Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2007-06-20

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:00 japhb I guess the x'es qualify as NIY marks ... it would be nice to have two columns, one for pugs smoke and one for parrot/perl6 smoke.  Though I don't know if the parrot/perl6 people are quite ready for that yet.
00:00 japhb Or three, to include kp6 ...
00:02 ibrown looking at example.
00:03 ibrown so you need folks to run the tests, to populate these docs?
00:03 japhb info pulled from smoke runs.
00:03 ibrown or is the process just confusing me?
00:04 japhb ibrown: many cronjobs that make magic.  :-)
00:04 japhb see http://m19s28.vlinux.de/cgi-bin/pugs-smokeserv.pl
00:04 lambdabot Title: Pugs Smoke Reports
00:05 japhb Ignore the release lines, it's been a while since a release.  Go for repo snapshot lines.
00:07 ibrown ubuntu eq linux?
00:07 japhb ubuntu isa debian isa linux, more or less
00:07 ibrown yeah. figured as much.
00:07 ibrown just wanted to make sure.  i'm new to deb, migrated from slack
00:08 japhb welcome to the light side of the force
00:09 ibrown lol
00:09 ibrown yeah we all grow up.
00:09 ibrown its been slack since the early 90's tho.  so its like hard to part.
00:09 ibrown all these damn pretty pictures.
00:11 thoughtpolice i'm an archlinux guy myself. although i also like gentoo and bsd. never tried just vanilla slack.
00:12 ibrown start with disk,video,keyboard.  
00:13 ibrown compile what you need
00:13 ibrown let it be a rock
00:13 ibrown albeit it does suck as a desktop
00:13 ibrown but my bnc wireless card is supported by ubuntu pretty easily, so i said wth
00:14 kunwon1 joined #perl6
00:16 ibrown so they release a svn #.  you check it out, do the tests, and commit the results which are stored in .pod files?
00:16 ibrown then that information is given back to the assembler interpreter guys? who take it and elaborate their project?
00:20 japhb I believe that when you 'make smoke' it uploads your test results to the smoke server, and when the pods are converted to HTML, the test sources and smoke results are incorporated
00:20 ibrown ah
00:22 ibrown so how do i get commit access to help out?
00:22 japhb ask anyone for a commitbit.  You need to give us your preferred user name, and an email address
00:23 japhb private message works best.
00:23 japhb Hold on, lets see if I can get into the invite manager ....
00:24 japhb (the invite manager and the server its hosted on seem to be at odds, resulting in VERY slow response times, so hold on)
00:24 japhb s/its/it's/
00:25 ibrown i still don't exactly follow i guess what the tests are doing per say.  and the like but i'm sure i'll get it after watching a min.
00:25 [particle] what the tests are doing?
00:26 [particle] there's a utility called 'smartlinks.pl' that takes as input the test files and spec files and matches smartlinks in the test files to locations in the spec files.
00:27 ibrown well.
00:27 ibrown here's a good example:
00:27 [particle] html is generated which includes the test file source inside the htmlified pod
00:27 ibrown i run grep -RL 'L<' t/* | $PAGER
00:27 ibrown which i understand, kind of,
00:27 ibrown what do i do with $PAGER
00:27 ibrown to dtermine the results.
00:28 ibrown unelss that just meant | less lol
00:28 [particle] is $PAGER set to 'less'?
00:28 ibrown or more or whatever
00:28 japhb er, I think $PAGER is supposed to mean 'less'
00:28 ibrown that's what i figured.  ok that makes sense.
00:28 [particle] yes, it's an env var, like $EDITOR
00:29 ibrown tru i don't normally call it by its var my fault on that one
00:33 japhb OK, ibrown, I'm ready to invite. /msg me your username and email?
00:38 japhb ibrown: having trouble /msg'ing?
00:39 ibrown i msg'd you
00:39 ibrown did it not go through
00:39 japhb nope, didn't arrive.
00:40 japhb Are you signed in with nickserv?
00:40 ibrown neg
00:40 japhb required on freenode to msg ...
00:40 ibrown oh
00:40 ibrown that explains.
00:40 japhb You can also send in-channel ... our logbot hides emails.
00:40 japhb most people just prefer not to send in-channel, of course ....
00:40 ibrown ibrown@dontpatronize.us ibrown: is fine
00:40 japhb k, hold on.
00:40 ibrown i don't care.
00:41 ibrown whatever, spam me some more!
00:41 japhb heh
00:42 * offby1 launches a can of spam ibrown-ward
00:42 japhb The server is chewing on your invite
00:44 japhb So here's the deal.
00:44 japhb (brought from priv msg land)
00:44 f0rth joined #perl6
00:44 japhb The tests are just Perl 6 scripts.
00:45 japhb They have various bits of magic in them, including these 'smart links'.
00:45 ibrown yeah little routines testing functions it looks like.
00:45 japhb Once the smart links are added once, they don't need to be updated again, unless the spec changes overmuch.
00:45 TimToady heh, heh
00:45 japhb Once added, they are automatically managed by the cronjobs.
00:45 japhb so the spec annotation that you see is all automated.
00:46 japhb We just need more granularity and completeness to the 'smartlinks'.
00:47 japhb Perhaps it will help to consider them HTML <a> tags.  Someone needs to put a lot of anchors in the tests.  After that, the hyperlinking Just Works.
00:47 japhb Oh, and don't listen to TimToady, he's just here to pull your leg.  ;-)
00:48 TimToady I never do anything useful around here except cause more trouble.
00:48 japhb 'xactly
00:48 ibrown i feel yeah
00:49 ibrown i'm doing a lot of analysis lately and swear by perl
00:49 ibrown but this cat i'm workign with is pressuring me hardcore to give it up
00:49 ibrown and swtich to something else, i'm like ugh. no.  but the perl6 language i'm liking.
00:49 offby1 I swear at perl
00:49 ibrown aye me too
00:49 offby1 my cat doesn't care which language I use
00:49 ibrown collegue?
00:50 japhb My cat needs her nails clipped.  I think she just put a hole in my pants leg
00:51 ibrown i can def do the smartlinks
00:51 japhb <burns>Excellent...</burns>
00:52 ibrown i just hope there's a bit more daunting tasks up the road. =P
00:52 TimToady you could write me a regex engine in Haskell tonight if you like. :)
00:52 ibrown shiit. why didn't you just ask.
00:53 ibrown i believe there's one under the sofa. brb
00:54 offby1 mine's outside calling for his buddies
00:58 japhb dinner &
00:58 offby1 renice -20 %1
00:58 rintaro joined #perl6
00:58 offby1 ha!  now let's watch him eat really fast
01:06 nipotaway changed the nick to nipotan
01:11 rintaro_ joined #perl6
01:20 cognominal_ joined #perl6
01:25 offby1 gaah.  I wish #parrot were on this server
01:33 dolmans joined #perl6
01:34 uno_be joined #perl6
01:37 japhb Mmmm, good dinner.  And I did, in fact, eat really fast, so it must have worked.  I was done well before everyone else was half done.
01:37 offby1 well, you burned up all my CPU.  I tried to SIGHUP you but you blocked it
01:38 japhb heh
01:40 rintaro joined #perl6
01:44 * japhb yells at perl5: "What do you mean, 'Flash Gordon approaching'?!?"
01:49 * japhb fervently hopes that Perl 6 will make it stunningly easy to redirect open filehandles for child processes *without* involving the shell or requiring use of some module to do what should be easy.
01:55 rintaro__ joined #perl6
02:00 offby1 scsh is great at that sorta thing
02:06 rintaro_ joined #perl6
02:07 nipotan changed the nick to nipotaway
02:16 ibrown ok so not to beat a dead horse, but checkout source, configure,make,make test, make smoke, add smartlinks, checkout
02:25 REPLeffect joined #perl6
02:26 TimToady joined #perl6
02:28 ibrown timtoady: so got a quick question if you got a spare second.
02:29 TimToady I do
02:30 ibrown so i run grep in 01-sanity and it opens up: 02-counter.t which to me appears to be an increment test. : then i open up the document for 02 on that URL you submitted, and find in the document where it references increment.  then I create a link to the test result and that's essentially what's being committed
02:31 ibrown i guess i'm just confused at what document i reference for the link. i understand the link syntax now etc.
02:32 TimToady well, the link is from the .t back into the various synopses
02:33 ibrown so like http://feather.perl6.nl/syn/S02.html that page
02:33 lambdabot Title: S02
02:34 ibrown true but where are they at in the src tree? docs?
02:35 TimToady it's probably working with the pod versions.  they're in the source tree but have different names there
02:35 japhb ibrown: see how its done in e.g. pugs/t/operators/eq.t
02:36 TimToady docs/Perl6/Spec has the pod files, but like Operators is really S03
02:36 japhb At the top of each spec doc is the correct number, they are numbered S<nn>
02:36 ibrown ok so docs/Perl6/Spec
02:36 TimToady you can also check out the most official ones under the S names from svn.perl.org
02:36 ibrown gotcha
02:36 ibrown been searching for that location
02:37 TimToady the official synopses live at https://svn.perl.org/perl6/doc/trunk/design
02:37 japhb As with that e.g. test file above, make sure you use S#/Section/string_to_match.
02:39 ibrown but there's not pods for everything yet?
02:39 TimToady semi-official and unofficial still live in docs/Perl6/Spec
02:39 sunnavy joined #perl6
02:40 ibrown can i check those docs out at that svn url?
02:40 TimToady yes
02:40 ibrown awesome.
02:41 TimToady and then grep is your friend
02:41 ibrown sorry for the annoying questions, i'm just starting to branch out into the community from my cave.
02:41 ibrown rgr
02:42 ibrown i'm checking out the docs. shouldn't be that big of a deal.
02:42 TimToady no problem; we're down with branching out
02:42 ibrown i just don't wanna switch to ruby.
02:43 TimToady boku mou...
02:43 mako132_ joined #perl6
02:43 ibrown ?
02:43 TimToady "me too" in Japanese...
02:43 ibrown rgr
02:44 ibrown if you want help with things on the regex engine let me know and i'll find some docs on the language you were speaking of earlier.
02:44 ibrown i'm not terrible with the regexes.
02:44 TimToady it was sort of a joke
02:45 TimToady if you haven't programmed in Haskell yet, it would be a long learning curve...
02:45 TimToady Haskell is even differenter than Ruby
02:46 ibrown that's pseudo amusing.
02:50 ibrown uhm.
02:50 ibrown yeah that's a curve buddy
02:55 rintaro joined #perl6
03:00 rintaro__ joined #perl6
03:02 japhb Hmmm, am I correct in reading the Parrot release announcement that Parrot is released *only* with Artistic License 2.0, and *not* a disjunction of AL 2 | GPL n?
03:09 Eidolos in my (admittedly crusty) parrot checkout it has both AL and GPL
03:10 PerlJam japhb: hmm. that's a good question.
03:11 PerlJam Eidolos: there was a release today.
03:11 Eidolos yeah
03:12 Eidolos it's a lot of work to change the license on an open source multimultideveloper project though
03:12 Eidolos see angband :)
03:14 Eidolos "Updated Parrot distribution to Artistic License 2.0, from dual Artistic 1/GPL license."
03:14 Eidolos hmmm
03:19 * PerlJam asks #parrot and the mailing lists
03:19 Eidolos wow
03:19 Eidolos it looks like they did drop the GPL
03:20 justatheory joined #perl6
03:20 Eidolos LICENSES/ became LICENSE which now contains just the AL2
03:22 PerlJam Ah ... http://groups.google.com/group/perl.perl6.internals/browse_thread/thread/655185fb525fb994/41ca94fc8ba6d29b?lnk=gst&amp;q=Parrot+license&amp;rnum=3#41ca94fc8ba6d29b
03:22 lambdabot Title: [TODO] replace dual Artistic/GPL license with Artistic 2.0 only - perl.perl6.int ..., http://tinyurl.com/2go2ge
03:22 PerlJam thank you lambdabot!
03:22 * offby1 <3 automatic tiny-url-ification
03:23 the_dormant joined #perl6
03:24 japhb Ah, that's cool
03:24 japhb Someone have a link to LICENSE?  My parrot checkout is WAY out of date, and I don't feel like waiting for pull ...
03:25 PerlJam http://foundation.perl.org/legal/licenses/artistic-2_0.txt
03:25 japhb Thanks, PerlJam
03:25 PerlJam See section 4 is the part that's relevant to why they aren't dual licensing anymore I believe.
03:25 PerlJam 4.c.ii specifically
03:26 * japhb reading it now
03:27 Eidolos perl6 is dual-licensed under AL2 and GPL1+
03:27 Eidolos or was at rfc-response time
03:28 Eidolos only 7 years out of date, not such a good source, I think I'll just go to bed then :)
03:28 TimToady AL2 is officially "GPL compatible" so doesn't need to be dual licensed
03:29 Eidolos ahh
03:34 japhb TimToady: Is there a list of what licenses match 4.c.ii?
03:37 TimToady dunno, you'd have to ask Allison probably
03:37 TimToady I try to stay away from licenses these days.  :)
03:39 beppu i can has public domain?
03:40 TimToady I don't stay that far away from licenses...
03:40 PerlJam TimToady: heh
03:47 dduncan indeed
03:47 dduncan AL2 is considered a free non-copyleft license, and so is GPL-any compatible
03:47 dduncan so I agree this change makes sense
03:48 dduncan compatible without special cases too, so a AL2 work can on its own be relicensed GPL
03:49 dduncan on that note, the general Pugs license could probably stand to see a similar change
03:50 dduncan the Pugs readme currently says dual AL2/GPL, could just be AL2
03:50 dduncan though on that note it should be AL2, not some pre-release AL
03:50 dduncan assuming that AL2 is finalized already and legal to use
03:50 dwave joined #perl6
03:51 dduncan that is, the Pugs README still says "2.0beta5" and that seems out of date, should just say 2.0 if that is final
03:52 PerlJam "legal to use"?
03:52 beppu maybe "production ready" would make more sense to our programming-warped minds.
03:53 beppu has AL2 gone through QA and been thoroughly debugged?  ;-)
03:53 PerlJam beppu: indeed
03:57 dduncan I would think so, if Parrot and that Oreilly book are now officially released with it
03:58 dduncan so is there any reason for the Pugs repo and docs to now not just reference the final AL2?
03:58 dduncan for that matter, would anyone complain if I made that change, or should we wait for audreyt to do it?
04:00 PerlJam dduncan: I wouldn't complain.  In fact, I was going to suggest someone just do it earlier (being to lazy myself :)
04:00 PerlJam s:2nd/to/too/
04:00 dduncan right then ... so I'll get started on that now ...
04:02 perlbot joined #perl6
04:09 mako132_ joined #perl6
04:15 devogon joined #perl6
04:16 amnesiac joined #perl6
04:28 justatheory joined #perl6
04:28 dduncan so I've replaced the license file itself, and updated references in README and ChangeLog
04:30 dduncan I'm also assuming that docs/01Overview.html isn't a frozen point-in-time doc, so that will be updated too
04:34 pmichaud joined #perl6
04:38 mako132_ joined #perl6
04:39 monomorph joined #perl6
04:42 dduncan committing ...
04:43 dduncan done
04:43 svnbot6 r16694 | Darren_Duncan++ | the Pugs main license is now simply Artistic 2.0 (final) rather than Artistic 2.0b5 disjoint GPL
04:43 dduncan in case there are any concerns about having the right version, the new license text is from http://www.perlfoundation.org/page/perlfoundation/legal/attachments/20061112063921-3-29352/artistic-2_0.txt
04:43 lambdabot http://tinyurl.com/28rwar
04:43 dduncan which afaik is the official source
04:45 mako132_ joined #perl6
05:31 ibrown so after you add a smartlink, and run the smartlink --check *.t it spits out a html file correct?
05:39 |Azure| joined #perl6
05:39 |Azure| left #perl6
05:40 |Azure| joined #perl6
05:41 _Azure_ joined #perl6
05:42 _Azure_ hello
05:42 _Azure_ changed the nick to Azure
05:45 Azure left #perl6
05:49 rintaro__ changed the nick to rintaro
05:49 rintaro joined #perl6
05:53 ibrown i don't get why i created a link in S02, but when smartlinks ran the output was named S26.html
05:53 ibrown diff page.
05:54 hirschnase joined #perl6
05:58 _Azure_ joined #perl6
06:02 _Azure_ How to contribute
06:02 _Azure_ I'm the first time use irc
06:03 diakopter hi _Azure_
06:03 diakopter how would you like to contribute
06:03 _Azure_ I'm come from perlchina
06:04 diakopter do you have a commit bit for pugs?
06:04 _Azure_ no
06:04 diakopter have you registered your freenode nickname with NickServ?
06:05 _Azure_ how to do that
06:05 diakopter (so you can send a /msg to myself with your email address and preferred username)
06:05 hoowa joined #perl6
06:05 diakopter _Azure_: http://freenode.net/faq.shtml#nicksetup
06:05 lambdabot Title: freenode: frequently-asked questions
06:07 diakopter _Azure_: /msg nickserv register <your-password>
06:08 _Azure_ I see
06:08 _Azure_ have a try
06:08 diakopter or, you can just say your email address and username here..
06:13 kanru joined #perl6
06:14 |Azure| joined #perl6
06:14 |Azure| changed the nick to Azure
06:15 Azure left #perl6
06:15 Azure joined #perl6
06:16 diakopter Azure: did you find a nickname that hadn't previously been registered?
06:16 Azure no
06:17 diakopter well then, good luck.  try _|Azure|_ ?
06:17 Tene Heh.
06:18 diakopter changed the nick to _|Azure|_
06:18 _|Azure|_ changed the nick to diakopter
06:19 Azure when I input /msg nickserv register  second times it show  azure already registered
06:20 Azure success?
06:20 diakopter try sending a msg to me
06:20 diakopter /msg diakopter test
06:21 perlbot joined #perl6
06:21 diakopter probably it is registered by someone else.  you'll need to change your nickname to something else and then try registering again
06:21 diakopter I just tried _|Azure|_ - it's not taken. ;)
06:22 diakopter /nick _|Azure|_
06:22 Azure joined #perl6
06:23 Azure changed the nick to azure
06:23 azure changed the nick to Azure
06:24 Azure why I login with no password
06:25 diakopter you're not yet logged in. those nicknames are not registered to you, it appears.
06:25 Azure how to login
06:27 |Azure| joined #perl6
06:30 |Azure| ping
06:31 |Azure| changed the nick to azure1st
06:31 azure1st changed the nick to aaaasfsfsdf
06:31 aaaasfsfsdf changed the nick to azure1st
06:32 azure1st changed the nick to acprp
06:38 uno_be changed the nick to unobe
06:49 BooK joined #perl6
06:49 acprp changed the nick to azure
06:49 azure changed the nick to acprp
06:49 acprp changed the nick to azure
06:50 azure changed the nick to Azure
06:50 buetow joined #perl6
06:51 Azure changed the nick to acprp
06:51 acprp changed the nick to Azure
06:52 Azure changed the nick to azure1st
06:53 nipotaway changed the nick to nipotan
06:54 |Azure| joined #perl6
07:01 unobe changed the nick to unobe_away
07:03 franck__ joined #perl6
07:04 acprp joined #perl6
07:06 acprp changed the nick to Azure
07:07 iblechbot joined #perl6
07:07 franck__ joined #perl6
07:07 Azure changed the nick to wwwwwwwwwwwwwww
07:11 |Azure| joined #perl6
07:11 |Azure| azure1st is me how to kill it
07:12 franck___ joined #perl6
07:12 |Azure| I want register it
07:13 |Azure| changed the nick to acprp
07:13 acprp changed the nick to azure1st
07:18 azure1st "you have already identified" I've register success?
07:18 azure1st ping
07:18 Tene Hi!
07:20 Tene By the way, everyone, if you have trouble with unregistered people not being able to message you, /msg nickserv set unfiltered on
07:20 |Azure| joined #perl6
07:21 |Azure| ?
07:22 Tene Hi!
07:22 diakopter Tene: thanks
07:22 * Tene is experiencing failure loading commitbit
07:22 * diakopter has commitbit loaded
07:23 |Azure| why azure1st not me
07:23 Tene I've never had any problems with freenode spammers, so I've had my nicks et unfiltered almost since the filtering was first put in place.
07:29 jisom joined #perl6
07:30 bsb left #perl6
07:30 rho joined #perl6
07:37 |Azure| changed the nick to acprp
07:37 acprp hi
07:38 Tene Hi!
07:46 acprp changed the nick to auzre1st
07:47 auzre1st changed the nick to azure1st
07:50 azure1st Hi  I already registered. How to contribute
07:52 |Azure| joined #perl6
07:54 tlb joined #perl6
07:54 Tene diakopter: still around?
07:58 mico joined #perl6
07:59 diakopter joined #perl6
08:03 daxim joined #perl6
08:20 svnbot6 r16695 | Darren_Duncan++ | ext/Muldis-DB/ : updated Operators.pm, fixed bugs in AST.pm and PhysType.pm
08:27 dduncan left #perl6
08:45 svnbot6 r16696 | tene++ | Minor change in tutorial.
08:46 diakopter changed the nick to diakopter_afk
08:48 Tene yay, git-svn works!
08:48 Aankhen`` Wassat?
08:53 Tene use svn repositories through git.
08:56 Aankhen`` Ah.
08:56 Aankhen`` I wonder how many different SCM systems are actually used to interface with the Pugs repo.
09:06 BenjaminZou joined #perl6
09:33 buetow joined #perl6
09:36 Southen_ joined #perl6
09:43 isaacd joined #perl6
09:44 chris2 joined #perl6
10:18 rfordinal_ joined #perl6
10:33 gaal joined #perl6
10:48 lidden joined #perl6
10:59 ruz joined #perl6
11:17 lumi joined #perl6
11:45 nipotan changed the nick to nipotaway
12:00 araujo joined #perl6
12:06 Lunchy joined #perl6
12:12 diakopter_afk changed the nick to diakopter
12:35 qmole joined #perl6
12:42 BooK_ joined #perl6
12:53 isaacd_ joined #perl6
13:06 ofer joined #perl6
13:34 unobe_away joined #perl6
13:51 diakopter changed the nick to diakopter_afk
14:04 diakopter_afk changed the nick to diakopter
14:11 explorer joined #perl6
14:18 vel joined #perl6
14:26 ruoso joined #perl6
14:34 BooK joined #perl6
14:57 fglock joined #perl6
14:59 fglock TimToady: re "write me a regex engine in Haskell tonight" - http://svn.pugscode.org/pugs/perl5/Pugs-Compiler-Rule/lib/Pugs/Emitter/Rule/Parsec.pm might be a good starting point for a quick implementation
14:59 lambdabot http://tinyurl.com/24mutj
15:10 deq` joined #perl6
15:14 REPLeffect joined #perl6
15:18 perljunkie joined #perl6
15:20 kanru joined #perl6
15:30 buetow joined #perl6
15:31 iblechbot joined #perl6
15:49 mj41 joined #perl6
15:56 fglock joined #perl6
15:58 diakopter joined #perl6
15:59 marmic joined #perl6
16:03 franck__ joined #perl6
16:04 agentzh joined #perl6
16:04 charsb___ joined #perl6
16:05 agentzh Aankhen``: no, i'm no longer seeing the strange onload behaviour in the smartlinked pages. you've fixed it? hehe
16:05 lambdabot agentzh: You have 1 new message. '/msg lambdabot @messages' to read it.
16:05 agentzh @messages
16:05 lambdabot moritz said 2d 3h 1m 31s ago: I converted your chinese translation of S01 to HTML using Pod::Tree::HTML, and the output is horribly garbeled <http://moritz.faui2k3.org/pugs/docs/zh-cn/syn/S01.pod.
16:05 lambdabot html> - do you know any pod2html converter that can handle your pod?
16:06 agentzh moritz: i've been using Pod::Simple::HTML here: http://perlcabal.org/syn/zh-cn/S01
16:06 lambdabot Title: S01
16:06 agentzh moritz: it looks fine to me :)
16:06 agentzh the Chinese version of S01 is in UTF-8.
16:07 sunnavy agentzh: will you translate all of syns?
16:08 agentzh sunnavy: well, not really
16:08 agentzh but i do love to help ;)
16:08 agentzh .oO( translating is a daunting task )
16:09 agentzh afaik, fayland has translated most of the synopses to Chinese in his personal repos: http://www.fayland.org/Perl6/Synopsis/zh_cn/
16:09 lambdabot Title: Index of /Perl6/Synopsis/zh_cn
16:09 agentzh it's a bit old though
16:09 agentzh he did the translation in early days of pugs.
16:11 sunnavy I think it's a bit hard to translate
16:11 barney joined #perl6
16:11 agentzh pretty hard
16:11 agentzh i've finished a Chinese transcript for S26
16:11 agentzh in my notebook
16:11 agentzh haven't entered them into my machine...
16:12 agentzh but...translating is a good way to study the synopses...
16:12 moritz agentzh: thanks, I'll have a look at it
16:12 sunnavy especially some technical terms, it's hard to find corresponding chinese words.
16:12 agentzh audreyt used to learn by translation...
16:12 agentzh moritz: okay :)
16:12 agentzh sunnavy: indeed!
16:13 agentzh i've been creating terms along the way
16:14 agentzh The slides for my talk "Contribute to Pugs" at YAPC::Beijing 2007 is now on feather:
16:14 sunnavy I read syns last year but not finished, and I am going read through them soon.
16:14 agentzh http://perlcabal.org/agent/slides/contr_pugs/contr_pugs.xul
16:14 agentzh sunnavy: great
16:15 agentzh sunnavy: maybe you can submit a lot of patches to p6l while reading
16:15 agentzh that's what i did.
16:15 sunnavy what's p6l?
16:15 moritz perl6-language
16:15 moritz the mailing list
16:15 agentzh mailing list
16:16 sunnavy well, wish I will :-)
16:16 moritz perl6-language@perl.org
16:16 agentzh see http://dev.perl.org/perl6/lists/ for other info
16:16 lambdabot Title: Mailing Lists - perl6
16:16 sunnavy your presentation on YAPC::Beijing 2007 is great, I love that.
16:16 agentzh sunnavy: thanks!
16:17 agentzh sunnavy: do you have a commit bit to pugs?
16:17 sunnavy no
16:17 moritz want one?
16:17 sunnavy sure :-)
16:18 agentzh wow, moritz is faster than me ;)
16:18 moritz sunnavy: /msg agentzh your email address, and he'll invite you
16:18 sunnavy agentzh: do you like BBS?
16:18 sunnavy moritz: thanks
16:19 moritz I would if my proxy timeout wasn't that tight :(
16:19 agentzh sunnavy: not very
16:21 agentzh sending out commit bits is no longer a trivial task these days...
16:21 moritz yay, commitbit works for me, so you can send me your email as well (if you want two accesses ;-)
16:22 agentzh moritz: commit bit sent?
16:23 agentzh it's very slow here...
16:23 sunnavy ok, I'll send my email to moritz.
16:25 moritz sunnavy: ok, tell me if the mail arrived. If it did not arrive in 15min or something I'll retry
16:26 ayrnieu joined #perl6
16:27 sunnavy moritz: I got it :-)
16:28 sunnavy thanks
16:28 agentzh sunnavy: add your name to AUTHORS to test the commitbit
16:28 sunnavy ok
16:28 agentzh and...welcome aboard!
16:29 sunnavy I'm glad to help. I love Perl.
16:30 agentzh yay
16:32 amnesiac joined #perl6
16:33 agentzh wow, 162 people in the room...
16:33 agentzh more than before...it used to be 148 or so.
16:34 moritz agentzh: more people reading our bullshit? *g*
16:34 agentzh lol
16:36 moritz and I have about a hundret hits per day on the web irc logs
16:36 fglock joined #perl6
16:37 agentzh that's cool
16:37 agentzh moritz: when will full-text search land to your irclog?
16:37 moritz agentzh: good question... next question... ;)
16:38 agentzh will the search results be properly paged?
16:38 moritz which of those two features would you like to have implemented first? *g*
16:38 * agentzh dodges.
16:39 moritz I think the paging is easier to implement for a start
16:40 agentzh that might be
16:41 jisom joined #perl6
16:41 justatheory joined #perl6
16:41 agentzh end of day for me &
16:41 agentzh left #perl6
16:48 svnbot6 r16698 | sunnavy++ | add sunnavy to AUTHORS
16:48 sunnavy seems commit bit worked :-)
16:48 moritz yay
16:48 moritz sunnavy++
16:48 drupek12 joined #perl6
16:49 sunnavy thanks, moritz
16:49 sunnavy or should I say moritz++ ?
16:49 moritz whatever you want ;)
16:49 sunnavy moritz++
16:52 sunnavy I'll dive into pugs and do something soon, wish the `something' is not bad ;-)
16:53 moritz sunnavy: are you more a perl or more a haskell coder?
16:54 sunnavy moritz: perl
16:55 sunnavy I know a bit about haskell but I don't think that's enough to hack pugs :-)
16:55 TimToady even perl programmers are welcome here.  :)
16:56 moritz sunnavy: same for me.. maybe you/we could help fglock with kp6...
16:56 moritz or helping the parrot folks to reorganize/verify/update the test suite
16:57 chris2 joined #perl6
16:58 [particle] last night at seattle.pm somebody mentioned that he started coding perl so he didn't have to write large c programs anymore.
16:58 sunnavy TimToady: so much honour to you! I love your language very much.
16:58 [particle] i recalled that i started for the same reason.
16:58 [particle] then i realised that i'm now writing a large c program (parrot)
16:58 pmichaud me also
16:58 [particle] ...so i can program in perl!
16:59 pmichaud (perl to avoid writing large C programs :-)
16:59 sunnavy moritz: I'll make a dicision soon, but not now :-)
17:00 moritz sunnavy: ok
17:00 moritz ... what was perl written in again? *g*
17:00 Jmax MS VB?
17:03 weinigLap joined #perl6
17:07 TimToady It's all about being Lazy rather than lazy.
17:08 TimToady People who are Lazy work very hard at it.
17:10 shachaf TimToady: Haskell is very Lazy. :-)
17:10 [particle] you mean i have to use the shift key? i'll stick with lazy :P
17:11 [particle] btw i won't be doing much string concatenation in perl 6. ~ is hard to reach.
17:14 Tene [particle]: just remap it to your 'a' key
17:15 TimToady ++ is twice as hard as ~
17:15 spinclad for me it's RAlt+Shift+~, which is a bit of a stretch for one hand
17:15 TimToady (well, not really...)
17:16 TimToady in any case, I don't mind discourage people a bit from using ~
17:16 TimToady *discouraging
17:16 TimToady usually it's much more readable to use interpolation
17:17 [particle] yeah, even though interpolation requires shift to get a quoting op, my muscle memory takes care of that, and i only need two quotes when i may need multiple tildes
17:17 spinclad ?eval ~ reverse 4!
17:18 evalbot_r16681 changed the nick to evalbot_r16698
17:18 evalbot_r16698 Error: ␤Unexpected "!"␤expecting "_", fraction, exponent, term postfix, operator, ":" or ","
17:18 TimToady O(2) often beats O(n) for large values of n
17:18 TimToady it's tempting to make postfix:<!> a builtin
17:18 fglock I need help with a piece of perl 5 code - http://svn.pugscode.org/pugs/v6/v6-KindaPerl6/lib/KindaPerl6/Perl5/Pad.pm
17:18 lambdabot http://tinyurl.com/ysae7e
17:19 TimToady I don't know anything about Perl 5
17:19 fglock I guess the equivalent Perl 6 code will have the same problem
17:19 fglock which is how to create closures incrementally
17:20 spinclad ?eval sub postfix:<!> { [*] ^ $^n }; ~ reverse 4!
17:20 evalbot_r16698 Error: ␤Unexpected "!"␤expecting "_", fraction, exponent, term postfix, operator, ":" or ","
17:20 fglock as in the compiler environment with BEGIN blocks and declarations
17:21 fglock I've got working code, but it could look better
17:21 spinclad oop
17:21 spinclad ?eval sub postfix:<!> { [*] ^ $^n }; reverse ~ 4!
17:21 evalbot_r16698 Error: ␤Unexpected "!"␤expecting "_", fraction, exponent, term postfix, operator, ":" or ","
17:23 Mankod joined #perl6
17:23 Psyche^ joined #perl6
17:29 TimToady sorry, a bit distracted today; one of our kitties is in the final stage of kidney failure and will probably not last till tomorrow... :(
17:30 justatheory joined #perl6
17:35 fglock TimToady: sorry about that
17:36 fglock TimToady: I think I found a way to fix the technical problems I had with my p6regex-to-p5regex compiler (backtracking with large match objects was expensive)
17:36 TimToady cool!
17:37 * spinclad says metta for kitty
17:38 TimToady spinclad: you want:
17:38 TimToady ?eval sub postfix:<!> ($n) { [*] 1..$n }; reverse ~ 4!
17:38 evalbot_r16698 "42"
17:38 TimToady placeholders don't work there yet
17:38 TimToady and [*] ^$n is always 0
17:38 Tene Heh.
17:40 spinclad ah.  (right: 1 +>> ^ $^n or so.)  here i was enjoying the confusing flow of punctuation...
17:40 Psyche^ changed the nick to Patterner
17:41 rindolf joined #perl6
17:42 TimToady according to current spec ^$n + 1 should do that too
17:42 TimToady but we don't really have Range objects yet
17:43 [particle] $deer..$antelope
17:45 justatheory joined #perl6
17:45 spinclad $home ~~ ^$^n
17:45 justatheory joined #perl6
17:47 TimToady shouldn't that be $home x 2
17:47 spinclad and in o/~ brackets o/~
17:49 spinclad o/~ ($Home x 2, $deer, $antelope) ~~ ^$^n o/~
17:50 spinclad loses a bit in this recension, maybe
17:50 TimToady that's a discouraging word
17:51 spinclad :(
17:51 TimToady fortunately it's seldom heard
17:52 [particle] $skies !~~ 'cloudy all day'
17:52 ruoso joined #perl6
17:53 vel joined #perl6
17:53 spinclad for @day -> $_ { !cloudy(@skies) }
17:55 lidden ?eval sub postfix:<!> ($n) { [*] 1..$n }; 0!
17:55 evalbot_r16698 1
17:57 lidden ?eval sub postfix:<!> ($n) { [*] 1..$n }; -1!
17:57 evalbot_r16698 1
17:57 spinclad lost a pole
17:57 spinclad ?eval sub postfix:<!> ($n) { [*] 1..$n }; 0.5!
17:57 evalbot_r16698 1/1
17:57 spinclad damnit.  i wanted 1/sqrt(pi) or whatever it is
17:58 diakopter Tene: returning your msg
17:58 TimToady -1! should have come out -1 because ! should have defaulted to ++ precedence
17:58 TimToady which is tighter than -
17:59 Tene diakopter: you were mentioning having commitbit open right before.  The person asking for commit access seemed to finally have his client situation sorted out.
18:00 [particle] how would you assert than $n is a positive integer?
18:00 Tene 'where'
18:00 [particle] i mean, what's the perl 6 idiom for that, in a subroutine definition
18:00 Tene I believe
18:00 Tene subset
18:01 [particle] foo (Int $n where $n >= 1)?
18:04 Tene I believe so.  Might want to check synopses.  Don't know if pugs implements it yet.
18:04 [particle] yeah, that's what i'd do if i weren't lazy :)
18:12 dduncan joined #perl6
18:15 isaacd__ joined #perl6
18:17 * fglock wonders how to split big questions into smaller answerable things
18:18 moritz fglock: use split() ;-)
18:21 fglock may I try to explain the problem to you? this would help me to write a readable question
18:22 moritz fglock: of course
18:23 moritz though I'll probably not able to answer it ;)
18:23 fglock when the compiler finds this:  {  my $x;
18:24 fglock it needs to construct a runnable representation
18:24 fglock one way to do it is:  $env = eval ' sub { $x } '
18:25 fglock I mean: $env = eval ' sub { my $x } '
18:25 fglock ok?
18:26 moritz I don't quite understand why that works
18:26 moritz $env is closure after that, right?
18:27 moritz does this do scoping right?
18:27 fglock yes - $env is a representation of the current state of the program under compilation
18:28 moritz but doesn't that overwrite existing variables in that (dynamic) scope?
18:28 moritz or are they stored in another $env?
18:29 fglock no, because the variables only exist inside the closure
18:29 moritz but other could, that were previously stored in $env
18:29 fglock things outside $env are not affected, so the compiler and the program don't mix (much)
18:29 moritz ok
18:30 fglock yes, I'll get there
18:30 moritz go ahead, then
18:30 fglock the compiler proceeds, and it sees the next statement:  {  my $x; BEGIN { $x = 1 }
18:31 fglock now you have to tell $env to change the value in $x
18:31 fglock you can't to this easily, so I wrote a module that works around it
18:32 moritz is it really a good idea to store local variables in closures?
18:32 moritz as opposed to, say hashes?
18:33 fglock that's the point - I don't really know; I'd like to discuss that
18:33 fglock one advantage of closures is that you can execute real code inside them
18:34 fglock hashes would require some emulation
18:34 moritz and they are fast, I presume
18:34 fglock otoh, closures require an eval(), which can be expensive
18:34 moritz but you could store ASTs in the hashes, and execute / emit them when you need the value
18:35 moritz was  that the question you wanted to ask initially?
18:36 fglock yes - I have a working implementation, but I'd like to sanity check the architecture
18:38 moritz so how much magic (or braindamage) do you need to modify the closures?
18:38 fglock re AST - I keep both the AST and $env, because I need them for different things
18:39 fglock it's easier to explain if you look at the code
18:39 moritz in v6-KindaPerl6 ?
18:40 fglock http://svn.pugscode.org/pugs/v6/v6-KindaPerl6/lib/KindaPerl6/Perl5/Pad.pm
18:40 lambdabot http://tinyurl.com/ysae7e
18:41 moritz looking...
18:42 moritz so this is specific to the Perl5 backend, right?
18:42 fglock it's a prototype, I'll rewrite it in MiniPerl6
18:43 moritz so you are limiting possible backends to languages that have closures?
18:44 fglock yes, but languages that don't have closures would have to emulate anyway; this is solved at the MiniPerl6 level
18:44 moritz ok
18:44 fglock you create a Pad with something like: my $env1 = Pad->new( outer => undef, lexicals => [ ::Var(...) ] );
18:45 fglock where ::Var is the AST for a variable
18:45 moritz ok
18:45 moritz the could is very readable
18:45 moritz s/could/code/
18:45 fglock outer is a possible outer pad
18:45 moritz or undef if none
18:46 fglock you add more lexicals with: $env1.add_lexicals( [ ::Pad(...) ] )
18:47 iblechbot joined #perl6
18:48 buetow joined #perl6
18:48 fglock the hack is that when the closure is defined, it returns a sub that does an eval()
18:48 moritz which sounds scary
18:48 fglock this way you can execute things in the closure environment
18:49 moritz but that again is very perl5 specific, isn't it?
18:49 fglock it works under Perl 6 too
18:49 fglock which is what I need, for 6-in-6
18:49 moritz ok ;)
18:51 fglock correction - you add more lexicals with: $env1.add_lexicals( [ ::Var(...) ] )
18:51 moritz that makes more sense, yes
18:53 spinclad (having read Pad.pm -- yes, it's good clear code)
18:54 fglock this way, you have access to the whole pad stack, with introspection; and you can incrementally execute code inside that environment
18:54 moritz sounds clever, but somwhat... still scary ;)
18:55 spinclad it makes sense to me you have to build an 'evaluator' closure
18:56 moritz *shudder* somebody on #debconf just posted and `uptime` output - with a load > 3000 ;)
18:56 moritz on one of the debian webservers
18:57 spinclad i could imagine that on a Thinking Machine, with its N thousand processors...
18:57 fglock moritz: what does it mean?
18:57 rindolf Hi moritz
18:57 moritz fglock: maybe you could/should ask the same question on p6c, perhaps partly using copy&paste from the irclogs...
18:58 moritz fglock: if you break down that "compiler review" question into smaller questions that can be answered in half an hour each, you'll get more feedback
18:58 moritz load = number of processess waiting for CPU and/or IO resources
18:58 fglock ok - I'll try this question first
18:59 fglock then I'll come back here with the other ones :)
18:59 moritz of course ;)
19:00 spinclad ... and i see you're taking advantage of Perl 5's  eval($expr)  in a package context ...
19:01 moritz fglock: and be sure to add the direct link to Pad.pm ;)
19:02 unobe_away changed the nick to unobe
19:03 fglock spinclad: yes, it needs to execute in the appropriate run-time package - the current package is the compiler's one
19:03 fglock s/package/namespace/
19:05 fglock what's the appropriate mailing list? perl6-language is generic, and perl6-compiler is for Parrot things, right?
19:06 spinclad ... i look for a way to build a Perl 5 structure for the pad, have a fixed algorithm to lookup($var) in it, and build up translated expressions with $var ==> lookup($var)
19:06 moritz no, parrot folks have their own mailing list
19:06 pmichaud definitely not perl6-language
19:07 pmichaud perl6-compiler would seem most appropriate.  It's not necessarily Parrot specific
19:07 fglock pmichaud: ok
19:07 moritz fglock: I think p6c is fine
19:07 obvio171 joined #perl6
19:08 fglock spinclad: thanks :)
19:09 fglock pmichaud: I wonder if you had a similar problem in parrot
19:09 pmichaud I haven't been following the discussion that closely
19:09 pmichaud chances are that I haven't encountered the problem in parrot, though
19:11 fglock in short, how do you handle the compiling environment - where do you store lexical things that were just compiled
19:11 fglock such that BEGIN block can see it
19:11 fglock blocks
19:11 pmichaud oh.  I haven't dealt with that yet
19:12 fglock cool - we could try to find that out together :)
19:12 pmichaud indeed!
19:13 fglock I have a working fix, but it looks too complicated
19:13 spinclad whatever structure (AST) you translate code into, you need to be able to .eval it and capture its effects in the compiler, as well as defer that to runtime
19:14 pmichaud spinclad: exactly.  That makes it a bit tricky :-)
19:14 fglock spinclad: yes
19:15 fglock pmichaud: this covers 2/3 of my question in the list :)
19:15 fglock the other 1/3 is easier
19:17 fglock pmichaud: if you have some time, I'd love to guide you through kp6
19:17 pmichaud fglock: I'll be happy for a tour, but it may be Friday before I can easily get to that
19:18 fglock sure
19:18 pmichaud fglock: if not Friday or Sat, it may need to wait until Tuesday
19:18 pmichaud I'm trying to get all of my stuff ready for YAPC::NA
19:19 fglock np - I need kp6 for YAPC::EU :P
19:19 moritz I'll to to be there (EU)
19:19 fglock moritz: cool
19:19 moritz s:1st/to/try/
19:20 pbuetow joined #perl6
19:21 fglock pmichaud: I'm looking for a more solid architecture, which would allow kp6 to evolve to full-p6
19:23 pmichaud fglock: sure thing -- walking through kp6 would probably help both of us in that respect
19:25 fglock pmichaud: I found a way to plug PGE into the architecture, such that kp6-parrot can use it (and I need to update the plan)
19:25 pmichaud oh, excellent!
19:26 pmichaud I just added (but haven't committed) some features to PGE to let it handle things like {*} at runtime :-)
19:26 theorbtwo joined #perl6
19:26 fglock can you plug a language parser for the thing inside {...} ?
19:27 fglock PCR does this with subclassing
19:27 pmichaud yes
19:27 fglock nice
19:27 pmichaud you can pass another object that has methods to be called when the {*} is encountered
19:30 fglock interesting - how about multiple {*} occurrences?
19:30 pmichaud it understands the "#= key"  comments as well
19:30 pmichaud and passes the key to the method
19:31 fglock :)
19:31 pmichaud so then the method can switch off of the key
19:32 fglock looks useful - it will end up being specced :)
19:37 svnbot6 r16699 | moritz++ | [irclog] fixed adding of linebreaks to long URLs to prevent horizontal
19:37 svnbot6 r16699 | moritz++ | scrolling
19:54 fglock pmichaud: {*} should actually be removed by a preprocessor; implementing it is a workaround, right?
19:55 pmichaud fglock: I'm thinking {*} might end up in the spec
19:55 pmichaud there could be some advantages to being able to call a grammar with its actions disabled
19:55 spinclad a preprocessor could replace it with its method call, perhaps?
19:56 spinclad and an underived grammar would have those methods as stubs
19:56 pmichaud at any rate, calling {*} when no special object is passed equates to a no-op, as if it had been removed by a preprocessor
19:57 pmichaud (or an empty method stub)
19:57 TimToady in theory, the only effect {*} has to have is to terminate a longest-token match
19:58 pmichaud as PGE currently has it, it would be able to do this
19:58 TimToady so a preprocessor might need to treat it as a ::
19:58 pmichaud essentially PGE is saying that it'll understand {*} even if it's not preprocessed out, and do something intelligent with it
19:58 TimToady sounds good
19:59 pmichaud and thus we can compile the grammar even if it has {*} in it, and use different "backends" for the stubs at runtime
19:59 pmichaud this turned out to be easier (for now) than writing the preprocessor :-P
20:00 TimToady phone call?
20:00 pmichaud earlier I was playing with using something like  {*#<key>}   instead of {*} # =key
20:00 pmichaud but as it stands now, PGE uses the #= key format
20:00 pmichaud (phone...on my way)
20:01 justatheory joined #perl6
20:01 fglock hmm -  &{ action() }
20:02 vel joined #perl6
20:07 tlb joined #perl6
20:09 PerlJam fglock: {*} is much simpler than &{ action() }
20:09 PerlJam :-)
20:11 fglock yes, but it adds yet another type of comment
20:12 fglock how about {*(key)}
20:12 fglock or simply <action(key)>
20:12 fglock this is very regexy
20:13 pmichaud isn't   #<foo> a comment already?
20:14 japhb joined #perl6
20:15 fglock I mean, why not introduce a new word for "action"; comments are "out of the language"
20:16 fglock <action {...}>
20:16 fglock <action {*}>   # placeholder
20:16 PerlJam What was ... meant to be exactly?
20:16 PerlJam actual code?
20:16 fglock yes
20:17 PerlJam in what language?
20:17 fglock the language is pluggable, just like normal {...}
20:19 fglock placing the {...} inside <action> makes it possible to turn the {...} execution on/off
20:19 fglock it's just a conditional
20:19 pmichaud this would mean that <action ...> needs special parsing, to be able to add the rule name into the call
20:20 fglock yes, or you can use <action: {...}>
20:21 PerlJam not putting actual code in the grammar also allows the exact same grammar to be used with multiple languages easily.
20:22 pmichaud regexes already allow other languages to be in the {...}  via the :lang adverb
20:22 fglock you could use a method for that; language-specific inheritance will define the actual code
20:23 [particle] ho! look what i walked in on. time for a trip to scrollbackland!
20:24 fglock I was missing the old long technical discussions :)
20:24 moritz fglock: time for YAPC ;)
20:25 pmichaud not putting actual code in the grammar allows allows the grammar to be re-purposed for different purposes easily (yes, subclassing can do this also)
20:25 pmichaud (however, subclassing also requires re-writing the rule)
20:25 fglock <my_action>  # subclassable
20:26 thoughtpolice joined #perl6
20:26 pmichaud fglock:  how would you distinguish
20:26 fglock <My::action>  # defined at link-time
20:26 * PerlJam suddenly expects Damian to appear and start PODing the grammar  (speaking of re-purposing things :)
20:26 pmichaud regex foo { foo <my_action> }
20:26 pmichaud regex bar { bar <my_action> }
20:26 weinigLap_ joined #perl6
20:27 fglock it's actually: regex bar { bar <bar_action> }   # if it is a different action
20:27 thoughtpolice anybody have a particular news reader they like? i've been trying to find one; if it runs from the terminal then the better I suppose. :)
20:27 fglock just like a different comment
20:28 pmichaud fglock:  and how do we indicate that it's okay for <bar_action> to be missing?
20:28 pmichaud part of the point of {*} is that it's okay to have a null action
20:28 PerlJam pm:  <> seems logical in fglock's fantasy
20:28 pmichaud fglock:  and do we really want....
20:29 pmichaud rule prefix_meta_infix:sym<*> { <sym>  <prefix_meta_infix_action> }   ?
20:29 fglock pmichaud: maybe with:  method bar_action {*}
20:29 PerlJam in any case, regex foo { foo <foo_action> } regex foo_action { ... }  # could cause some accidental conflict if we aren't careful.
20:30 Aankhen`` @tell agentzh I didn't fix it… I couldn't even reproduce it. :-S
20:30 lambdabot Consider it noted.
20:30 PerlJam {*} has the benefit of being ordinarily illegal (except for the specific purpose of being ignored or substituted)
20:30 [particle] fglock: pmichaud briefly considered: action bar {...}
20:30 pmichaud anyway, I'm sticking with {*} until Larry changes it :-)
20:30 offby1 left #perl6
20:30 fglock pmichaud: sure
20:30 [particle] but i like {*}, and it's what's specced anyway?
20:30 [particle] s/\?/:\)/
20:30 pmichaud actually, I don't think {*} is "specced" outside of STD.pm yet
20:31 PerlJam yeah, specced by convention
20:31 [particle] de facto
20:31 [particle] the head cheese put it in writing.
20:31 pmichaud PGE goes a step further by actually parsing it in non-preprocessed-regexes and doing something with it
20:31 riffraff joined #perl6
20:31 moritz specced the perl 5 way
20:31 pmichaud as opposed to having a pre-processor do it
20:31 riffraff hi
20:31 moritz hi riffraff
20:32 [particle] pmichaud: is it desirable to make that configurable?
20:32 pmichaud [particle]: configurable how?
20:32 PerlJam [particle]: if a preprocessor has gotten to it before PGE, there's nothing to configure  :)
20:32 pmichaud you mean, such that PGE strictly honors the spec and treats {*} as a closure?  Sure
20:32 [particle] well, i suppose preprocessor selection, but maybe pge won't control that
20:32 [particle] and yes, also what you said
20:33 fglock pmichaud: in any case, if you are doing it the OO way there is no need for {*} or <action>; it just works
20:33 [particle] that's what i was initially thinking of.
20:34 Alias_ joined #perl6
20:42 mico joined #perl6
21:09 explorer joined #perl6
21:13 polettix joined #perl6
21:13 jferrero joined #perl6
21:15 hirschnase joined #perl6
21:21 explorer joined #perl6
21:31 dduncan Do any of you know of a web-connected perltidy utility, so one can just paste perl code into a form and have it processed by the site ... as an alternative to installing perltidy on one's own machine?
21:32 [particle] maybe a nopaste somewhere?
21:33 moritz dduncan: I could set one up, it would just take me some time
21:33 dduncan don't worry about it
21:33 dduncan and I didn't think of nopaste, since this was just for my own reading, not for sharing ... though I could pick a bogus channel
21:34 [particle] some nopastes allow you not to specify a channel at all
21:34 dduncan or it has a 'none' option ... nice
21:37 dduncan hm, the 2 pastebots I tried didn't have any filtering options ... will look for others ...
21:39 weinigLap joined #perl6
21:40 [particle] joined #perl6
21:44 meppl good night
21:48 Limbic_Region joined #perl6
21:48 pasteling "test" at 24.69.53.198 pasted "test" (76 lines, 1.7K) at http://sial.org/pbot/25693
21:49 dduncan oops
21:49 dduncan forgot 'none'
21:49 dduncan anyway, I got it to work, thanks anyone
22:16 Yaakov This is a great spot.  I really like it.  http://www.youtube.com/watch?v=ANfFwFqTS9o
22:18 Psyche^ joined #perl6
22:18 Psyche^ changed the nick to Patterner
22:18 ibrown joined #perl6
22:18 ibrown what's good guys?
22:18 Psyche^ joined #perl6
22:18 Psyche^ changed the nick to Patterner
22:19 ibrown i got a question, i tried to add a smartlink, how do i check it's correct before commit?  i did a svn committ but the url: http://dev.perl.org/perl6/doc/design/syn/S02.html doesn't reflect the changes.
22:19 lambdabot Title: Synopsis 2: Bits and Pieces - perl6:
22:19 Tene ibrown: that's updated daily, I believe.
22:19 moritz ibrown: see t/README
22:19 [particle] ibrown: i think there's an option on smartlinks.pl to check your links
22:20 ibrown yeah i did the commit this am, or last night i thought.  just didn't know if there was a better way to verify the smartlink was in place right.
22:20 ibrown yeah there's a --check option... it outputs a file called S26.html in whatever dir i run it in tho
22:20 ibrown after it states, "3 smartlinks found 0 errors"
22:20 moritz which means it worked ;)
22:20 ibrown cool.
22:20 ibrown i was more worried about placement ;)
22:21 moritz ibrown: ah, that. Look at http://perlcabal.org/syn/
22:21 moritz that's where you can see the smartlinks
22:22 ibrown oh awesome.
22:22 ibrown i can see where it added the counter test to S02.
22:22 ibrown that test passed on my box tho, that's where the make smoke comes in?
22:22 Tene Maybe we should set it to be rebuilt on every svn commit.  feather can handle the extra load, right? ;)
22:22 moritz Tene: no, it can't :(
22:23 [particle] at the current rate of commits? sure! :P
22:23 moritz ibrown: there's a daily smoke on feather.perl6.nl
22:23 moritz ibrown: that provides the success/failure markings
22:23 perljunkie I have one application. It is split into 2 parts. A data collection suite, and data explorer. Currently split into 2 databases. generally. Would a 1 or 2 database design be standard?
22:23 moritz ibrown: and your own reports are at smoke.pugscode.org
22:23 perljunkie damn. wrong channel. sorry
22:25 ibrown well.  it appears as if i get it now.
22:25 ibrown hoorah.
22:25 lambdabot joined #perl6
22:25 ibrown nifty process from what i can tell too.  
22:26 moritz indeed ;)
22:26 moritz Aankhen``++ , agentzh++
22:27 [particle] (spec-based testing)++
22:28 ibrown what's the integers on the http://m19s28.vlinux.de/cgi-bin/pugs-smokeserv.pl page.  after the % ok? is that test-suite information?  like # passed vs. # failed etc?
22:28 lambdabot Title: Pugs Smoke Reports
22:32 moritz looking...
22:32 Limbic_Region ibrown - number of tests total, passed, etc
22:32 ibrown rgr
22:32 Limbic_Region if you click the >> link to expand the view
22:32 Limbic_Region it will make much more sense
22:33 moritz pugs revison : spec revison  fail pass todo
22:33 moritz or something
22:33 Limbic_Region nope
22:33 Limbic_Region after the %
22:33 Limbic_Region not before
22:33 moritz Pugs 6.2.13    18 Oct 2006 02:09 Wed  90.03 min  100.00 % ok  17954:  17954,  0,  4638,  747,  0
22:34 [particle] oh, yeah.
22:34 moritz there the 17954 look like revison numbers to me
22:34 [particle] revision first
22:34 Limbic_Region total test case, passed cases, failed cases, todo cases, skipped cases, unexpectedly succeeded cases
22:34 [particle] then failed, total, todo, skipped, iirc
22:34 [particle] right, unexpected success, thanks lr
22:34 ibrown so from what i can tell PIR is perl6 on parrot, which is the way i've got pugs running i thought... so the last make-smoke done on a linux kernel was 10 feb 2k7.  maybe you need a new smoke?
22:35 Limbic_Region oh, my eye just skipped over the rev:
22:35 Limbic_Region heh
22:35 moritz ibrown: not quite
22:35 moritz ibrown: PIR is just parrot assembler
22:35 moritz ibrown: and the PIR backend of pugs is currently b0rked
22:35 [particle] right-o
22:35 ibrown oh.
22:35 Limbic_Region actually, ibrown is talking about "PIR (Perl 6 on Parrot)"
22:36 moritz ok
22:36 Limbic_Region which if I had waited for moritz to finish
22:36 Limbic_Region and the PIR backend of pugs is currently b0rked
22:36 Limbic_Region It wouldn't have mattered
22:36 * Limbic_Region is slow on the uptake today
22:36 moritz and Perl 6 on Parrot lives in the parrot repository
22:36 moritz in languages/perl6/
22:36 Limbic_Region ibrown - the good news is that pmichaud has a perl 6 implementation on parrot written in PGE that is now passing all the sanity tests
22:37 Limbic_Region and very soon going to be taking on the rest of the perl 6 test suite in pugs
22:37 Limbic_Region again, moritz is beating me to the punch
22:38 moritz ;)
22:38 Limbic_Region and even better news is that pmichaud has been gaining ground with contributors
22:38 [particle] i'm modifying Test.pm in perl6/parrot now, and pm is writing a script to download selective tests from the pugs repo
22:39 [particle] soon you'll be able to write perl 6 in perl 6, on parrot
22:39 moritz you can do that with kp6/mp6 as well
22:39 moritz on perl6, that is
22:39 moritz though mp6 is bootstrapped to parrot as well, iirc
22:40 [particle] no, not fully.
22:40 moritz how sad
22:40 moritz I thought it was done at a YAPC hackaton
22:40 [particle] patrick and flavio are working together now, though.
22:40 moritz which is a great thing as well
22:41 [particle] we should soon see the same kind of convergence that gets folks hot about the iphone :)
22:41 moritz p6-on-parrot is now evolving so fast that I have trouble svn up'ping in a timely manner ;)
22:41 pmichaud moritz: thanks for your comments on p6c.  Any further thoughts about how we might organize pugs-specific (and parrot-specific) tests?
22:41 [particle] pmichaud: i say impl-specific tests should not be in the official repo
22:42 moritz pmichaud: perhaps specially regression-test files
22:42 moritz pmichaud: things that are known bugs in one implementation
22:42 [particle] since the official repo is in pugs now, perhaps we can put it in a separate dir
22:42 moritz of course all others have to pass them as well ;)
22:42 pmichaud well, we can use things like :pugs<todo> and :parrot<todo> to mark specific tests as per-implementation todo
22:43 [particle] i'm sorry, i was thinking about implementation-specific non-spec behavior
22:43 moritz pmichaud: I know, but that's beside the point...
22:43 [particle] duh.
22:43 pmichaud [particle]: yes, that's what I'm thinking about also
22:43 moritz pmichaud: the ternary test you came up with is just _weird_
22:43 pmichaud here, the test is really testing for a pugs-specific parsing bug
22:43 moritz right
22:43 [particle] *unspecced behavior
22:43 [particle] yeah.
22:43 [particle] that test should not be in the official repo
22:43 pmichaud so, I'm thinking (hoping) it can go somewhere else
22:44 moritz so it is specced, and the test is needed, but its' not primarily a ternary test
22:44 weinigLap joined #perl6
22:45 moritz maybe we should move that discussion back to p6c...
22:45 pmichaud moritz: probably we should.  I'm not sure what to add at the moment (but have also been focused on other things today)
22:45 moritz before I move such tests I'd like to have a green light from TimToady or audreyt, because it's a rather large policy change
22:46 moritz pmichaud: I'll write another mail before going to be
22:46 pmichaud oh, I've already been given a bit of a green light to refactor tests
22:46 moritz s/be/bed
22:46 pmichaud moritz: yes, I'd appreciate that very much
22:46 moritz pmichaud: ok
22:54 moritz [X] Mail sent
22:55 pmichaud moritz++
22:56 moritz cool, blead passes all smoke tests on my maschine ;)
22:56 larsen_ joined #perl6
22:56 moritz pmichaud: I help wherever I can without any compiler building knowledge... so if I can help more with the test suite, be sure to let me know
22:58 ibrown so where does perl6 really need dev help?
22:59 ibrown i guess is a good question from what i read while away
22:59 moritz ibrown: what do you want to code?
22:59 moritz ibrown: haskell, c, perl5, perl6?
22:59 ibrown well.
22:59 moritz doc?
23:00 ibrown i do perl5, php, ruby, some C, i could probably adapt to C++ my english and explanatory skills are lacking, normally i delegate doc projects lol
23:00 ibrown haskell i looked at yesterday for the first time, and first thing i thought was, how fucking cool, second thing was i agreed with toady, big learning curve from a more atypical structured language
23:00 obvio171_ joined #perl6
23:00 moritz ibrown: there are quite a lot things you can do...
23:01 ibrown i mean here's the deal.  i don't want to switch from perl.
23:01 moritz ibrown: you could help fglock with his v6.pm, mp6 and kp6 compilers
23:01 ibrown so whatever i can do to ensure that A: it gets released. B: it doesn't fail on the web i'm kosher for
23:02 moritz ibrown: or you could learn perl 6, and port helpfull modules from p5 to p6...
23:02 ibrown i was hoping to get involved in some module writing, for 2 reasons, A: to do it, and B: to ensure that i understand perl6 at a much greater level than 5
23:02 ibrown well, the .pm's i know the most are dbi, and ldap/ad pms.  heh.
23:02 moritz then translating modules from 5 to 6 might be a very good idea to start with
23:03 ibrown i heard a podcast that the guy that wrote the DBI module is actually writing the perl6 dbi module, is that confirmed or was he speaking out his arse?
23:03 moritz ibrown: maybe you should start with a simpler one ;)
23:03 [particle] tim bunce is working on dbi 2.0 for perl 6
23:03 moritz cool ;)
23:03 [particle] he wants a jdbc-like api
23:03 [particle] but it hasn't been designed yet, and help is welcome there
23:03 ibrown haha oh i dont' want to write DBI/LDAP off the bat, i'll start with more simple things, data::dumper, and b.s. like that, along with whatever other standardized modules need work.
23:04 ibrown any spot to see a good TODO?
23:04 moritz ibrown: Data::Dumper is not needed, you have the .perl method ;)
23:04 moritz ?eval <a b c>.perl
23:04 Limbic_Region well, actually Tim said one of the things blocking him was a Java -> Perl 6 translator
23:04 evalbot_r16698 changed the nick to evalbot_r16699
23:04 evalbot_r16699 "(\"a\", \"b\", \"c\")"
23:04 ibrown hrm.
23:04 Limbic_Region and phil crow took up that torch
23:04 Limbic_Region and has put his work to date on CPAN
23:05 ibrown so do print Dumper($a) now you just ?eval <a>.perl ?
23:05 Limbic_Region it mostly just does interfaces ATM - not classes
23:05 moritz ibrown: the '?eval' part is only to start the bot...
23:05 daxim what's the recent talk on the free perl6 book?
23:05 ibrown oh my bad lol
23:05 moritz ibrown: so Dumper($foo) is $foo.perl or perl $foo
23:05 weinigLap joined #perl6
23:05 ibrown haha that's awesome
23:05 moritz daxim: that's "Perl 6 and Parrot Essentials"...
23:05 Limbic_Region daxim - the Perl 6 and Parrot Essentials from O'Reilly has been re-licensed if that's what you mean
23:06 Limbic_Region mortiz - stop that
23:06 Limbic_Region :P
23:06 ibrown so finish reading perl6, and then pick a pm and get to cruxing eh?
23:06 moritz daxim: you can find the perl6 parts in the pugs repository in pugs/docs/tutorial/
23:06 daxim so much I know, what are the details?
23:06 moritz ibrown: right
23:06 moritz daxim: http://www.nntp.perl.org/group/perl.perl6.language/2007/06/msg27767.html
23:06 ibrown along with documenting the test scripts ;)
23:06 lambdabot Title: Perl 6 & Parrot Essentials as project documentation - nntp.perl.org, http://tinyurl.com/ynvlty
23:06 Limbic_Region ibrown - I wrote Config::Tiny in Perl 6 before pugs supported an object system - not hard at all
23:06 moritz daxim: and http://moritz.faui2k3.org/pugs/docs/tutorial/ch01_overview.pod.html
23:07 lambdabot Title: , http://tinyurl.com/2s96ys
23:07 ibrown limbig_region: suggestions for .pm's that need immediate attention?
23:07 moritz that's the first chapter as .html
23:07 moritz perhaps HTTP.pm?
23:07 ibrown http.pm right.
23:07 moritz but you should ask Juerd about that
23:08 Limbic_Region ibrown - I don't have any suggestions but you should browse the ext/ directory in pugs as well as the examples/ directories
23:12 ibrown kk running to bodega, i'll check into some stuff til dinner time with the grrl when i get back.  i'll shoot my .pm decision through the IRC channel and then throw it up on the perl-internals list?
23:13 ibrown or is that a perl-language related issue?
23:15 moritz you could use perl6-users oder perl6-compiler
23:16 moritz or perl6-language if you have questions about perl 6
23:25 thestarslookdown joined #perl6
23:27 moritz anyway, I'm gonna get some sleep
23:28 REPLeffect joined #perl6
23:32 felipe joined #perl6
23:33 marmic joined #perl6
23:39 prism joined #perl6
23:42 REPLeffect joined #perl6
23:45 Eidolos joined #perl6

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

Perl 6 | Reference Documentation | Rakudo