time, if possible
23:02 clkao sure
23:02 fglock thanks :)
23:02 clkao you've got CHANGES /version etc all set?
23:03 clkao you didn't add a failing test for the thing you fixed ;)
23:03 jrockway hmm, is moose using Sub::Multi yet?
23:03 jrockway that would be pretty cool
23:03 fglock looking
23:03 clkao now again
23:04 clkao how the fuck do i generate meta.yml
23:04 clkao why does it all stop working
23:04 diakopter_afk changed the nick to diakopter
23:05 fglock clkao: sorry, no tests in Sub::Multi; the failing test belongs to v6.pm
23:05 jrockway WriteAll(meta => 1) ?
23:05 fglock http://dev.pugscode.org/changeset/15766
23:05 jrockway then rerun Makefile.PL
23:05 clkao right, i mean there should be unit test in s::m :)
23:05 lambdabot Title: Changeset 15766 - Pugs - Trac
23:05 clkao no worky
23:12 clkao fglock: there you go
23:12 svnbot6 r16926 | clkao++ | trunk:
23:12 svnbot6 : Sub-Multi 0.003
23:12 pugs_svnbot r16926 | clkao++ | trunk:
23:12 pugs_svnbot : Sub-Multi 0.003
23:12 pugs_svnbot diff: http://dev.pugscode.org/changeset/16926
23:12 lambdabot Title: Changeset 16926 - Pugs - Trac
23:13 bonesss joined #perl6
23:14 fglock clkao: thank you :)
23:14 clkao fglock: you are comaint now as well
23:20 vel joined #perl6
23:41 svnbot6 r16927 | fglock++ | trunk:
23:41 svnbot6 : [kp6]
23:41 svnbot6 : - added 'perl5rx' target:
23:41 svnbot6 :   $ perl kp6-perl5.pl --perl5rx < examples/token-test.pl | perl -I lib5regex -I lib
23:41 svnbot6 : - same-name captures now correctly capture to array
23:42 pugs_svnbot r16927 | fglock++ | trunk:
23:42 pugs_svnbot : [kp6]
23:42 pugs_svnbot : - added 'perl5rx' target:
23:42 pugs_svnbot :   $ perl kp6-perl5.pl --perl5rx < examples/token-test.pl | perl -I lib5regex -I lib
23:42 pugs_svnbot : - same-name captures now correctly capture to array
23:42 pugs_svnbot diff: http://dev.pugscode.org/changeset/16927
23:42 lambdabot Title: Changeset 16927 - Pugs - Trac
23:54 jameswf joined #perl6
23:56 bsb joined #perl6

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

Perl 6 | Reference Documentation | Rakudo