pugscode.org <<Overview Journal Logs>> | r1773/winxp: 587/4256 smoke/win2k (occasional): xrl.us/fqum | pugs.kwiki.org
Set by theorbtwo on 11 April 2005.
Ovid Crap. My HTML::Entities still fails. 00:00
stevan Ovid many of those are force TODO so they dont really pass 00:16
the tests are a few weeks (even months in some places) ahead of development
Ovid Makes sense. I still wish I could figure out why HTML::Entities fails, though. When I have time, I'll try to produce a minimal test case for checking somewhere. 00:17
stevan Ovid: After I finish all the test stuff and changelog I will take a look (assuming I dont just pass out) 00:19
Ovid Heh :)
I wish I could get away with that. Unfortunately, work beckons.
Good work, though, so I can't complain.
stevan I am in between projects,.. so I got tuits 00:20
Ovid Wow. It's really pouring down outside. I love it!
stevan but soon my next project starts to my commit level will go down
Ovid Well, I want to devote my weekends to this, but the guy who's working on the new parser for AI::Prolog thinks he might have it done soon, so I'll need to get that polished. 00:22
stevan BTW - very cool stuff with that AI::Prolog 00:23
I have always been a fan of Prolog 00:24
Ovid Thank you. After I get it a bit more stable, I want to port it Parrot.
stevan and Erlang, which looks a lot like Prolog and whose first prototype was actually in prolog
Ovid: that would be very cool
Ovid I would prefer to do a Warren Abstract Machine, but I don't know them very well. A first pass with an engine I know would be better.
stevan I have high hopes for Parrot,.. it is all the potential I saw in the JVM but which never happened
Ovid Actually, I'd be rather interested in Mercury. I've heard good things about it. 00:25
stevan I have read a bit about WAMs, but that goes over my head pretty quickly
I think I have heard of that
you know shapr is a language fetishist too
he sent me a few links to some really weird stuff
not Brainf*ck weird either,.. PhD thesis weird 00:26
Ovid Have you read www.vanx.org/archive/wam/wam.html? It's a tutorial reconstruction of the WAM. It makes more sense, though, if you're familiar with Prolog and how it's backtracking works (understanding choicepoints and all that)
stevan follows the link 00:27
mugwump Oz is based on a WAM AIUI 00:28
stevan hmm, these are the kind of PDFs which make me miss access to corporate laser printers
Ovid Really? 00:29
(I mean about Oz)
mugwump so buy one second hand, they're throwing them away these days
stevan mugwump: not a bad idea actually :)
mugwump I was tempted by a ~$400 new colour HP laser with JetDirect 00:31
crysflame mmm, neat 00:32
you'll bleed toner bills for life
mugwump Then I remembered that HP is an ink company that sells PCs and servers on the side
:)
jabbot pugs - 1829 - First pass at cookbook entry for interpo 01:13
pugs - 1828 - First pass at reversing chars and words
pugs - 1830 - First pass at cookbook entry for indenti 01:23
hlen maybe they should create a perl6 section in perlmonks 01:33
jabbot pugs - 1831 - more force-TODO and test cleanup
mugwump dear god no!
jabbot pugs - 1832 - First pass at cookbook entry for trimmin 01:43
buu Is there anyway to use the current pugs to limit execution time and such? 01:51
02:37 metaperl_ is now known as metaperl
ninereasons maybe I should understand your question, buu; but I don't 02:45
buu I wish to embed a language 02:46
I'd love to embed perl6
But I need to control how long the program executes, system calls, memory, etc
ninereasons hm.. i might be mistaken, buu; but I don't think that there is a way to do that yet. 02:47
you might look through the test suite though, and see if the features you need are being tested, and are succeeding. 02:48
buu =[
I don't think this is part of the test suite 02:49
ninereasons should it be, do you think? 02:50
buu er
Obviously *I* think it should be, but I'm not sure anyone else does
ninereasons I'm not sure that what anybody else thinks really matters . If you're able to write appropriate tests, then it seems to me that you should do so. 02:51
buu Er
I don't think the features I want are even planned, how would I write tests for them?
ninereasons you have a point there, if that's the case :) 02:52
buu I don't think anyone has contemplated it yet
It's more of an interpreter issue than a language issue 02:53
ninereasons good luck with it :)
buu ...
Thanks 02:54
autrijus hi lambdacamels 03:00
buu: what's your host language? 03:01
buu autrijus: Probably perl
Or possibly pike 03:02
Or pathenthetically python!
Hell, I might do it in p6 if it's advanced enough yet
buu pokes autrijus 03:04
autrijus hrmph. 03:08
buu: your best bet is to help us working on pugscc --parrot
buu I demand you drop everything you're doing and solve my problem right now!
Er, what is pugscc?
autrijus then you get to use embed parrot for free
and just feed it IMC.
buu And can parrot do those tricks?
autrijus yes.
buu Really?!
Where is that documented
autrijus while GHC also can, embedding GHC is trickier. 03:09
buu How about pugs
autrijus pugs is just a compiler.
buu Er, what executes the compiled code? 03:10
autrijus you don't hear "embed gcc" that often :)
either GHC, or Parrot.
pugscc has different backends.
see mod_parrot for concrete example
buu Freaky.
autrijus author of mod_parrot says he'd be working on mod_pugs. 03:11
which seems to be exactly what you want.
www.parrotcode.org/docs/embed.html # embedding parrot
buu hrm
Sounds keen.
autrijus www.smashing.org/mod_parrot/ # mod_parrot
buu How far along is pugscc? 03:13
mugwump mod_segfault 03:17
autrijus buu: well it can compile mandelbrot :) 03:18
but needs more work.
buu heh 03:20
Well 03:21
I'm reading embed parrot and such, I don't really see how to do what I want
autrijus you want pdd18 03:22
# dev.perl.org/perl6/pdd/pdd18_security.html
also see dan's recent posts
buu Where? 03:24
autrijus good question. I can't find it anymore. 03:26
ask around? I know Dan is still continuing the design for quota 03:27
<- not really a parrot person
trying to be, but not quite there
stevan autrijus: hello
autrijus hi stevan! 03:28
stevan the tests have been cleaned
although I see a few that need my attention
buu So eh, it's not implemented yet?
autrijus buu: I'm sure it's not implemented in full 03:29
I don't know which hooks are already in place.
buu hrm
autrijus stevan: I'm trying to kill the mutual recursion now 03:32
seems GHC goes bonkers over it
stevan autrijus: ok, I am cleaning a few more tests that showed up
(I just did a fresh 'make' and 'make test' 03:33
autrijus cool. and I reasoned yesterday that the new binding code for slurpy array is likely to be broken
also normal scalar bindings may also be broken
it seemed to assume "is copy" by default
it's easy to fix, fortunately
so let me tackle this mut.rec. first
stevan ok
is openfoundry down? 03:43
autrijus no... dns however looks weird 03:44
please add this entry 03:46
140.109.17.109 rt.openfoundry.org
I'm looking for fix
or rather 03:53
140.109.17.109 rt.openfoundry.org svn.openfoundry.org
stevan I cant add DNS entries
I have no access to that stuff actually
autrijus that's a real problem then 03:54
stevan I am running another 'make test' to make sure I got everything
autrijus sec.
svn switch --relocate svn.openfoundry.org/ foundry.autrijus.org/ 04:01
that should do.
stevan not working 04:03
autrijus what's the error msg?
stevan svn: PROPFIND request failed on '/'
svn: PROPFIND of '/': 302 Found (foundry.autrijus.org)
autrijus svn switch --relocate svn.openfoundry.org/ pugs.autrijus.org/ 04:08
this should now work.
stevan svn: PROPFIND request failed on '/pugs'
svn: PROPFIND of '/pugs': Could not resolve hostname `pugs.autrijus.org': No address associated with nodename (pugs.autrijus.org)
I can just send you patches 04:09
buu So sad.
autrijus try final time? 04:11
svn switch --relocate svn.openfoundry.org/ openfoundry.dyndns.org/
stevan worked :) 04:12
autrijus =) lunch &
luqui what happened to svn.openfoundry.org? 04:18
stevan luqui: DNS issues
try this:
svn switch --relocate svn.openfoundry.org/ openfoundry.dyndns.org/
luqui cool
stevan :)
luqui it's been awhile... what have you guys done... 04:20
er, should I say, what haven't you done :-)
stevan luqui: oh nothing
luqui got rules yet?
stevan the ChangeLog is fairly up to day
rx:perl5 only
luqui awww 04:21
stevan limited too
buu Hmph
stevan we got CGI.pm :P
sockets, threads
buu Any one have an idea how much of parrot-security has been implemented
Ovid We *sort of* got CGI.pm.
stevan nope, you could try asking on #parrot though
Ovid is much pickier about his CGI than I am :)
Khisanth is this CGI.pm better than the old one? :) 04:22
stevan Khisanth: could it be any worse :P
Khisanth stevan: YEeES
things can always be worse
Alias_ Yeah, CGI could be even LARGER 04:23
stevan Khisanth: it is not OO, but I tried to be cleaner about things
and Ovid is working on it now (which is a very good thing)
luqui I think once grammar munging is available, we'll very soon see PHPerl
Ovid It will be. My thought is that, amongst other things, HTML generating functions will be handled separately and we won't have the odd mixing of sub calls and methods.
luqui (that is, inside-out perl)
Khisanth bleh
buu Ovid: That'd be nice.
stevan Alias_: I will make a special one /just/ for you :)
Alias_ oh joy 04:24
Ovid I've also been working on HTML::Entities, but for some reason, Pugs stops parsing my file partway through. It doesn't throw any errors until I try to access the things that it failed to parse.
stevan Ovid: yes,.. no HTML,.. templates are much smarter anyway
Alias_ Can you port Matt Wright's Formail as well?
buu Where is #parrot?
Ovid Sure! I can see how many of his security holes I can stuff into one script!
stevan Alias_: I thought you were starting the perl6-MSA effort :)
Ovid I should add that the current modules being implemented in ext/ are just to test Pugs. At least for the stuff I've seen, I hope no one intends that they be considered "real". 04:26
stevan I think all of Pugs is still "use at your own risk" status :) 04:27
ingy hola kids 04:28
stevan hey ingy
luqui hi ingy
Ovid Hi ingy.
luqui buu: irc.perl.org
Ovid And hi luqui!
ingy hey luqui, nice to be seeing you in Toronto
hi stevan, Ovid 04:29
luqui but not in portland :-(
ingy when?
Ovid You can't make it this time? That's too bad :(
ingy ah, luqui is not going...
I thought you meant I was not going
luqui yeah, didn't get a talk accepted
Allison is trying to get me a vip pass, but haven't heard anything yet 04:30
ingy thinks it is funny that she is the one giving the pugs talk
oh hell luqui, if that's all you need I wouldn't sweat it 04:31
Ovid You could always sign up as a volunteer. Yeah, you miss a couple of sessions, but you still get in free.
ingy you're definitely high enough on the Perl food chain I would think
luqui :-) 04:32
brb
Ovid stevan, you there?
stevan Ovid: yup
Ovid I'm reducing HTML::Entities to a minimal test case right now so we can figure out why it's not parsing properly.
stevan make sure to try the latest build too 04:33
almost all tests will now pass 04:34
Ovid Sure. As soon as I can check it out :)
stevan Failed 6/203 test scripts, 97.04% okay. 44/4236 subtests failed, 98.96% okay. to be exact
Ovid I'm on 6.0.14
stevan svn switch --relocate svn.openfoundry.org/ openfoundry.dyndns.org/
Ovid Updating now! 04:35
luqui b 04:36
stevan wow,.. a lot of movement on the cookbook 04:38
nice
luqui where are we doing this?
ingy luqui: no comments on S26 yet. have you heard any mumblings?
stevan luqui: doing what? the cookbook? examples/cookbook/*
luqui oh, yeah, I didn't read that. I would have if Aaron did it, but I trusted you, so I was lazy about it 04:39
Ovid Well, I was whoring the cookbook out on Perlmonks.
ingy I guess I'll just have to arm wrestle Larry at the hackathon
luqui do we require that the cookbook examples run?
Ovid no
stevan luqui: nope (but it does help)
luqui I'll look it over one of these days
jabbot pugs - 1833 - even more test cleaning and force-TODO-i 04:40
pugs - 1834 - * add unicode include headers to avoid w
pugs - 1835 - * remove warnings from hsplugins probing
pugs - 1836 - More progress on S28draft.pod
pugs - 1837 - the last of the force-TODO and test clea
ingy goes off to lift weights
stevan ingy: no CPAN drinking game?
luqui You should arm wrestle gloria
ingy stevan: niiice
stevan: we played a CPAN drinking game in Taiwan
autrijus won actually 04:41
stevan ingy: yes I have heard,.. although no one has detailed the rules of the game yet
ingy stevan: I'll fill you in when we next meet up
stevan ingy: sounds good 04:42
ingy hmm... Maybe I'll write it up as S27!!
The Bible should definitely cover drinking
did pugs get released this week? 04:43
stevan ingy: not yet,.. we are getting 6.2.0 ready right now
ingy maybe I'll get S27 in on time then
but first I must dance! 04:44
swing dancy dancy &
Ovid stevan: I have HTML::Entities down to an 8 line program that fails to parse properly. 04:49
stevan Ovid: paste it to me
perlbot nopaste
perlbot Paste your code here and #<channel> will be able to view it: sial.org/pbot/<channel>
Ovid Actually, just got it shorter. Hold on.
jabbot pugs - 1838 - Add $string.chop and fix a string delim
pasteling "Ovid" at 66.167.139.35 pasted "minimal test case for HTML::Entities failure" (36 lines, 1.1K) at sial.org/pbot/9253 04:51
stevan Ovid: bare returns like that dont work 04:52
luqui Do we have a place to put perl-in-perl implementations of perl builtins? 04:53
Ovid stevan: doesn't matter. Have 'em return anything, even "return 'foo'" and you'll get the same error. However, if you delete that for loop, it finds the sub.
Something about the for loop is killing it. 04:54
luqui pugs -e 'sub foo () { say "bar" } module Baz; foo()' # prints 'bar'
are namespaces implemented, or are they just a noop? 04:55
stevan Ovid: I got it to compile 04:58
pasting in one second
Ovid Yeah, it compiles fine, but the test fails saying it can't find the subroutine.
OK.
pasteling "stevan" at 67.186.136.119 pasted "HETest working" (43 lines, 798B) at sial.org/pbot/9254 04:59
Ovid stevan, I think I just found the problem. $char_to_entity is never declared, but Pugs is silently ignoring this. 05:02
Er, there's a deeper problem. I through that in my original HTML::Entities and I get the same error :-/ 05:04
teamjet i am really embarassed, looking for spoon-feeding here 05:09
perl 6 out yet?
Ovid No, not yet. We're working on a prototype named pugs.
jabbot pugs - 1839 - some ChangeLog additions
teamjet thanks 05:10
Ovid No problem.
luqui ack, how do I flatten an array in pugs! 05:16
my $a = [1,2,3]; my @b = *@a; say +@b; # 1
er, *$a
Ovid Generally when I get stuck like that I grep through the tests to see if there's something compatible. 05:17
luqui doing that now
jabbot pugs - 1840 - Continuing work on S28draft.pod 05:19
Ovid Woo hoo! I go the the test in HTML::Entities to pass. Finally :) 05:24
stevan ok,.. I am off to sleep 05:25
Ovid Night stevan. Thanks for all of your help. 05:26
stevan autrijus: I updated the changelog, but you might want to add more
Ovid: your welcome ... did the whole thing compile? or just the mini version?
theorbtwo Morning, all.
Ovid I have a much larger mini-version.
stevan autrijus: and any tests which still die I thought you might be able to fix, otherwise just add them to force_todo 05:27
Ovid: excellent
btw - t/force_todo is so much nicer than manual todo/un-todo
morning theorbtwo
theorbtwo Is the plugins detection killing anybody now? 05:28
stevan ok, going for real, g'nite all see you in about 7 hours
theorbtwo g'night, stevan.
gaal morning all. what's that i hear about openfoundry down? 05:29
theorbtwo Mornin, gaal. If you didn't see, BTW, evan linked to your LJ post.
gaal i just made coffee, i didn't see *anything* yet :)
h 05:30
oh, that is
that evan, i like him :) 05:31
his later post in evan_tech, coming from him, sounds like an april fool's joke though :)
theorbtwo Odd, I didn't notice that somehow. 05:33
gaal insufficient coffee maybe?
05:33 castaway_ is now known as castaway
gaal smokes 1840 05:34
theorbtwo Insufficent tea, quite likely.
gaal king crimson must have been lambdafolk. they have that line in their first album where they mention insufficient schemes. 05:35
beh, why is yaml_harness running rules/Disabled? 05:36
Khisanth gaal: smoking 1840 packs of cigarette a day is a bit unhealthy isn't it? :) 05:39
gaal who said it was cigarettes :)
Khisanth smoking anything in that quantity has to be bad 05:40
gaal we've reached the Victorian age if you think of r numbers as years AD. 05:41
Khisanth waits for 3000
theorbtwo In the Year Two Thousand... </voice> 05:46
gaal to2, i'm tweaking testgraph a wee bit, is it okay to always say 'type=text/css' in the LINK element? 05:47
theorbtwo I'm not 100% sure. 05:48
gaal oaky, so i won't.
theorbtwo I didn't originally have it that way; it was suggested as a fix to some problem or another, but I don't remember what, or if it worked. 05:49
I don't really see where it'd be a problem, though.
gaal it was suggested to fix a problem with my webserver. it didn'
t fix it :)
jabbot pugs - 1841 - A bit more work on S28draft.pod
gaal i'm adding an option to embed the css file; the command line remains compatible with old usage. 05:50
dvergin__ Time for bed. S28draft.pod is tucked in for the night and my eyeballs are all a-twiggle. 'Night, all. 05:57
gaal to2, which version of the cataloger is preferred now? jmm or the other one? 06:10
theorbtwo The other one.
jmm no longer exists.
gaal ah, thanks
theorbtwo (Because it got mv'd over the other one.)
gaal (i guess i still have it around because of a leftover hack attempt created a conflict and svn was smart enough not to remove my local copy) 06:11
in that script, do you mean "cd .." instead of "ls .." ? 06:14
and, whoprovides pugs_svn_src? 06:15
theorbtwo wonders what gaal is talking about. 06:17
gaal look at util/catalog_tests.pl:26 06:18
theorbtwo Oh. That's nothingmuch's code; I haven't looked at it much since he added that. 06:19
gaal ah, sorry.
so do i need 'svn co' instead?
theorbtwo I think he means cd .. ; svn co ...
gaal yeah
fixing
Ovid Night all. 06:20
chady looks around and gets worried that he has a lot to learn now to keep up 06:23
luqui & bed 06:25
jabbot pugs - 1842 - * type casting errors are now more descr 06:40
gaal forum2.org/gaal/pugs/tests_winxp_msys.html and forum2.org/gaal/pugs/tests_linux.html - avec test xrefs 06:43
Ovid Does anyone know how to get Unicode characters to show up on the Mac OS X terminal? 06:46
jabbot pugs - 1843 - * downgrade Win32 failures to warnings. 06:49
pugs - 1844 - * =<> now works again (sort of)
pugs - 1845 - * Typeable fixes.
pugs - 1846 - allow embedding CSS (for broken webserve
pugs - 1847 - * "my $x = rhs" should impose rvalue con
autrijus stevan: cool. I'm still working on test auditing and mutrec 06:52
Ovid Does Pugs have issues with printing unicode characters? 07:04
autrijus worksforme.
Ovid Hmm ... I can't get HTML::Entities to quite work but printing debug output seems to be weird. 07:05
mj41 WinXP build failed:
Compiling Compat ( src/Compat.hs, src/Compat.o )
src/Compat.hs:55:13: Not in scope: `trace'
src/Compat.hs:81:0:
Duplicate type signature:
src/Compat.hs:81:0-37: statFileSize :: FilePath -> IO Integer
autrijus where's the code? freepan?
mj41 src/Compat.hs:105:0-37: statFileSize :: FilePath -> IO Integer
Ovid autrijus: was that question directed at me? 07:06
autrijus yeah
mj41: try r1848
Alias_ Ovid: You were mentioning reflection before. I forgot to mention that you may want to take a look at Class::Inspector
mj41 Compact.hs done 07:07
Ovid It's nowhere right now. I am trying to add HTML::Entities to ext/ and I have decode_entities partially working, but encode_entities is giving me fits. I haven't checked anything in because I understood that ext/ was only for working code and I've no idea where to check it in at. 07:08
autrijus ok. check in to ext/HTML-Entities/
and I'll fix it right away, how about that :)
Ovid OK :) 07:09
jabbot pugs - 1848 - * fix Compat.hs on win32
Ovid autrijus: I just checked it in. 07:11
mj41 pugs -e "return 1;"; 07:13
cannot return() outside a subroutine
Val (VList [VInt 1])
Is this ok?
autrijus $ perl -e 'return 1' 07:14
Can't return outside a subroutine at -e line 1.
I guess it's okay.
Ovid: wonderful. let me look 07:15
I'll rename them.
Ovid Thanks. I started working on that so I could check in CGI recipes to the cookbook :)
autrijus to HTML-Entities
Ovid Lots of stuff is commented out while I was trying to get it to work, so it looks sloppy.
Regrettably, I need to hit the sack. Night all! 07:17
mj41 my mistake, same as perl5, but what about
#conf.pl:
return { 1; }
pugs -e "require 'conf.pl'"
perl -e "require 'conf.pl'"
autrijus Ovid: I see why.
you need to use utf8 for your src code.
Ovid Oh?
Er, it wasn't? 07:18
autrijus it was latin1.
Ovid It looked fine to me.
autrijus I fixed it.
r1852
Ovid Hmm ... how do I tell?
Aankhen`` Night Ovid.
autrijus depends on what your editor is.
Aankhen`` (whenever you pop off)
autrijus for vim, "set fenc=utf8"
Ovid Thanks.
I'm using vim 7.0 (alpha) for Mac OS X. 07:19
Aankhen`` vim on Mac? Oh, the irony... ;-)
jabbot pugs - 1849 - First pass at horribly crippled HTML::En
pugs - 1850 - * rename directories
pugs - 1851 - * change File-Spec name change
pugs - 1852 - * use uutf8 literals.
Ovid Yeah, I've heard it before :)
OK, now I'm going to bed and I really mean it! 07:20
chady errm.. no ghc-6.4 for debian?
autrijus chady: look at experimental 07:21
jabbot pugs - 1853 - * and tests should use utf8 too 07:29
chady autrijus: thanx 07:30
mj41 damn it 07:32
perl -e "my $c={ a=> 1}; print (keys %$c)" # a
pugs -e "my $c=( a=> 1); say $c.keys" # a
pugs -e "my $c={ a=> 1}; say $c.keys" # 0
autrijus mj41: fixing.
ingy hola
autrijus hey ingy 07:33
mj41 autrijus: thanks
ingy hi autrijus
Aankhen`` ōæ½xABautrijusōæ½xBB: The Perl 5 HTML::Entities appears to be encoded in Latin-1. 07:37
autrijus Aankhen``: Perl 6 apparently is unicode baesd. 07:39
based, even.
it may be a day that we guess latin1.
but I do not want it to be the default.
Aankhen`` Yeh, I was just pointing it out in case there were any problems stemming from that fact... 07:40
autrijus ok :)
Aankhen`` Do multi subs work?
autrijus they should/
they should.
Aankhen`` Cool.
encode_entities seems to be a good candidate. 07:41
Juerd autrijus: Didn't Larry say latin1 bytes work as their utf8 characters? 07:54
(I wonder *how*, though)
ingy autrijus: where is the long module list in cpan? 07:57
rgs ingy: www.cpan.org/modules/02packages.details.txt.gz ? 07:58
ingy tanks
Aankhen`` Eek. GHC is 47 MB? 08:01
Aankhen`` watches his connection choke.
ingy Perl6 got a BIG motor
Aankhen`` Heh. 08:02
autrijus Juerd: larry may have said that. I don't see a way to do that cleanly. 08:03
so I punted.
Aankhen`` ōæ½xABautrijusōæ½xBB: Are you still working on HTML::Entities? 08:06
autrijus Aankhen``: yes, about to commit now 08:07
Aankhen`` Alright.
Aankhen`` waits with bated breath.
rindolf Hi all! 08:08
autrijus hi rindolf! 08:09
Aankhen`` Hiya.
rindolf Hi autrijus
Hi Aankhen``
Aankhen`` What's shakin'?
rindolf Aankhen``: I'm fine. I finished writing an article for Perl.com.
Aankhen`` Cool.
Link?
autrijus working HTML::Entities landed. 08:10
Aankhen`` Or is it yet to be submitted?
Khisanth can the various unicode files be easily distinguished from each other?
autrijus enjoy
rindolf Aankhen``: I can /msg you a link.
autrijus r1854
Aankhen`` Sure, that'd be great.
autrijus rindolf: /metoo
rindolf autrijus: OK
Aankhen`` Yey for autrijus!
autrijus Khisanth: utf8 from utf16 etc, sure
Khisanth: but a file can be both valid utf8 and valid latin1.
and I don't want to trigger reparse.
so that's a problem.
if you have thunk of a clever algorithm, I'm all ears 08:11
Alias_ with or without prescanning? :)
Khisanth what about endian? (I am was just looking at the menu in Mozilla)
autrijus reparse = 2pass ;) 08:12
Khisanth: endian ha sBOM
has BOM
so that's clear
gaal a file is valid utf8 and valid latin1 iff it is ascii, no?
autrijus no.
any utf8 file is also valid latin1 file. 08:13
actually, any file whatsoever is a valid latin1 file.
gaal ah, yes of course
silly me. only it's nonsense latin1, likely
autrijus aye.
problem being, one cannot easily tell. 08:14
gaal all you need it a test for nonsesicality then. though you still need anoth pass, to tell whether the utf8 interpretation yields nonsense too, for example if the author was being silly. 08:15
Khisanth perhaps larry meant something else?
autrijus gaal: 2pass is not acceptable for a streaming parser.
I can lookahead finitely, but I don't want to do infinite lookahead.
Khisanth: perhaps. there's no way to tell except ask :) 08:16
gaal autrijus: neither is an AI-complete dependency :)
(i was saying that tongue in cheek :)
Alias_ autrijus: Can you reparse at all, or is it just expensive? 08:17
My "is this a binary" stuff generally just pulls the first chunk ( 1/2/4k of the file and checks that )
gaal is latin-1 some sort of preferred 8-bit encoding in perl6? 08:18
autrijus Alias_: that means a 4k lookahead up front
that is doable.
it is perhaps the most sensible thing to do.
Alias_ autrijus: Which if it matches the page size is very fast
As long as you _can_ reparse if you have to...? 08:19
But I'd imagine that BEGIN would make that very hard
autrijus yeah, especially BEGIN that affects encoding.
use encoding()
rindolf gaal: hi! It's me, Shlomi.
gaal: what's up?
jabbot pugs - 1854 - * require "" now works on utf8 files.
autrijus gaal: I don't know. I hope not. I can cope if it is 08:20
Alias_ But really, can you ever actually know?
nothingmuch small poll: who is the youngest person here? some friend is asking
gaal hi Shlomi
nothingmuch he's assuming me, i said lightstep
rindolf nothingmuch: I'm 27.
Alias_ how old?
autrijus definitely not me. I'm 23, I'm old ;)
rindolf Hi nothingmuch
nothingmuch hola
rindolf nothingmuch: what's up?
theorbtwo is back. 08:21
autrijus hey theorbtwo.
rindolf nothingmuch: how old are you?
nothingmuch 19
autrijus I'm currently fixing the infinite loop in array_ref
nothingmuch lighstep is 18
not much
theorbtwo is 24.
nothingmuch or maybe even 17
gaal is 64
autrijus mmm base 7? 08:22
gaal j/k
rindolf nothingmuch: I talked with a 16-year old Perl programmer yesterday
nothingmuch rindolf: the person is looking for someone interested in p6
rindolf nothingmuch: I see.
nothingmuch if that's relevant, then i shoot
autrijus next week I'll turn 24, which means I would have programmed perl for 10 years.
nothingmuch it's for some interview
rindolf nothingmuch: are you exempt from service? Or are you an Atuday?
nothingmuch exempt is true 08:23
rindolf autrijus: I've been programming Perl since 1996
nothingmuch feels younger all of a sudden
rindolf nothingmuch: do you know your way around Module::Build?
nothingmuch rindolf: i've hacked something into it once
and i use it
but i don't know it per se
theorbtwo feels much older. 08:25
rindolf nothingmuch: do you know how I can install a custom file under /usr/share/xml/ ?
nothingmuch nope... usually i resort to doing that in __DATA__
this is because /usr/share might not be /usr/share
rindolf nothingmuch: well, something along these lines. 08:26
nothingmuch i'd put it under the same libpath you put your .pm in
for consistency
rindolf OK.
nothingmuch because the file can be looked up in perl
and because you don't have to learn anything new =) 08:27
rindolf Of course, I'm considering changing my strategy and writing a module with a function that will return the content I want to put it in the file.
nothingmuch that's normally a better approach
even if the function is just a thin wrapper for a file
rindolf nothingmuch: OK. 08:28
castaway hey there young people :) 08:36
08:36 scw is now known as scv
nothingmuch hola, old woman! 08:36
08:36 scv is now known as scw
castaway you young whippersnapper you :) 08:36
castaway wonders if shes the oldest here ,) 08:37
rindolf castaway: how old are you? (if I may be prudent enough to ask)
castaway 31 ,)
(ancient ,)
not that I feel this old.. 08:38
rindolf Is there anyone here older than 31?
nothingmuch our search is in the other direction
castaway suspects chip may be..
theorbtwo is Sparticus! 08:39
rindolf Chip Salzberg is here?
nothingmuch yes, he is the parrot lead architecht
tect
castaway occasionally
hmm, Chip: Back in 1988, I wanted to use Perl 2.0 on a .. 08:43
rgs is 30 08:44
since several years
rindolf castaway: perl 2.0?
nothingmuch damn, i was 3
rindolf castaway: oh right Nineteen-_Eighty_-Eight.
I was 11 08:45
castaway right
theorbtwo was 8.
castaway was umm.. 15
nothingmuch oh crap, you had to bee a teenager in the 80s?
i'm so sorry!
castaway so I'm betting he's got a few years on me
no reason to be, they were good :)
rindolf castaway: what did you want to run Perl on? 08:46
castaway No, that was Chip, not me, I was quoting
rindolf castaway: I see.
castaway but for the record, it says: Xenix/286 08:47
www.foo.be/docs/tpj/issues/vol2_4/t...-0004.html
rindolf Xenix... %-)
Xenis was a single-user UNIX.
castaway SEE still running, nm? 08:51
castaway attacked it yesterday
nothingmuch should be
i can access it
castaway nifty 08:52
its getting as far as sending me doc lists.. just need to parse them :)
nothingmuch all from seemacs? 08:53
theorbtwo She's dealing with the lunch-guy now, I think. 08:56
Sounds like all from seemacs.
castaway yes (back) 08:58
(from 5 floors down)
autrijus $a = 5; *$a
is that an error or just (5) ?
castaway yes from seemacs.. ie it takes my connection, thinks Im another SEE instance, and offers me a doclist.. :)
rindolf What's seemacs? 09:00
theorbtwo emacswiki.org/wiki/SubEthaEmacs
castaway An attempt to implement SubEthaEdit type multi user editing, into emacs 09:01
theorbtwo Not just SEE-type, but wire-compatible with SEE. 09:02
castaway umm, right
castaway should update that page 09:03
autrijus infinite cyclic structures implemented. 09:07
nothingmuch autrijus++
rindolf perlbot top 20 karma
perlbot The top 20 karma entries: C: 77, autrijus: 71, nothingmuch: 48, stevan: 46, ~brad2901: 29, theorbtwo: 28, ingy: 28, gaal: 21, $i: 20, iblech: 16, beth: 16, Darren_Duncan: 14, jagerman: 14, ~wegstar: 13, (: 13, shapr: 12, Yaakov: 12, Somni: 12, Corion: 11, Juerd: 10
autrijus my $a = [0]; $a[0] = $a;
nothingmuch $i =) 09:08
nothingmuch whines about redhat and minimizes IRC again 09:09
Khisanth wonders what is (:++
castaway an odd smiley?
Khisanth or wait it's (++
theorbtwo Hm, trying to say $a, say $a.perl, or say +$a hangs.
Khisanth .perl? 09:10
theorbtwo Think Data::Dumper.
Khisanth oh 09:11
theorbtwo (I'm not sure if the first two hanging is OK, but the last should just return 1, no?)
shapr whoa, I got 12 09:12
shapr boings happily
theorbtwo Allo, shapr. 09:13
castaway ruffles shapr 09:14
shapr allo!
shapr looks like a nice set of curtains with ruffles
castaway *g*
hey there bouncy
castaway dubs shapr "tigger"
shapr heh 09:15
gaal goes "oo" 09:17
jabbot pugs - 1855 - * saner context casting; infinite cyclic 09:19
autrijus my $array_ref7 = [ $array_ref1[[2, 1, 0]] ]; 09:20
this tests looks very wrong...
kungfuftr moo
shapr boings 09:21
castaway woof!
autrijus oh well. I'll change it to 09:24
my $array_ref7 = [ $array_ref1[*[2, 1, 0]] ];
jabbot pugs - 1856 - * more array_ref work; all tests now pas 09:29
ingy autrijus: just commited S27 09:30
autrijus ingy: woot
ingy looks forward to playing a few rounds at YAPC 09:33
castaway playing? 09:34
ingy castaway: read S27 09:35
castaway tries to absorb it my osmosis :)
autrijus hmm
my @lines7 = readline($in7);
my @line8 = 1..Inf;
ingy svn.openfoundry.org/pugs/docs/S27draft.pod
autrijus somehow some people want the first to be strict
and the second to be lazy
talk about conflicting expectations. 09:36
castaway :)
theorbtwo notes that a circle is not a polygon.
autrijus it is a infinygon.
theorbtwo I thought it was a unigon.
nothingmuch autrijus: i think this can be done by doing laziness as a trait
it's a trait of the value 09:37
that affects how it's containers behaves
autrijus I'll just mandate readline() is strict in list context for now.
nothingmuch rvalues that occur from readline(...) etc, will be eager
or strict (i think that's a conflicting term)
autrijus eager == strict in here 09:38
nothingmuch i.e., be eager where we suspect it should be done
yes
you are right
be lazy only when we're sure it's safe
and how are we sure it's safe? we know some primitives, like ... Inf are pure
or some explicit constructs are lazy
strictness is a contageous trait
that affects everything it touches 09:39
including delazifying things that are lazy, but are interacting with strict values now
jabbot pugs - 1857 - S27 - Beer Drinking Hackers Must Read!
pugs - 1858 - * add pair.keys and pair.values back.
castaway ah, perl culture, of course 09:40
hmm, poor ole non-drinking hackers :) 09:42
theorbtwo You can be the offical.
rindolf ingy: hi.
castaway hmm, ta :)
as long as I get tea :)
theorbtwo And a copy of 02modules! 09:43
castaway of course
elmex wtf is S27 ??
i hate beer
castaway the Perl Culture chapter
elmex i only drink whiskey
castaway sorry, it only says beer
ingy elmex: add your own game :P 09:44
elmex s/beer/alcohol/ s/alcohol/drugs/ s/drugs/drugs&sex/
Khisanth caffeinated drinking game should be more productive ;)
ingy but less fun
hi rindolf 09:45
kungfuftr vodka+redbull!
nothingmuch redbull?! *choking sounds*
too sweet!
ingy tequilla+vodka+redbull!!
TVR baby!
nothingmuch jagermeister (2-3 shots), then beer
(as much as it takes to stay happy but not completely drunk) 09:46
ingy nothingmuch: TVR is a traditional JAPH drink
nothingmuch (normally 3-5 1/2 liters an evening)
elmex scottisch whiskey!
castaway Scotch
elmex oh, thanks ;)
nothingmuch has never tried whiskey
elmex scotch whiskey 09:47
nothingmuch: my fav. is glenfiddich ;)
nothingmuch perlmonks.org/?node=edan
there's a list there, maybe you count
rindolf ingy: what's up? What is keeping you busy nowadays? 09:48
jabbot pugs - 1859 - * readline() is now strict in list conte 09:49
ingy rindolf: Perl 09:53
rindolf ingy: what in Perl? 09:54
ingy Kwiki, YAML, Perldoc, Kwid, FreePAN, Pugs to name a few 09:55
you know, top level namespaces stuff ;)
(except Pugs) 09:56
rindolf Heh.
Can FreePAN install Python modules?
ingy wiki.freepan.org/index.cgi?FAQ 09:57
<-- bedtime
rindolf Oh the horror! search.cpan.org/~ingy/Kwid-0.00_01/
ingy: aren't you British?
ingy no
rindolf ingy: I see.
ingy: good night.
ingy nighty night 09:58
kungfuftr anyone got much experience with perl and shift_jis encodings?
(OT i know)
rindolf kungfuftr: don't look at me.
rindolf is innocent (really) 09:59
autark-jp kungfuftr: only experience I have is converting it to something sane ;) 10:01
kungfuftr autark-jp: how bout converting the MS version of shift_jis to utf-8 and back again? 10:03
Aankhen`` Is builtins/undef supposed to fail a few tests?
autrijus I'm going thru all tests now
please bear with me
autark-jp don't know if I have touched MS version, but sjis->utf8->sjis, yes.
Aankhen`` OK.
rindolf Aankhen``: welcome back 10:04
Aankhen`` Just compiling Pugs for the first time, and the tests are taking longer than the compilation. :-P
Thanks, rindolf. :-)
autrijus that's usual. we have lots and lots of tests
Aankhen`` Yeah, I figured.
Anyway.
I'll wait till you're done with going through the tests. 10:05
ingy autrijus: who played CPAN Drinking besides acme, autrijus, clkao, gugod, ingy, mugwump and obra?
autrijus yeah
ingy was hcchien there?
autrijus no 10:06
ingy whiteg was there but did not play
nothingmuch perl golf should be mentioned, it being the basis for pugs development driving
ingy nothingmuch: patches welcome!
rindolf ingy: what's CPAN Drinking?
autrijus s/patches/commits/
nothingmuch thinks outloud more than he should
ingy rindolf: read S27 10:07
theorbtwo asks questions less then he should.
nothingmuch rindolf: see s27 in docs
autrijus rindolf: built pugs yet? :)
rindolf There's a typo there. 10:08
autrijus: no, not yet.
EQUIPTMENT. 10:09
twice
ingy rindolf: fix it please
I must sleeeeep
theorbtwo G'night, Ingy.
jabbot pugs - 1860 - * join () now passes all tests; ().join(
pugs - 1861 - Added history of CPAN drinking game
ingy night theorbtwo & 10:10
autrijus ingy: fixed.
rindolf ingy: night 10:11
theorbtwo Mornin, meta, cwest. 10:15
autrijus hey.
kungfuftr reruns smoker 10:16
jabbot pugs - 1862 - * fix typo, spotted by rindolf. 10:19
rindolf pets jabbot 10:20
gaal who was the official for the game?
autrijus whiteg I think 10:21
nothingmuch jabbot: seen stevan
jabbot nothingmuch: stevan was seen on Tue Apr 12 05:28:46 2005 GMT
nothingmuch goody goody gumdrops
i'm forking off test graph
i have some ideas:
or maybe it doesn't, maybe it just builds a structure 10:22
yaml harness should be TAP::YAML, or something along those lines
the actual harness should use it
it is an object that takes straps and builds yml
and the yaml harness uses YAML and that to make a yml
TAP::Visualize::$foo will be testgraph.pl 10:23
what is $foo? i don't know what kind of graph that is
theorbtwo TAP::Visualize::HTML ?
nothingmuch there are many ways to use HTML to visualize it
theorbtwo The name wasn't meant to invoke "graph" so much as "graphical".
nothingmuch i'm not sure that is accurate enough
i'll make a Tap::Visualize::$foo::Consolidating, which takes several YAMLs 10:24
since consolidation is basically a display issue
but it will use TAP::Consolidator, i think
we could say it's Tap::Visualize::HTMLMatrix?
i think that gives a better idea of what it does 10:25
theorbtwo Works for me.
nothingmuch okay
so what should the TAP::DOM thing name be? 10:26
TAP::Structure? that is a bit vague? 10:27
castaway is still puzzling what "TAP" might be :) 10:28
nothingmuch search.cpan.org/perldoc?Test::Harness::TAP
theorbtwo Test somethingoranother Protocol.
nothingmuch test anything protocol ;-)
theorbtwo Perl's interface between testing modules like Test::More and the test harness Test::Harness is a simple text-based format called TAP, the Test Anything Protocol. This is its story.
castaway aha 10:29
jabbot pugs - 1863 - * reverse.t now all passes.
theorbtwo The modules should probably be under the Test::* namespace, rather then TAP::*. 10:32
nothingmuch TAP::Model is the best name I can think of
i don't want to imply they're used for testing...
maybe Test::TAP::*?
theorbtwo Why don't you want to imply that? 10:33
nothingmuch because they aren't test modules per se
they are reporting tools
customarily Test::* gives you a set of assertions, or whatever
it's very useful that you can get a dense result list
and normally know exactly what a module does in the Test:: namespace
i feel a bit ooky about polluting test results 10:34
since I know it will bug me a little
the Test::Harness mess (one for every dist that bundles it) is annoying enough already
but either way, Test::TAP is good enough for me
Aankhen`` BBL. 10:35
nothingmuch because you learn what TAP is once, and then you can filter quickly
ciao Aankhen``
rindolf Aankhen``: bye 10:37
nothingmuch hola lightstep 10:38
lightstep is it just me, or does t/data_types/hash_ref.t contain an infinite loop?
rindolf lightstep: hi
lightstep y0
pugs -e 'my $a; $a<a>=1' <-- loops 10:39
nothingmuch then that's a bug 10:40
ping autrijus
lightstep -> bong 10:47
nothingmuch sshhh!
you might be bugged
with all this anarchy going about the feds are probably monitoring #perl6
autrijus whew. 10:56
just implemented mergesort.
@a.sort:{$^b <=> $^a} works
as doea
sort {$^b <=> $^a} @a
nothingmuch yay!
this could be lazified 10:57
autrijus provided you know that the closure does not cause side efects.
maybe you want the side effects.
nothingmuch as a heapsort
i thought we don't care about all that stuff in p6
=P
autrijus ;)
you wish ;) 10:58
nothingmuch no i wish otherwise
but thats what i sort of seem to observe
the general mantra is that somehow things will work out 10:59
jabbot pugs - 1864 - * sort {} now works.
castaway "things"? 11:00
castaway decides its lunchtime
pjcj autrijus: you want to specify when and how often and with which parameters a sort sub will be called? 11:03
autrijus not really.
nothingmuch autrijus: what is your opinion on laziness?
autrijus I think laziness is a good thing in general. :) 11:04
is this vague enough?
pjcj ah - good. It sounded as though you did.
xern damn, my remote machine is down..... cannot write code now....
nothingmuch i love laziness
pjcj laziness is still a virtue, even in Perl6, no?
nothingmuch but i don't want to give up robustness
i prefer being explicitly lazy
Aankhen`` Is t/data_types/array_ref supposed to take ages? 11:11
It seems to be stuck at "ok 35/41". 11:12
autrijus hm? it worksforme
Aankhen`` Strange. 11:13
I'll try again in a few minutes.
rindolf nothingmuch: Ran Eilam demands his book back. 11:17
nothingmuch ran eilam can give me his mailing address like he promised 3-4 times
and then i will gladly mail it 11:18
but i'm not driving to TA on anti-isthemines
rindolf nothingmuch: are anti-isthemines a kind of drugs?
nothingmuch yes
they repress a part of the immune system
castaway histamines, no?
(for hayfever)
nothingmuch could be, i have no clue, castaway
typically allergic people produce histamines for the wrong reasons 11:19
rindolf nothingmuch: you can take a bus.
jabbot pugs - 1865 - * empty scalar autovivification.
nothingmuch rindolf: no, i can't, because if i get back from work, and then go, by the time i'm done i have no bus home
please don't patronize me, i can think of this stuff myself 11:20
rindolf nothingmuch: I see.
nothingmuch anyway, i have dermatitis, which is like allergies in that histamines are produced wrongly, but the reasons for this symptom are different from allergies
Aankhen`` t/builtins/string/length went mad...
rindolf has prepared rice with lentils 11:21
Aankhen`` "Failed 36/15 tests, -140.00% okay"
nothingmuch Aankhen``: Test::Harness is funny that way
if the plan doesn't match reality in some way
Aankhen`` OK. 11:22
nothingmuch run pugs t/builtins/string/length
.t
and fix the plan stuff
and also make a bug report to Test::Harness about that
Aankhen`` t/data_types/array_ref is stuck on "ok 35/41" again.
Aankhen`` salutes! 11:23
SIR YES SIR!
nothingmuch double time
autrijus it doesn't for me :-(
nothingmuch ;-)
Aankhen`` Hrm.
I'll fix the plan thingie first.
autrijus did you "make optimized"?
cool
nothingmuch believes in god, and believes that god believes in Aankhen``
Aankhen`` nmake, actually; Win32, VC++. 11:24
No optimized though.
nothingmuch thinks his GC zone ideas are materializing very well, but doesn't have the time to impl 11:25
jabbot pugs - 1866 - * hash() and pair() works as described i 11:29
Aankhen`` Whaaa? 11:31
It's running 51 tests...
Hrm.
It's running each test three times.
Inside the loop, atleast.
theorbtwo wonders why pugs -e 'require_haskell("/usr/src/SHA1/blib/arch/SHA1__0_0_1.o"); say &sha1' still doesn't work. 11:32
nothingmuch gaal++
castaway what does it do? 11:33
theorbtwo Undeclared variable &sha1
Val VUndef
castaway hmm, how does p6 know thats supposed to be a func call in SHA1*.o ? 11:34
theorbtwo Because SHA1__0_0_1.o tells it so. 11:35
castaway automagically adds to the namespace?
theorbtwo Supposed to. 11:36
castaway can you check the symbol table or whatever, and see if its there? 11:37
can someone give me a url to look directly at the files in the pugs svn?
theorbtwo svn.openfoundry.org/pugs/head/
castaway thanks 11:38
Aankhen`` Um... how do you de-reference an array reference?
castaway hmm, head/ didnt work, but the rest was good :) 11:39
Aankhen`` Nevermind.
castaway looks at doc/class ducktyping stuff, but cant understand it
jabbot pugs - 1867 - * infinite cyclic hashes.
kcwu pugs> ?1 11:42
Segmentation fault (core dumped)
Aankhen`` I give up. 11:43
How do you de-reference an array ref?
theorbtwo The cabal broke @$ar?
Aankhen`` Doesn't seem to work. 11:44
autrijus list $ar 11:45
Aankhen`` Ah.
Thanks.
autrijus @{$ar} probably should work too
write TODO tests. 11:46
Aankhen`` Alright.
Does next() work?
autrijus no.
kcwu it seems that, any line begin with '?' will make pugs crash. "? <exp>" doesn't work at all
Aankhen`` Darn. 11:47
autrijus kcwu: right. had not have time to fix
Aankhen`` There's something strange about t/builtins/strings/length.t
Hrm, wait. 11:48
autrijus but, dinner first. &
castaway wonders why P6::Bible requires Spiffy
Aankhen`` Enjoy your meal.
theorbtwo Bon appitite, autrijus. 11:49
nothingmuch castaway: most of ingy's stuff does
theorbtwo BBIAB, have to reboot.
castaway aeh, but I thunk P6 bible is just a bunch of pod files.. 11:50
(and is it supposed to give errors parsing the pod files? :)
11:51 chady is now known as megalomania
Aankhen`` Eek. 11:52
This array is being flattened. O_O
Er.
Sorry.
Since the tests for multidimensional arrays are all TODO'd in data_types/array.t, I guess they're not supported. 11:55
WB. 11:57
castaway snuggles theorbtwo.
theorbtwo thanks Aankhen, and kisses castaway.
Aankhen`` The more I look at length.t, the stranger it seems... 11:59
gaal and now it only has the *easy* cases, Aanken`` :)
s/k/kh/
jabbot pugs - 1868 - * hash.t all passes.
pugs - 1869 - * repair junction.pick
Aankhen`` Heh.
That's nice. ^_^
Welp, atleast now it's clear why it's running 36 tests. 12:00
Sorta clear, at any rate
.
jabbot pugs - 1870 - fix indentation 12:09
nothingmuch wants proper delegation already! 12:18
Class::Delegation doesn't really cut it
rindolf nothingmuch: what does delegation mean? 12:20
nothingmuch sub new { ....; $self->{some_slot} = $object }
and then you somehow declare that $object handles methods xyz for you
rindolf nothingmuch: can't AUTOLOAD or whatever do that? 12:21
nothingmuch and you don't have to wrap it in sub method { $self->{some_slot}->method(@_) }
or use scary AUTOLOAD tricks
AUTOLOAD only goes so far
Class::Delegation uses a very scary variation on the AUTOLOAD theme
iblech nothingmuch: has $.some_slot handles <method1 method2> 12:23
nothingmuch in p5, not p6
if i didn't know how lovely delegation in p6 will be, i wouldn't be complaining ;-) 12:24
iblech :)
pdcawley Okay, how do I tell ghc where to find hs-plugins? 12:25
rindolf nothingmuch: but you can manufacture methods to do it using eval,
nothingmuch rindolf: no thank you
ninereasons make install && make register, pdcawley
pdcawley Ah...
nothingmuch rindolf: i code to promote readability 1st
i've done this stuff, and i do it for fun still
pdcawley That'll teach me not to read the whole README 12:26
nothingmuch but this code is code for others to read and see, and i'd rather type a bit more, than to confuse someone whose reading it
i can do it with regular @ISA inheritence, and it's not that aweful
obra nothingmuch++
nothingmuch it just adds a bit of interface cruft
rindolf nothingmuch: I meant that write a method factory and then use it like (delegate([qw(method1 method2 method3)], 'myslot'))
nothingmuch this is not very reliable, it suffers from the infamous INIT/BEGIN time mess 12:27
rindolf: again, that's what Class::Delegation basically cooks up for you, it doesn't matter how it's implemented
(not friendly to mod_perl users)
ninereasons gaal, I hope that I didn't mess up length.t too much. If I did, would you tell me what I did wrong?
nothingmuch as this module will most likely be used for generating reports, i don't want anyone sufferring
pdcawley really, really wishes the INIT/BEGIN thing could be fixed up in Perl 5. 12:28
There's so much cool stuff you can do when you use that stuff, and then it all falls to pieces when you try to use mod_perl
nothingmuch Export::Simple for example 12:29
that's sooooo beautiful
jabbot pugs - 1871 - length.t add to force_todo
pdcawley Almost anything that uses attributes falls apart in mod_perl if memory serves. 12:30
obra agh! piers! you said that hateful word
pdcawley obra: What hateful word?
obra
nothingmuch funny thing is that attributes usually help organize wide apps like web ones very well
obra is a bit bent out of shape about the Apache2 renaming
pdcawley nothingmuch: Indeed.
obra "Here! well into RC land, we're going to break all your apache2 code" 12:31
pdcawley Imagine being able to write 'sub whatever :handles('/this/path') {...}'
obra and people have wondered why I didn't want to port RT to mp2 until it was released
pdcawley is so glad he isn't writing web apps any more.
nothingmuch is glad he never really started
pdcawley notes that Rails is tempting him to start again, but this time in Ruby. 12:32
The P6Rails is going to kick *much* ass.
nothingmuch yeah 12:33
with WALK, and hyper dispatch
and junctions
and all that beautiful stuff that really helps to simplify spaghetti logic
and, err, proper OOP support
duh
obra has been luring autrijus toward writing webapp framework
pdcawley was thinking more along the lines of all the lovely domain specific metaprogramming that Rails has.
gaal back 12:34
pdcawley but a big 'yeah' to the other stuff...
gaal ninereasons, what did you think you messed up? :)
nothingmuch pdcawley: example? i don't know rails
i barely know of it
ninereasons I was confused by an edit conflict, for one thing.
nothingmuch ninereasons: 3-way-diff is your best friend 12:35
ninereasons I checked in a G file - bad
gaal where, length.t?
nothingmuch too bad vimdiff doesn't know how to do it
pdcawley When you're defining your class attributes, for instance, you can declare some fields with 'has :whatever', 'belongs_to :whatever_else' and it uses this info to work out the database table structure it's working with.
ninereasons fixed now, though.
(well, I think so, anyway)
nothingmuch pdcawley: ah, pretty
gaal sorry, i was eating and i seem to lack some context
pdcawley And, while it does it, it sets up appropriate accessor methods. 12:36
nothingmuch pdcawley: i have a pet project that's sort of stalled because real work isn't pushing it forward, that does very funky stuff with Class::DBI schemas, using SQL::Fairy
ninereasons sorry, I didn't get enough sleep and probably didn't proved intelligible context, gaal
nothingmuch err, sqlfairy, that is SQL::Translator
ninereasons *provide
pdcawley Oh, it's all stuff that can be done in Perl, but it's so well thought through in Rails. 12:37
ninereasons svn up to get length.t and forece_todo and see what you think about the alterations I made.
nothingmuch will take a look at it some day, if he ever gets to needing to do web development
gaal no worries :)
pdcawley And because you don't have to roll all your own 'internals of the object' cruft for every method, you can just get on with writing the app.
obra yeah. rails is _pretty-
and the code is _pretty_
pdcawley obra: It's quality all the way down.
ninereasons gaal, ok :) 12:38
kungfuftr pdcawley: how's the pgce going?
pdcawley Ruby has beautifully 'light' OO syntax; there's very little scaffolding so every bit of code you read is relevant.
kungfuftr: Dropped out.
kungfuftr pdcawley: heh, perl addict! 12:39
pdcawley Not that. Just a hellish workload.
jabbot pugs - 1872 - resolve conflicting edit
nothingmuch how does ruby support YAML? 12:40
pdcawley notes that Ruby is one of the things that really excites him about Perl 6's OO facilities.
nothingmuch: Very well, thank you.
nothingmuch goody
nothingmuch might be tempted to really try and use it when it's time
pdcawley The testing fixtures in Rails are specced using YAML.
nothingmuch but again, management will have to unstick that project
pasteling "gaal" at 192.115.25.249 pasted "msys build fail (latest r)" (5 lines, 192B) at sial.org/pbot/9259 12:42
gaal who broke the build? :) 12:43
kungfuftr gaal: you need to make clean
gaal (okay on linux, where i have hs-plugin fwiw)
but i did :) 12:44
kungfuftr hhhmmm... odd
kungfuftr waits for his testrun to finish
pdcawley Tests are failing here 'Cannot cast from VRef <Scalar> to [Char]' looks common...
OSX + hs-plugin
ninereasons don't build 'make optimized' pdcawley, that might solve the problem.
gaal oh: on linux, src/Eval.hs:286:34: Empty 'do' construct 12:45
pdcawley just did 'make; make test'
no make optimized in sight.
ninereasons hm. i'm not helping then :( 12:46
gaal oh: on linux i forgot to make clean
so ignore that complaint
but msys is certainly b0rken.
kungfuftr latest smoke build on my boxen: kungfuftr.com/pugs-smoke.html 12:47
ninereasons look at all that perty green :) 12:48
gaal i just noticed something i don't like about the testgraph: you can't copy the test description. 12:50
i wonder if there's a way to avoid that w/o javascript
kungfuftr not really i'm afraid 12:51
gaal oh, crap: the Eval error still happens for me in latest r
on linux
whoops, sorry, no i had a stale file again. careless. :( 12:52
kungfuftr *slaps* gaal
gaal ouch!
kungfuftr reruns test builder 12:54
ninereasons there's no way for pugs-interactive to wait on the following line for the completion of an expression, is there? I've tried the obvious '\' and that doesn't work
stevan make optimized just worked for me 12:55
(so far)
ninereasons wow.
stevan hola nothingmuch
the test are running (with some errors though) 12:56
kungfuftr junctions seem to be pretty borked
stevan kungfuftr: really??
kungfuftr stevan: yeah, 18/30 failed... 7/23 ran, etc. 12:57
nothingmuch hola stevan
which do yo prefer, stevan, HTML::Template, or Petal?
stevan nothingmuch: stevan mostly, I dont respond when people call me Petal :P 12:58
nothingmuch: I have never used Petal
nothingmuch stevan++
stevan HTML::Template used to be my favorite
nothingmuch exactly
it's nice, but no one uses it
stevan but now I am addicted to Template Toolkit
nothingmuch IMHO we don't need a "real" templating thing, and i'd rather have one that everyone can read easily 12:59
stevan I like being able to do [% object.method() %]
nothingmuch i think HTML::Template and Petal both satisfy that requirement
in petal that is also allowed, which is why i prefer it to HTML::Template
stevan I would go with HTML::Template it is more common
nothingmuch in general it is more forgiving of OOP design
okie dokies
stevan and still very usable (just more work on the perl side)
this is for the Test::TAP::Madness::* module? 13:00
castaway Text::Template++ :)
nothingmuch Test::TAP::HTMLMatrix will be a gutted testgraph.pl
stevan just finished his first cup of coffee and the backlog :)
nothingmuch insert ::Visualize in there
shapr writes Madness::Cthulhu
stevan stevan queues up Madness::One::Step::Beyond 13:01
shapr grins
kungfuftr nothingmuch: TT++
nothingmuch Object::Meta::Plugin
stevan nothingmuch: in answer to your age query,.. I am 31 (soon to be 32)
nothingmuch kungfuftr: yeah, i rather like it too, but it has a 30 minute learning curve. i want a 5 minute one.
stevan nothingmuch: I dont know if I agree witht hat
I saw TT and it made immediate sense 13:02
castaway stevan, how soon? :)
stevan (on the template side that is)
castaway: Aug 12th
kungfuftr you may not need those stupidly advanced things now, but further down the line you will and when that happends it's a huge extra cost to have to either hack it into Petal (etc) or move everything to TT
castaway damn :)
stevan hopes he is the oldest
castaway nope, missed by a month
ninereasons pdcawley reported lots of tests failing without 'make optimized', stevan ..
nothingmuch splash? wtf? MOMMMY!)
ninereasons pdcawley Tests are failing here 'Cannot cast from VRef <Scalar> to [Char]' looks common...
nothingmuch it's also a beast to install
(what?! XS? DBI? /usr/share/WHAT?!
shapr oldest what?
nothingmuch #perl6 lurker
stevan ninereasons: I got that too 13:03
shapr I'm 33
1971
castaway blinks
stevan DOH!!! the geezer from sweden beat us
shapr grins
castaway darn :)
Steve_p Geezer???
stevan castaway: it does explain why he is so wise
shapr haha
castaway stevan umm.. :)
but not the bouncy bits :) 13:04
stevan Steve_p: can you beat shapr's 33?
ninereasons is childlike in perl; at 48
shapr w00
castaway ooh
Steve_p Actually, no, I can't
stevan 48, 48,... so I hear a 50?
stevan sometimes feels like his is 50 after chasing the kids all day
Steve_p 1971 is the perfect year
shapr Steve_p: which month? 13:05
stevan Steve_p: I prefer 73 myself :)
kungfuftr 83!
Steve_p Wait a second, yes I can beat 33
(old age) -- # ;)
shapr, January
shapr Ah yes, I'm September.
stevan kungfuftr: I remember '83,.. those were the days,..
castaway September++
stevan got my first pair of Fat Laces 13:06
castaway right, stevan :)
shapr I got on the internet in the 80s
nothingmuch 's parents didn't even meet in 83
stevan puts on Melli-Mels "The Message"
shapr: how could you? Al Gore didnt invent it until the early 90s 13:07
nothingmuch hah
stevan nothingmuch: I need to get you those Mingus MP3s 13:08
nothingmuch indeed 13:09
stevan you can't play Jazz bass without listening to them
nothingmuch don't worry, i like chambers
and i like doug watkins
i'm missing a lot
but i have some education, still ;-)
stevan John Coltrane - A Love Supreme ++
shapr I like John Lee Hooker, Jimi Hendrix. 13:10
nothingmuch don't have that either
shapr and Led Zeppelin, of course.
stevan shapr: I got some nice Robert Johnson MP3s recently too
nothingmuch stevan, shapr if you don't have any concienceous objections, should i start a bt tracker for us?
stevan and some early Woody Guthrie commie/union stuff
shapr I don't do copyright stuff, if that's what you mean by conscientous objections. 13:11
ninereasons and Raaaay Charles
nothingmuch normally does that, and then buys it if he's still listening after a short while
shapr I just don't do it at all, and I encourage others to do the same.
stevan nothingmuch: me too,.. MP3s are only so good quality wise
shapr Because, the current laws are waaaay stupid.
stevan nothingmuch: I have no objections either 13:12
rindolf ninereasons: Hit the Road Jack, and don't you come back, no more, no more, no more, no more.
shapr And the only way they'll get fixed is by killing off the Media Mafia.
stevan shapr: you need only look at many of the 80s-90s hip hop artists to see how wrong the system is
shapr I agree
nothingmuch anyway, i'll do that bt thing, and who ever wants the cops at their door can join it =)
stevan (although that no longer applies now)
nothingmuch: in the words of Ice Cube, Dr. Dre and all of NWA ..... F*ck dah Police 13:13
nothingmuch fsck /dev/pig
ninereasons rindolf, Well I'm through with you, too bad you're blue I said move on, I said move on
nothingmuch bwahaha
damn, i'm so lame
stevan nothingmuch++ 13:14
nothingmuch anyway, back to work
stevan I am still getting errors with plain 'make' 13:18
stevan goes back to 'make optimized'
kungfuftr needs a faster box 13:20
shapr rob__: are you rob nagler? 13:24
rob__ shapr: actually im not mr. nagler :) 13:25
kungfuftr assign.t is hanging
shapr ok, whew.
I can skip the flaming then. 13:26
rob__ heh
kungfuftr on test 2... odd
bah... meetings to go to 13:28
masak shapr: now that pugs supports multidimensional data structures, what do you think is the feasibility of writing a board game in it? 13:32
shapr err 13:33
stevan masak: you might want to start with Tic-Tac-Toe first
masak :) good idea 13:34
shapr Who elected me clueful? =)
stevan Pugs still consumes a lot of memory and CPU at times
masak shapr: you and autrijus are the only ones i know in here :) 13:35
shapr Oh, ok then.
stevan masak: now you know me too :)
masak stevan: yes, hello :) 13:36
stevan hello masak nice to meet you :)
masak i'm the stereotype minimal committer. i have committed six lines of test code in t/subroutines/sub_named_params.t 13:37
stevan masak: anything is something :) 13:38
masak true :)
i'm not sure how to become more involved, short of reading up on haskell
and i have little time, otherwise i would do this full time
stevan masak: but you help when you have time... me I am in between projects so I have time to spare 13:39
but when my next project starts (in a week or so) my commit rate will go down 13:40
the ebb and flow of OSS :)
masak sometimes i wish i was out of a work so i could do this all day
living on air, coding pugs
stevan masak++ # prefering unemployment 13:41
castaway if only..
cognominal masak: I am going thru www.haskell.org/tutorial/. It is a good tutorial 13:42
stevan masak: shapr's HaskellDemo is nice too www.haskell.org/hawiki/HaskellDemo 13:43
shapr cheers
stevan shapr: I added it to READTHEM as well
shapr spiffy!
masak cognominal & stevan: thx 13:45
ninereasons pugs> my ($foo, $bar) = ('FOO', 'BAR'); 13:47
({ref:<Scalar>}, {ref:<Scalar>})
autrijus greetings!
nothingmuch in p5, is an assignment an lvalue?
autrijus yes, an assignment is an lvalue. 13:48
nothingmuch i.e., can i say ($foo ||= '') .= "blah"?
autrijus yes.
nothingmuch autrijus++
clkao hi autrijus
autrijus hey clkao.
masak hi autrijus 13:49
stevan autrijus: good morning/afternoon/evening
autrijus hey stevan. lots and lots of core changes.
jabbot pugs - 1873 - * removed pretty much all vCast from Pri
shapr yay 13:50
stevan stops his build and runs svn up
pdcawley does the same.
shapr does that too!
nothingmuch darcs get nothingmuch.woobling.org/Test-TAP-Model
shapr peer pressure!
autrijus heh... I'm not done yet :)
stevan autrijus: should I update the changelog again,.. or is more to come?
pdcawley Have we got objects yet? 13:51
autrijus stevan: you should update the changelog again... all my work currently is getting the lvalue model in
pdcawley: you have ref() and isa() and MMD, so yes
but you don't have classes ;)
nothingmuch pdcawley: sort of... they're basic infastructure
their
stevan autrijus: sounds good
pdcawley Right, so if I'm going to do the 'minilisp in perl6 on haskell' thing, then I need to write it in a functional styly for now?
stevan pdcawley: yes 13:52
pdcawley: there is some 'functional'-style stuff in examples/
nothingmuch pdcawley: yep
but wait a week ;-)
pdcawley Why? You're doing it already?
autrijus pdcawley: however basic class support is real soon now. I think by the end of this month :)
pdcawley Or we'll have objects by then?
nothingmuch pdcawley: not me, but autrijus 13:53
maybe me if i have more time, but i doubt that will make a diff
pdcawley I think if I can get my data structures and functions designed effectively, the conversion to an OO implementation should be straightforward.
stevan wow... over 30 commits since I went to bed last night (only 8 hours ago) 13:54
pdcawley However, a 'call_with_current_continuation' type thing, exposed at the language level wouldn't suck...
stevan pdcawley: t/unspeced/cont.t
pdcawley Ooh...
stevan (I think that is what you want)
autrijus there's also &sub.goto 13:55
pdcawley Want more than escape continuations.
autrijus you can have full continuations iff you can get larry to spec them.
:)
@larry, that is
pdcawley Dang.
stevan pdcawley: and t/subroutine/curry.t (more functional stuff)
osfameron why is larry now @larry ? 13:56
autrijus osfameron: design team
stevan autrijus: I think ingy is planning on arm wrestling larry re: POD
autrijus cool
osfameron autrijus: aha, thanks
stevan autrijus: you can challange him with the CPAN drinking game
nothingmuch stevan: is that a good thing or a bad thing?
nothingmuch thinks it's a good thing
design team needs shaking up IMHO
stevan nothingmuch: we just all need to watch out for the blue ice cubes,... 13:57
nothingmuch and the anarchistic approach of "just do it instead of them and see what they say" seems to be the sportiest way
ninereasons t/operators/assign.t still hangs for me 13:58
stevan ninereasons: yeah me too (with make optimized that is)
ninereasons still {ref:<Scalar>}) 13:59
my $a = 'FOO'
'FOO'
my ($a) = 'FOO'
({ref:<Scalar>})
that's why the swap hangs, I suppose. 14:00
recursive references
gaal btw if anyone is looking to write tests but doesn't knwo which, there's plenty to do in goto.t 14:02
Steve_p My guess is that there is refactoring to be done on just about any of the test files 14:03
stevan Steve_p: dont you know it :)
I did some cleaning last night, but it was far from sufficient 14:04
ninereasons frankly, it was awesome, stevan; but I'm sure you're right that there's more to do. 14:05
stevan ninereasons: thank you,.. but what I saw along the way,.. oh the horror,.. the horror :)
ninereasons wraps a blanket over shocked and trembling stevan 14:06
autrijus I'm fixing infinite loop
gaal stevan, any particular things you want us to work on?
lol, nothingmuch++ "# FIXME - the whole Test::Harness::Straps model can be figured out from this" 14:07
stevan gaal: hmmm, anything listed in t/force_todo would be a good place
gaal i mean, you mentioned refactoring 14:08
nothingmuch shit, i dropped my eraser in my coffee
stevan some of those are broken, and some are missing Pugs features, and still others are just messy
shapr nothingmuch: might taste better
gaal congratulations, you now have decaf
nothingmuch shapr: actually i was done with it 5-6 hrs ago
but it's turkish crap, with the gunk on the bottom 14:09
stevan yummy
nothingmuch just deep enough for it to be completely submersed
shapr I don't think that piece of plastic will erase marks anymore.
stevan turkish coffee marinated erasers
Aankhen`` Do multidimensional structures work?
Data structures, even.
nothingmuch Aankhen``: you mean by ref? 14:10
or like 'my @array is shape()'?
Aankhen`` By ref, I guess, e.g. my @array = ([1, 2, 3], [3, 2, 1]);
nothingmuch my pugs is not compiled right now, so i can't sya 14:11
but i am 90% sure that yes
stevan Aankhen``: that should work now (assuming you can compile) 14:12
Aankhen`` Ah. 14:14
Since which revision?
I compiled 1853.
It simply flattened the AoA (both references, BTW) in builtins/strings/length.t. 14:15
BBIAB. 14:16
stevan Aankhen``: svn up because we are now on 1873
autrijus infinite loop fixed. 14:17
about to commit ---
but I'm not sure of the logic in monads.hs line 232 at all.
stevan autrijus: I updated the ChangeLog (its now 70 lines long) 14:18
nothingmuch . o O ( aquaaariiuuuuss! )
gaal is the moon in the 7th house?
ninereasons Aankhen``, r1835 has ({ref:<Scalar::Const>}, {ref:<Scalar::Const>})
r1865 has ({ref:<Array>}, {ref:<Array>})
nothingmuch *horn sounds*
gaal: when?
gaal now. is it? 14:19
ninereasons svn up should cure your problem, Aankhen``
nothingmuch no clue
where is jupiter?
gaal dunno. aligned with mars?
nothingmuch no peace, though, to guard the planet
autrijus r1875.
jabbot pugs - 1874 - more ChangeLog additions 14:20
pugs - 1875 - * new context propagation logic to disti
cognominal ". o O ( aquaaariiuuuuss! )": nothingmuch betrays his age
nothingmuch cognominal: beh. I like duke ellignton's stuff, where the horns sound like electric guitars
age has nothing to do with it, good education does ;-)
cognominal :)
nothingmuch had, musically speaking, very good parents and friends 14:21
stevan cognominal: he got it from his grandmothers record collection :)
nothingmuch not only musical edu, actually. Mom said to me once 'did you ever watch hair? it's a good movie'
and dad took me to see Forman's guard changing parade thing in prague, because he thought i might appreciate it 14:22
ninereasons yay. works for me, autrijus 14:24
autrijus woot. 14:28
ninereasons i wish I could get these plugins installed on freebsd 4.11. It bugs me. 14:29
jabbot pugs - 1876 - * bad casting for "Any".
nothingmuch is invisible 14:39
autrijus hey mtve.
whew. assign.t all passes!
which means we have a rather sane semantics now
jabbot pugs - 1877 - * assign.t all passes, including weird o
Steve_p heh 14:40
autrijus runs another round of tests 14:41
ok. now I think I have time to look at SHA1...
castaway yay :) 14:43
nothingmuch does anybody have 10 minutes of sisyphic work time to donate?
shapr to what?
nothingmuch Test::TAP::Model
Steve_p I can take a look at it in a couple hours. What's needed? 14:46
nothingmuch test coverage
:w 14:47
damn
and lots of predicates
for convenience
the coverage is really for these
Steve_p predicates?
nothingmuch i have an object:
autrijus ok. can someone run another round of test for r1878 and post url?
nothingmuch Test::TAP::Model::File
autrijus I need my laptop's cpu power for hs-plugins :) 14:48
nothingmuch it wraps around the hash structure in Test::TAP::Model
and gives you convenience methods
$t->passed_tests
$t->bailed_out
etc
autrijus: will be ready in 10 mins
autrijus danke
nothingmuch or whatever it takes for the cycle
jabbot pugs - 1878 - * restore join() to sanity. 14:50
Steve_p OK, I'll take a look at it
nothingmuch Could not find module `Plugins':
use -v to see a list of the files searched for
(imported from src/External/Haskell.hs)
make: *** [optimized] Error 1
Steve_p: i'll submit my progress as soon
autrijus the probing is still bad?
omg.
nothingmuch this is after make purge 14:51
autrijus I'll take over ;) 14:52
nothingmuch ssh to eris
i'll chown /usr/local/src/pugs
autrijus please svn up
perl Makefile.PL
and try agian
see if r1879 works
Steve_p Where is Test::TAP::Model hiding?
nothingmuch Steve_p: darcs get nothingmuch.woobling.org/Test-TAP-Model
i have yet to submit the File.pm file
ok, seems to work 14:53
jhorwitz autrijus: any progress on the parrot backend lately? 14:57
autrijus hey jhorwitz!
jhorwitz emerges from lurking
autrijus jhorwitz: nope. the priority is to get 6.2 out of door
then write PA02
then PugsIMC
jhorwitz more time for me to grok haskell... 14:58
autrijus :D 14:59
jabbot pugs - 1879 - * brutal way to check hs-plugins
castaway brutal++ :) 15:00
pdcawley Mmm... brutality!
Steve_p Anyone here get darcs to compile with GHC-6.4 15:01
autrijus Steve_p: it seems to be a known no-go
nothingmuch Steve_p: not me. Be cheap and get a binary
or just wget -r,
castaway gets inline haskell disabled (but then I did before too) 15:02
autrijus require_haskell "/usr/local/lib/perl6/site_perl/mach/SHA1__0_0_1.o" 15:05
pugs: /usr/local/lib/perl6/site_perl/mach/SHA1__0_0_1.o: unknown symbol `SHA1_sha1_closure'
ok, I get that far.
Internal error while running expression:
resolveObjs failed with <<False>>
autrijus fixing
nothingmuch Steve_p: OK, File.pm added 15:09
jabbot pugs - 1880 - remove obsolete comment
nothingmuch I'll write Test::TAP::Model::Subtest now, and then refactor ::File to use that for extraction 15:10
Steve_p nothingmuch, got it :)
nothingmuch thanks 15:11
rindolf Everybody join #perl! We're 15 people short of having 400 participants. 15:12
14
nothingmuch rindolf--; # beh, this is so pointless 15:13
rindolf nothingmuch: I was joking.
perlbot karma rindolf
perlbot Karma for rindolf: 2
pdcawley rindolf-- 15:14
rindolf--
Steve_p rindolf--
pdcawley Because neutral is good.
Dang.
rindolf++
nothingmuch oh my
pdcawley karma rindolf?
jabbot pdcawley: rindolf? has neutral karma
nothingmuch rindolf: you can ++ yourself ;-)
rindolf pdcawley: thanks
nothingmuch (j/k)
autrijus require_haskell works. 15:15
rindolf perlbot karma rindolf
perlbot rindolf doesn't have any karma
autrijus cleaning up and about to commit.
pdcawley perlbot karma pdcawley
perlbot pdcawley doesn't have any karma
rindolf pdcawley++
nothingmuch btw, autrijus , make test didn't finish yet
autrijus nothingmuch: is it looping?
nothingmuch oh my, were having a karma orgy
t/data_types/hash_ref.t <-- loopy?
if yes, then I guess that yes
i've been looking at it for 30 seconds
no change
autrijus aw :-/ 15:16
looks yes.
terribly sorry. :(((
fixing
nothingmuch karma sutra
jabbot nothingmuch: sutra has neutral karma
nothingmuch autrijus: you can ssh to eris to debug this stuff
castaway wonders what autrijus did, and how it "works"
shapr Hail Eris! 15:17
rindolf nothingmuch: I won't like you anymore unless you bump my karma.
nothingmuch rindolf: why?
rindolf is kidding
10 people short of 400 15:18
nothingmuch with you it's always very hard to understand
normally to imply cynical tone people either exaggerate wildly, or stick in a bunch of ';-)' '=P' or ':)'s
with all your proper punctuation and stuff
autrijus I like proper punctuation. ;-) =P :) 15:19
Aankhen`` Heh.
rindolf autrijus++
Aankhen`` Everyone has a poker face on IRC.
:-| <--- my poker face.
autrijus do I get to be the joker?
pdcawley UWTLS?
jabbot pugs - 1881 - * require_haskell now runs... more work
ninereasons sorry, autrijus. the king cannot be the joker. 15:20
pdcawley realises that if he doesn't keep up with p6l he's not going to have a clue what most of it means.
autrijus "unless we take Luke's suggestion."
nothingmuch two hundred and fifty six!
pdcawley So EIWTLS is 'Even if we take...' 15:21
rindolf Why aren't we using CIA as a commit bot?
autrijus howto? 15:22
nothingmuch because pugs is not yet tracked by CIA
if you
if you have the know how to make it work, i reckon we'll all be pleased =)
autrijus yup.
rindolf nothingmuch: cia.navi.cx/doc/adding-projects 15:23
nothingmuch you're welcome to do it, i don't have the tuits to mess with stuff like that 15:24
or rather, i do
but i have higher priority things to fill up this time
rindolf nothingmuch: I think I need access to the physical repository. 15:25
nothingmuch if you need dirty work, like adding hooks, autrijus can probably help 15:26
if you need regular commit access, that's not an issue
if you tell me what to do i'll do that
but if it involves signing up/making decisions/communicating i'd rather write the code i'm working on right now
because the test graphing has been waiting for refactoring a long while, and i finally have an excuse to do it while at work 15:27
rindolf nothingmuch: I think I need to setup a hook.
nothingmuch i don't know how openfoundry manages this stuff
i have admin access, but i'm not 100%
autrijus rindolf: post the details on setting up svn hook on some url?
nothingmuch i'm not a subversion user, so i don't know what i can do
autrijus rindolf: I'll redirect foundry people to it.
nothingmuch 100% sure i can do it
rindolf autrijus: they're in the URL I pointed you to. 15:28
autrijus ah. I missed that link.
rindolf autrijus: cia.navi.cx/doc/adding-projects
And then the whole world and his mother will see the Pugs commits on #commits 15:29
machack666 hey all, is there a rev that builds all tests on Mac OS X? There was one that was close last nght (<60 tests failing), but now it looks like there are a lot more failing again (mostly to do with vCast, looks like)
rgs that'll double the traffic on #commits.
nothingmuch machack666: you can try before itypes 15:30
jabbot pugs - 1882 - Updated listification syntax for strings
nothingmuch but right now tests are hanging
autrijus: are you going to use eris, or can i keep it in the smoke loop?
machack666 thanks, nm -- anything I can help with?
autrijus nothingmuch: I'm not going to use eris; I'm almost getting the hanging fixed
nothingmuch machack666: errm, not that I know how to tell you to help
in general i can help 15:31
but not specifically
i can tell you: write tests
machack666 ok -- I'll just wait on bated breath. :)
nothingmuch but that's about it ;-)
autrijus machack666: examples/cookbook/ needs help :) 15:34
machack666 i'll take a looksie.
yes, it is fairly spartan. 15:36
is the idea to only translate when there is a distinction between the P6 and P5 solutions? 15:37
nothingmuch i think so 15:39
but i think it would be nice to translate to "good" perl6 where applicable 15:40
otherwise p6 might look just like p5 with a funnier syntax
autrijus yeah.
nothingmuch i.e., zip, junctions, hyper ops - these should really shine in cookbook examples
mj41 So. How to translate this to perl6?
foreach my $tut_num ( 0..$#$index ) {
$prev_tut_fn = $tut_fn;
$tut_fn = $index[ $tut_num ];
if ( $tut_num+1 <= $#$index ) { $next_tut_fn = $index[ $tut_num + 1 ]; } else { $next_tut_fn = undef; }
ninereasons go, go, nothingmuch preach
autrijus infloop fixed. 15:41
I think.
nothingmuch autrijus: do you know if there is an $?ITER var? 15:42
i resort to many 'my $i; for (@_){ $i++; ... }' things in p5
i don't like that
i think it's useful enough that we can subvert p6l
in which case this will look like:
autrijus r1884
nothingmuch for $tut_num (@index) {
no 15:43
acki
machack666 nothingmuch: wasn't there a discussion about for @ar.kv?
autrijus eh. you do
nothingmuch for $tut_num (0 .. @index.elems - 2){
autrijus for @index.kv { ... }
that even works now 15:44
nothingmuch ah, that's cool
ok, so
for @index.kv -> ($index, $value) ?
or do you need to take the pair apart in the loop?
autrijus for @index.kv -> $idx, $val { ... }
nothingmuch ok
autrijus or, implicitly
for @index.kv -> { say "$^i: $^x" }
nothingmuch $^key, $^value; # they sort ok ;-) 15:45
autrijus yeah, isn't the alphabets great.
nothingmuch: can you run another round of smokes?
I'm pretty confident it's not looping now
nothingmuch running
i think we should be really nasty with this one though 15:46
for (@index[0 .. @index.elems - 2] Ā„ @index[1...]) { 15:47
....
}
maybe the subscript slices can be niftier?
$^a is tut
$^b is next
what does the rest of the loop look like? 15:48
ninereasons with parens, is the -> arrow is not necessary for implicit vars; isn't that right?
nothingmuch i prefer $^former and $^latter
=)
i think it shouldn't be
but i have no clue
PerlJam -> is not needed for $^foo 15:49
machack666 if infinite slices do the right thingōæ½x81, does for (@index[0..Inf -2]) do the right thing as well?
nothingmuch can you say 'for (@index[0 .. -2] Ā„ @index[1...]) {'?
PerlJam they are completely orthogonal.
nothingmuch machack666: i'm not 100% sure, we had a discussion on that
on 0 .. -2 doing the right thing
that is, really meaning 0 .. @index.elems
PerlJam machack666: I would think not.
jabbot pugs - 1883 - * hash_ref infinite loop fixed. 15:50
pugs - 1884 - * do not allow storing into constant has
PerlJam nothingmuch: same to you.
nothingmuch but i think 0 .. Inf does not mean 0 .. @index.elems
i don't remember what the conclusions were
machack666 it seems like there could be a handy shorthand for that
PerlJam nothingmuch: it does.
well, it *can*
nothingmuch PerlJam: sorry, that came out stupid
i think it does not mean that syntatically
the consequences are the same
but Inf -2 is Inf, not @index.elems -2
woah, idea: 15:51
machack666 but only if - binds tighter than Inf
:)
nothingmuch @array![-2, -1]
theorbtwo Hm, if Inf-2 became Inf but something-2.
nothingmuch negated subscripts
or maybe [^ ]
theorbtwo Er, something<-2>
PerlJam nothingmuch: yeah, I thought of that too just now (I'd use ! but in a different place)
machack666 or something<..-2> : Inf from the other direction! 15:52
nothingmuch you mean @array[!-2]?
PerlJam nothingmuch: yep
integral Perhaps a smaller infinity that really is only as big as the thing indexed? So Inf-2 = Inf, LittleInf-2 = @whatever.elems-2
nothingmuch that is i-2 is '-2 but false'?
that's nice
i prefer yours, it's much more flexible
but then we'd have to include set operations on subscripts
and make them high order objects
(which i wouldn't mind, really, they should be very useful)
i'll write to P6l 15:53
autrijus $ ./pugs -e "require_haskell 'SHA1__0_0_1.o'; say sha1('')"
da39a3ee5e6b4b0d3255bfef95601890afd80709
theorbtwo Yey!
integral autrijus++ 15:54
autrijus theorbtwo++
dons++
rindolf autrijus: what about me?
autrijus now, to go from here to "require SHA1;"
Khisanth that is with the plugins?
autrijus Khisanth: yes, with hs-plugins
you can grab tpe.freepan.org/repos/ingy/SHA1/
install pugs, install that
Khisanth the 0_0 looks funny :)
autrijus and voila.
it will go away :)
it's version number.
Khisanth hrm it's a .o not a .so? 15:55
theorbtwo Don't we need it to support multiple installed versions of the same module?
autrijus we do
it will go away from the syntax.
not from the filesystem.
"require SHA1" needs to figure out which .o
theorbtwo (For that matter, shouldn't there be an "INGY" in there somewhere.)
Ah.
autrijus ingy++ # too
theorbtwo Ah! 15:56
The '&':name is what was missing.
autrijus yup.
entirely my fault. 15:57
also you need to point getConfig to it.
so it knows archlib etc.
also I instrumented Perl6::MakeMaker to install things from src/.
it should go to package-specifc dirs later.
but now I don't care :)
theorbtwo OK. 15:58
nothingmuch machack666: so what does the rest of that recipe look like? 15:59
jabbot pugs - 1885 - * require_haskell Really Works.
nothingmuch PerlJam: Message-ID: [email@hidden.address] 16:01
autrijus: assign.t seems to hang 16:02
ninereasons which version, nothingmuch? after 1880 ?
nothingmuch 1884
autrijus gah, it hangs _again_ ? 16:03
sorry :-(
gimme a sec.
nothingmuch yup
ssh eris
you know you want to
because I have to go home
autrijus okie. 16:04
nothingmuch i've setgid everything to my group, and chmodded so i have write, so we should be able to get along as long as your umask is sane
mj41 @a.elems-1 == @a.last_or_something_like ? 16:05
autrijus danke. sorry for causing this much trouble
nothingmuch that is, insane
no trouble
i opened the shell account precisely for this purpose
machack666: ping
PerlJam nothingmuch: you know you're going to get back responses like @array[all(0..,none(-2,-1)).list] (or whatever the syntax is to get a list of all of the values in a junction).
machack666 nothingmuch: mj41 was the guy with the recipe, I beieve 16:06
nothingmuch that sucks
ah
mj41: ping
;-)
sorry machack666
mj41 for $a.kv -> $idx, $val {
if ( $idx > 0 ) { $prev = $a[$idx-1] } else { $prev = undef };
if ( $idx + 1 < $a.elems ) { $next = $a[$idx+1] } else { $next = undef };
}
nothingmuch where is the tut_num example coming from?
machack666 i'm too busy doing productive work, like getting utf-8 support working with emacs. ;)
mj41 sorry
nothingmuch i'm not really sure exactly what it does 16:07
autrijus now, let's see if it works....
autrijus crosses fingers
$ ./pugs -MSHA1 -e "say sha1('')" 16:08
da39a3ee5e6b4b0d3255bfef95601890afd80709
yes! yes!
nothingmuch crosses fingers too
autrijus++
mj41?
PerlJam autrijus: excellent! You've laid the foundation for the SHA-NA-NA module! ;-)
PerlJam wonders if the humor is lost on this audience. 16:09
jabbot pugs - 1886 - * pugs -MSHA1 -e "say sha1('')"
theorbtwo s/NA/NaN/g 16:10
theorbtwo got the reference after a bit.
PerlJam Perhaps I'm just an old american.
nothingmuch ha ha ha ha! 16:11
i was away for a second, i promise
*cough*
mj41: see nopaste 16:12
perlbot nop
perlbot nopaste
perlbot Paste your code here and #<channel> will be able to view it: sial.org/pbot/<channel>
pasteling "nothingmuch" at 212.143.92.226 pasted "for loop, with $prev and $next" (6 lines, 146B) at sial.org/pbot/9269 16:13
nothingmuch err, the first subscript should be [email@hidden.address]
nothingmuch leaves 16:14
16:15 bradb is now known as bradb|out
mj41 $pugs.can(zip) ? 16:18
theorbtwo pugs -e "@foo=zip('a'..'f', 10..15); say @foo.perl" 16:19
stevan PerlJam: I got your reference, ... but 10 minutes late after reading the backlog :)
mj41 No compatible subroutine found: &zip 16:20
theorbtwo Hm? I thought it worked... 16:21
stevan mj41: that doenst mean it is not there,.. it just means you didnt supply the right args
t/builtins/list/zip.t
autrijus bad precedence for "zip" 16:22
someone wrote "left" where it should be "pre"
machack666 are here docs supported in pugs? 16:23
autrijus no.
machack666 which Synopsys talks about here docs? a non-existant one? ;) 16:24
i'm trying to translate some recipies and I know that at least the quoting mechanism has changed. I'd like to find more info about where they can go, etc...
pasteling "mj41" at 147.229.221.107 pasted "array prev and next if defined, good enough?" (12 lines, 356B) at sial.org/pbot/9270 16:25
autrijus zip() is in. 16:27
machack666: qq:to
S02.
machack666 autrijus: thx
jabbot pugs - 1887 - * zip(). 16:29
theorbtwo D'oh! I forgot the -i on my grep. 16:30
autrijus now going back to fix assign.t loop... 16:32
theorbtwo: so, now mut rec is not a problem for us, I think I'll release 6.2.0 with mutrec 16:33
is "make optimized" still bad?
theorbtwo Sounds good to me, autrijus. 16:34
stevan autrijus: trying make optimized right now 16:36
ninereasons S02 makes it sound like [email@hidden.address] is the literal equivalent of $#foo. is that correct? 16:37
stevan hmm, so what would (0 .. Inf).end() do? 16:38
autrijus looks so
stevan: it will probably return Infinity after either constant time or nonconstant time.
jabbot pugs - 1888 - * deref scalar in value context. 16:39
pugs - 1889 - * oops, remove debug
ninereasons I don't understand what larry means by [email@hidden.address] or [-1]' ? 16:40
theorbtwo grins.
autrijus it means that
theorbtwo I'd hope after constant time.
autrijus @foo[$#foo]
needs to written as @foo[-1].
hth.
stevan: I'm making a round of tests 16:41
I think we're good to preflight.
stevan autrijus: I am too (make optimized && make test)
autrijus wait -- lvalue zip? 16:42
whose idea is it? is it in the synopses?
stevan where do you see this? 16:43
autrijus zip.t
machack666 make && make test still has failures on mac os x...
autrijus failures is fine
infinite loop is not :)
stevan needs to get more RAM,... make is getting slow 16:44
machack666 ok 16:45
do the pipe operators work yet?
autrijus no 16:46
I think I see an infinite loop at t/pugsrun/06-dash-help 16:48
sigh
machack666 hanging on operators/assign (1/51) mac os x w/ hs-plugins 16:53
kungfuftr moo 16:55
autrijus machack666: probably not up to date
kungfuftr reruns smoke again 16:57
machack666 i think i'm using the alternate svn server. is there a sync delay?
autrijus there is.
Corion datenzoo.de/pugs/win2k.html - r1887
autrijus please use svn.openfoundry.org one
gaal 1890? 16:58
machack666 i don't really know svn well -- how to change without rechecking?
iblech [t/operators/assign.t] -- 51/51 passes @r1890
Corion Hmmm. In abs.t - it checks for Int, but type Num is just as good IMO, or even better as a result of abs()
Oh. But they mutate the type ... that's the bad thing about it, now I realize :) 16:59
nothingmuch evening
jabbot pugs - 1890 - * zip now uses length of longest; zip.t 17:00
Corion Eh... @INC.perl.say outputs "\('... " - what's the backslash doing there? 17:01
autrijus what's "boustrophedonic"?
Corion: it means that it's a reference to a list...
ninereasons maybe 'undef' behaves differently on win32 ? assign.t is failing there on r1887, for Corion
nothingmuch www.worldwidewords.org/weirdwords/ww-bou1.htm 17:02
Corion autrijus: Ah. Hmmm... which makes my reparsing test bork
ninereasons: It loops endlessly until I killed assign.t
cognominal An ancient mode of writing, in alternate directions, one line
from left to right, and the next from right to left (as
fields are plowed), as in early Greek and Hittite.
autrijus Corion: I think we can drop that... but then,
$foo = @INC; $foo.perl
will need the \
Corion autrijus: I don't mind changing my test :) 17:03
autrijus I don't think I can fix it right now
cognominal www.perlmonks.org/?node_id=45852
autrijus yay, please do :)
nothingmuch autrijus: did the tests pass on eris?
Corion Aiee - and somebody changed the option parsing :)
autrijus nothingmuch: had not rerun yet, still fixing bugs 17:05
Steve_p nopaste?
gaal autrijus: think printers
nothingmuch ok, i'm restarting the loop now
autrijus thinks printers
gaal old printers
nothingmuch makes noises
gaal that took a while to move the printing head
nothingmuch bzzzzzzzz, kashclick, bzzzzzzzzzzzzzz, kashclick, bzzzzzzzz
ah! 17:06
gaal if it's already at one side, why rewind it before printing another line?
nothingmuch trrz trrz trrzz trzzz
(the old one was a very quick laser printer)
gaal no CR
only LF
these guys were the earliest unix users!
or mac users, i can never remember.
nothingmuch gaal: join("\b\b", split(//, "other line"); 17:07
;-)
theorbtwo BTW, most modern inkjet printers print that way -- they just reverse every other line.
mj41 111 users 17:08
gaal so even modern printers are boustrophedonic. :)
nothingmuch mj41: so what does that example do?
jabbot pugs - 1891 - * split() should not strip leading field 17:09
pugs - 1892 - * length.t all passes; reenable io
Steve_p nothingmuch, --- Test-TAP-Model_0/t/basic.t 2005-04-12 10:12:01.939300464 -0500 17:10
Crap!
nothingmuch, paste.husk.org/3001
gaal we actually have the word boustrophedonic in the pugs source tree.
Steve_p Test coverage for Test::TAP::Model at 100%
nothingmuch Steve_p: wants to darcs send?
yay!
Steve_p++; # amazing work on p5p, and also here 17:11
ninereasons Corion does >= r1890 still hang for you? 17:12
Corion ninereasons: Dunno - just fixing a test then resmoking
Juerd @print = @lines[0... :by(2)] Y @lines[1... :by(2)]>>.reverse
nothingmuch dinnertime
autrijus 06-dash-help no longer borks.
Juerd Except real life wants CRLF to stay at the end, and not be moved to /^LFCR/
nothingmuch Steve_p: if you darcs send by the time i'm back i won't apply, there's a conflict so far
so i'll merge in a bit 17:13
ninereasons i was able to find a backversion of r1887, and it did hang as you describe, also on freebsd
autrijus looks some more at testgraph
wolverian Juerd: is p5 smart about that?
it is, apparently.
er, no it's not.
damn you reverse in list context.
Juerd wolverian: P5 doesn't have Y, P5 doesn't have ..Inf, P5 doesn't have :by, P5 doesn't have >> and it doesn't have a reverse method on arrays. What exactly is left to be smart about? :) 17:15
autrijus :D
Juerd++
wolverian Juerd: the CLRF :)
Juerd only now realises how hard this same thing would be in Perl 5 :)
wolverian: Oh! :)
wolverian: Just autochomp and add it again when printing 17:16
Easiest solution :)
s/reverse method on arrays/reverse method on scalars/, though it has neither.
And of course, you want LF only, not CR 17:17
ninereasons is '@ARGS is chomped' still current, or was it abandoned?
Juerd CR kind of defeats the purpose here
machack666 dumb question: will infix:<Y> be an actual synonym for infix:<&yen;> ? 17:18
Juerd machack666: It already is.
machack666 unicode, schmunicode! ;) 17:19
jabbot pugs - 1893 - * any junctions must be reduced.
pugs - 1894 - Restored -I test due to @INC.perl.say re
Khisanth wonders if Perl7 will have Z 17:20
pjcj wonders if Perl7 will be Z
Steve_p is Z 17:23
machack666 is going to digest some more on the synopses
bye all
hlen wonders what Y does
Juerd hlafarge: It is zip() 17:24
eh
s/hlafarge/hlen/
Corion My test smoke is still running, but didn't hang.
r1894
hlen Juerd, ok.. thanks, i'll look into zip()
gaal it's time to think of -j3 for make test/yaml_harness... a complete smoke is no longer very fast 17:25
ninereasons found 'is chomped' in E02
Corion gaal: Was it ever? :)
Juerd hlen: See also juerd.nl/files/slides/2005dpw/en/perl6op.html for a very quick introduction
gaal wasn't too bad a month ago 17:26
hlen Juerd, ok, thanks again =)
pdcawley Woo... summary's written. 17:27
Corion r1894 - datenzoo.de/pugs/win2k.html
pdcawley++ 17:28
Juerd pdcawley: Are you writing the p6[lci] summaries again now?
pdcawley Every other week.
Juerd Ah, good!
pdcawley Keeps the workload manageable for both Matt and me.
Juerd pdcawley++
pdcawley I know from experience that if you slip to fortnightly the work gets really, really, heavy.
nothingmuch gaal: i'll try to hack that into Test::TAP::Blah 17:29
Corion BTW, I get some unexpected successes - should I hunt them down and promote them ?
nothingmuch should be easy to do with forks, since the data is easily serializable
but i would like to do it in a subclass
Corion (given.t, sub_return_values.t, short_circuit.t)
elmex ack 17:32
autrijus I'll promote given.t. 17:34
nothingmuch returns
autrijus you can do the rest
Corion autrijus: Sir, YES! Sir! 17:35
nothingmuch Steve_p: ping 17:36
Steve_ping
autrijus: 06-dash-version.t 17:38
hangs
gaal not on r1890/msys 17:39
nothingmuch svn ups once more
Corion Weird - one "unexpected success" seems to be a Test.pm bug - it is reported as TODO, even though it is a plain "is()" call
nothingmuch Corion: that happens usually when there are '#' in the test desc 17:40
and it's a harness bug, not a Test.pm bug
and i'm not 100% sure it's a bug
Corion nothingmuch: I'm looking if it is mentioned in force_todo :)
Yep - they are still in force_todo :)
nothingmuch pugs t/blah.t | grep '#';
i saw a grep that highlighted matches 17:41
Corion force_todo should output a real TODO message (like, "for release")
nothingmuch wonders where that was
where is that?
autrijus Test.pm 17:42
Corion commits
nothingmuch is sooo out date
autrijus commits
looking good 17:43
nothingmuch reloops
autrijus ok, r1897. 17:44
shower now; I'll do preflight + release when I get back.
meanwhile, if anyone fancies reading at svn log and add the past few hours' work to ChangeLog... :) 17:46
brb &
jabbot pugs - 1895 - Untodoed some succeeding tests 17:49
pugs - 1896 - * given.t all passes, thanks Eris.
pugs - 1897 - * 06-dash-version should not match again
ninereasons it's nice that jabbot walks toward us when speaking now, instead of away as before. 17:51
shapr agrees
Hail Eris!
nothingmuch hail eris!
pasteling "Corion" at 217.86.61.174 pasted "Coroutine testbed" (63 lines, 1.3K) at sial.org/pbot/9271 17:52
Corion I've pasted my coroutine testbed (which I haven't even once run through Pugs), in case somebody wants to toy with it. I won't get to toying with it during this week.
gaal [OT] cuuute! perlnomic.org/ (i suppose i'm a fan of Peter Suber)
Corion It's supposed to become an interface to irc, of course, but I've left out the sockets stuff for the start. 17:53
Steve_p nothingmuch? 18:02
nothingmuch yes 18:03
Aankhen`` Yey, the array_ref tests aren't hanging any longer! 18:05
nothingmuch Steve_p? 18:07
gaal r1901, but this is a fix in the yaml harness and doesn't require another smoke round, folks. 18:08
jabbot pugs - 1898 - <Corion> force_todo should output a real 18:09
pugs - 1899 - * repair want.t
pugs - 1900 - * splat now flattens for real
pugs - 1901 - fix failing test xref on msys
autrijus *** Evl: App "circumfix:{}" ('aasd')
wha? 18:10
I didn't even know we have circumfix:
nothingmuch circumfix is like hyper, right?
autrijus oh wow we do.
no, it's { }
nothingmuch i think hyper is a circumfix op though 18:11
even if { } is too
autrijus oh. ok. 18:12
anyway, fixing pair.t
4 tests to go!
gaal msys: t/magicals/pid.t 18:13
'.' is not recognized as an internal or external command,
operable program or batch file.
and: t/pugsbugs/env_is_readonly.t 18:14
Can't locate File/Spec.pm in @INC (@INC contains: ext/Test/lib /usr/lib/perl6/msys /usr/lib/perl6 /usr/lib/perl6/site_perl/msys /usr/lib/perl6/site_perl . blib6/lib).
Val VUndef
we dont' support cygwin at all, do we? i should just s/cygwin/msys all over t/ 18:15
kungfuftr latets smoke on my box is up 18:16
gaal oh, my bad re:env_is_readonly, right? i should put blib in the search path.
Corion Heh - thanks to whoever added that fix upon my blathering 18:17
gaal so, autrijus, can i chuck cygwin for msys?
nothingmuch pugs: cannot cast from VList [] to AST.VCode
for every test
crysflame msys?
gaal Corion, the win32 couldn't xref blah? that was me
crysflame: msys ~~ mingw 18:18
Corion gaal: I meant r1989
crysflame ok
Corion err - r1898
gaal i don't know why they needed two names
ah
mj41 MSYS or Minimal SYStem is a POSIX and Bourne shell environment use with MinGW
gaal so is the arch ever mingw? on my box it's msys 18:19
i've never seen mingw
or could it be that it depends on the order you install them?
kungfuftr autrijus: know where to get binary ghc package for freebsd 5.3? 18:20
autrijus kungfuftr: pkg_add should do no?
clkao mentioned somewhere has it
kungfuftr autrijus: pkg_add -r ghc doesn't have any on the mirrors 18:21
gollem: ye cunt! 18:22
shapr blinks
kungfuftr =0)
gollem you know it
shapr Ƥr du sƤkert?
finns det nƄn annat?
kungfuftr gollem est my housemate... a fellow perl geek
shapr jasƄ 18:23
det Ƥr okej
mj41 MSWin32 - nmake - ActivePerl or
mingw - mingw32-make - MinGW and ActivePerl or
msys - make - MinGW, MSYS and ActiveState or
msys - make - MinGW, MSYS and MSYS-DTK (MinGW perl)
gollem nej , finns ingen annan
shapr Ƥr du svensk?
gollem japp
shapr hur kommer du till uk?
gollem hehe
bra frōæ½xE5ga
kungfuftr bah!
gollem fick jobb
shapr intressant 18:24
kungfuftr: cool, you live with a swede
gollem it's very very cool..
:)
shapr men jag ōæ½xE4r inte sōæ½xE4kert om jag lōæ½xE5ter svenska ocksōæ½xE5 18:25
jag kommer frōæ½xE5n USA, men bor i Boden
gollem ok
jag bodde i Luleōæ½xE5 fram till i december dōæ½xE5 jag flyttade hit
shapr well dang
I never meet the local geeks
that sucks
gollem hehe
shapr Or maybe I did meet you?
gollem doubt it 18:26
shapr Did you ever meet the crazy american person with a reputation for writing Haskell?
The unicyclist in Boden?
Ok, never mind :-)
gollem yes
err
hmm
in Boden? no ,maybe not
shapr I hung out in Luleōæ½xE5 some with an Irish friend of mine. 18:27
gollem i went to the Uni there
shapr Do you know Tony Scully in CDH/CDT? 18:28
gollem no
shapr ah well
tyvōæ½xE4rr
gollem what is an american doing in Boden anyways? 18:29
shapr I met this chick on icq, (50,000 words omitted) and now I live in Boden.
nothingmuch hah
kungfuftr what is anyone doing on boden is a better question
nothingmuch castaway and theorbtwo both have such stories
Corion Yay! The segfault when evaluating NaNs has gone too!
jabbot pugs - 1902 - * fix named vindings. 18:30
pugs - 1903 - * pair.t and multidum repaired.
pugs - 1904 - * junction_functions repaired.
gollem shapr: i hope she's worth it...
Corion nothingmuch: But they sound so alike. It's as if they copy their story.
shapr grins
It's not a great sacrifice to live in Boden. It's a beautiful place, and I have internet. 18:31
gollem how long have you lived there?
zuulvin_ wedwdw
dddddd
shapr two years in Boden, three years in Tornio before that.
puhutko suomea?
gollem ooh
autrijus Corion: win32 happy? 18:33
Corion autrijus: Yep 18:35
autrijus care to do another full test run?
I think I'm ready for preflight.
Corion There are some Win32 issues (PID, ENV), but nothing pressing :)
autrijus reviewing changelog now
nothingmuch autrijus: should i try to finish mine for once?
autrijus yeah, sure, as long as the tests are TODOed/skipped
nothingmuch: sure!
nothingmuch it is at oo
Corion datenzoo.de/pugs/win2k.html - r1901
(and I have to fix another of my tests that fails due to the changed string-eval semantics) 18:36
autrijus nod. I've fixed all the visible reds here
or forced todoed them
nothingmuch t/operators/binding.t
# Looks like you failed 1 tests of 9
autrijus ok, take your time
nothingmuch r1902 finishing soon
bad plan else where
kungfuftr autrijus: kungfuftr.com/pugs-smoke.html too
autrijus r1906 has bindings done
nothingmuch beh, you'll see 18:37
autrijus r1906 is all clean here
gaal t/builtins/io/io_in_for_loops...............ok
1/49 unexpectedly succeeded
nothingmuch ok, it'll update soon
autrijus I think, at least
kungfuftr still doing another run
nothingmuch kungfuftr: hopefully this will soon be consolidated
i'm working hard on this
if you want to help with Test::TAP::Model i'll be grateful
kungfuftr nothingmuch: oh... nice
nothingmuch darcs repo is up
kungfuftr yup, been trying to do test stuff last few days
real life-- 18:38
autrijus magickally, "make optimized" worked again
Corion Oooo - conflict :)
autrijus oooh exciting
kungfuftr oooh... cheese! 18:39
Corion Nah - I think the conflict was either resolved or gaal directly fixed the test-bug as well :)
jabbot pugs - 1905 - add msys to windows junctions 18:40
pugs - 1906 - * fix xor; clean the rest of tests
gaal t/magicals/pid..............................NOK 1# Looks like you failed 1 tests of 1
Corion gaal: PIDs are broken on Win32 18:41
(all PIDs are "1" :) )
gaal ah :)
Corion gaal: I added a test for that, because one day I intend to fix it :)
gaal good, good
Corion gaal: The test checks if the child PID differs from our $*PID :)
Corion does another smokeround
gaal t/operators/binding.........................NOK 7# Looks like you failed 1 tests of 9 18:42
Corion I hope that we get embedding soon, so I can start stealing from Perl5 and/or using Win32 DLL functions
gaal t/operators/filetest........................FAILED tests 2-3, 5-6, 23, 26
Failed 6/30 tests, 80.00% okay (less 4 skipped tests: 20 okay, 66.67%)
Corion Heh. Should we have #pugssmoke, where everybody does live reports of their test runs? :)
kungfuftr how longs kwiki.org been dead? 18:43
gaal t/operators/short_circuit...................FAILED tests 15, 19
tell me if i should stop sayig all this
we have a problem with how slow these cycles are
kungfuftr gaal: run the smoke and upload the html to somewhere 18:44
=0)
gaal t/pugsrun/01-multiple-e.....................FAILED test 7 # Corion, this one's for you :)
kff, but that has latency
Steve_p kungfuftr, seen reports from yesterday
jabbot Steve_p: I havn't seen reports , Steve_p
autrijus I force_todoed them
Corion gaal: Fixed in the latest commit I think
gaal how latest is the latest?
nothingmuch Steve_p: ping
Steve_p heh
autrijus r1909.
Corion r1908 is what I'm smoking 18:45
Steve_p hi nothingmuch
nothingmuch sent?
autrijus Corion: it will have one bad plan() in short_circuit
otherwise should be fine
I'm rerunning another round as well.
nothingmuch nothingmuch.woobling.org/pugs_test_status/
gaal how do i tell which ver i have co'ed?
Steve_p Yeah, darcs send said it did
gaal oh, right, svn info 18:46
kungfuftr gaal: `svn up` should tell you after it's finished
nothingmuch goes to clean the kitchen
gaal yeah, but i don't want to do that while in the middle of a smoke :)
nothingmuch Steve_p: i'll check mail soon
gaal okay, i had 1905. 18:47
see, it's impossible to test when there are four commits before the test is over :)
and this is a relatively fast machine!
autrijus t/pugsrun/01-multiple-e.....................ok 18:48
1/7 unexpectedly succeeded
nothingmuch 's is still compiling for 190x, dunno which
but optimized
autrijus t/pugsbugs/env_is_readonly..................ok
2/3 unexpectedly succeeded
known?
if yes, remove them from force_todo please
gaal autrijus, i think this may be a platform-dep issue, because for me that failed. 18:49
kungfuftr 2029!?!?
jabbot pugs - 1907 - * retract one too eager force_todo 18:50
pugs - 1908 - Fixed another test that broke due to the
pugs - 1909 - * fix test count
gaal kff, ?
autrijus CGI is failing:
t/basic.t 21 5 23.81% 7 9-12
t/params.t 30 9 30.00% 2 9 11-12 17 24-25 27-28
kungfuftr what url do i use to co again?
crysflame like, CGI.pm?!
autrijus I'll investigate CGI.pm. 18:51
kungfuftr: svn.openfoundry.org/pugs/
kungfuftr ta
gaal we need to teach perlbot that
kungfuftr odd... unrecognised url scheme so svn reports 18:52
autrijus you don't have neon built in?
kungfuftr don't think so
bellecks
autrijus back. 18:54
I'll deal with the 3 unexpected successes.
done. 18:56
making another round of r1910 tests 18:57
Corion r1908 at datenzoo.de/pugs/win2k.html
autrijus weird 18:58
t\builtins\io\io_in_for_loops.t
I thought I have force_todo there
can you run it separately?
it looks like that run has no idea of force_todo
ChangeLog done. 18:59
jabbot pugs - 1910 - * clean up unexpected tests. 19:00
pugs - 1911 - * bump version to 6.2.0; add ChangeLog
autrijus preflight will begin shortly.
Corion I'm currently smoking r1910 - should I stop/restart/retest ?
autrijus no, go ahead
Corion Weird. The force_todo file claims that #26 is todo, but the test output disagrees ... 19:01
kungfuftr right... doing a test run on my 5.3 speddier laptop 19:03
autrijus #26?
Corion Hah. The output is "(TODO for release)", but I think that the harness wants "#TODO reason"
autrijus: In io_in_for_loops.t or whatever it's called.
Hmmm. No, the output is even "# TODO for release". Weird. 19:04
autrijus preflight! 19:07
wagner.elixus.org/~autrijus/tmp/Per...2.0.tar.gz
castaway oooh
autrijus writing to p6c. 19:08
wolverian hmm, pugs has apparently added its Test.pm to my 5.8.4 lib dir. 19:09
autrijus it did that a while ago. 19:10
I believe it has been fixed. is this a current issue?
wolverian probably not.
removing them and retrying.
autrijus thanks and sorry.
wolverian no problem.
Corion r1910 - datenzoo.de/pugs/win2k.html (rsync++) 19:12
gaal Corion, scp -C is good too :)
autrijus Corion: grab preflight and "nmake test" by hand?
Corion autrijus: Willdo
gaal: Does it skip unchanged files? 19:13
gaal no, but it compresses the data.
autrijus 06-dash-version.t
it will fail, that's expected, I'm committing a fix.
it's regex matching against 6.0.x. 19:14
fancy that.
gaal anyone in tests already? i'm still compiling
kungfuftr r1911 here 19:15
Corion nmake test running
gaal btw how bad is it if i make optimized, then svn up a few files, and make w/o optimized? (not happening now, i've started from clean)
Corion (still compiling) # preflight
stevan (me too) # preflight
gaal ah, tests have started. 19:16
Corion t/builtins/io/io_in_for_loops...............ok 19:17
autrijus preflight is expected to have one test file fail (06-dash-version) and nothing else.
Corion ... so it's a problem of the yaml harness that it doesn't recognize the force_todo stuff
autrijus I will release the final version if Corion and stevan confirms this status.
Corion autrijus: Oy! Now I get your 6.0.x comment :) 19:18
autrijus :)
Corion autrijus: But I wanted to avoid the test to check against the version of the currently running pugs.exe :)
stevan testing testing testing ..... 19:19
theorbtwo Should you disable make optimized in the release?
Corion ... A project manager falls from the empire state building. As he passes the third floor, he is overhead to say "so far, so good".
kungfuftr r1911 is clean except for 06-dash-version.t with `make; make test`
jabbot pugs - 1912 - * isa_ok now OKs 19:20
pugs - 1913 - * fix version for .002000
pugs - 1914 - tests chars, but not adequately; see
pugs - 1915 - * oops, wrong regex
pugs - 1916 - * we are not 6.0.x anymore
autrijus theorbtwo: optimized now works.
theorbtwo Oh, OK.
gaal PID still fails here
autrijus for some strange reason
gaal: oh? which version? which test?
gaal: what's your
gaal t/magicals/pid..............................FAILED test 1 # msys
autrijus pugs -e "$?OS"
pugs -e "$?OS.say"
gaal ahem. 19:21
$ pugs -e "$?OS.say"
unexpected "O"
expecting "x", "d", octal digit, "o", "b", digit, fraction, exponent, term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
NonTerm SourcePos "-e" 1 2
autrijus pugs -e '$?OS.say'
;)
Corion magicals/pid.t fails here too. $*PID support is broken on Win32
gaal yeah, msys :) 19:22
Corion I'll demote that test to todo_
theorbtwo Hmmpf, can we get something saner?
gaal t/operators/arith...........................FAILED tests 149-150, 152-153
Corion autrijus: Or do you want to demote tests there?
gaal t/operators/filetest........................FAILED tests 2-3, 5-6, 23, 26
autrijus ok, I added skip.
theorbtwo (Then msys, that is.)
gaal huh? that's sane enough, to2
don't you expect 'cygwin' for $^O when you use that? 19:23
oops $?OS
Corion t/operators/arith: 13/179 skipped, but otherwise OK on Win2k
theorbtwo I expect "cygwin" on cygwin systems, "win32" on pure win32 systems.
gaal well, msys is very much like cygwin.
Corion t/operators/filetest.t fails FAILED tests 2-3, 5-6, 23, 26 19:24
mj41 WinXP infinite loop, IMO:
pugs -e "use File::Spec; say 'begin'; say abs2rel( '\a' , '\b' ); say 'end';"
Corion Likely because -s is still unimplemented on Win32
theorbtwo msys is only a well-known name for those highly familiar with cygwin, no?
stevan mj41: what does the perl5 File::Spec do? 19:25
gaal do we want force_todo to have platform smarts?
or is that the road to insanity?
Corion gaal: I think we should make the platform relevant tests simply todo_
t/pugsrun/06-dash-version...................FAILED tests 1-8
(as expected)
gaal Corion, but then you get unexpected oks on som platforms. 19:26
theorbtwo Um, giving "unexpectedly succeeded" on linux isn't much better then "failed" on win32.
stevan what about a compromise of todo_win32_* test functions
Corion Ah - I forgot about the other platforms, true.
Let's simply make the test smarter then.
gaal stevan, surely *that* is insane :)
todo_freebsd
todo_cray
theorbtwo todo_qnx!
gaal todo_zx_spectrum 19:27
Corion todo_linux_freebsd_but_not_netbsd
stevan todo_commodore_64
kungfuftr how bout as a parameter? for todo
Corion Let's simply add another junction to TODO, which contains the OSes :)
theorbtwo Allo, limbic.
Corion kungfuftr: You keep reading my mind!
autrijus don't do that to 6.2.0 yet :)
nothingmuch damn you Test::Harness!
Limbic_Region salutations theorbtwo
kungfuftr =0)
gaal no!!! skip if $?OS = any<known_bad>
theorbtwo L~R: wagner.elixus.org/~autrijus/tmp/Per...2.0.tar.gz 19:28
Limbic_Region very productive day for me - haven't been following Pugs at all today - so what's new?
gaal whoops, i'm slow :)
Limbic_Region heh - good timing theorbtwo - will have to check it out tomorrow though
stevan Corion: I like the Junction idea
Corion stevan: It's just disguised madness IMO.
mj41 perl -e "use File::Spec::Functions qw(abs2rel); print abs2rel( '\a' , '\b' );"
..\a
autrijus actually, realflight. 19:29
Corion Simply make the tests smart enough to care for their platform
gaal so, summary of failing tests on win32:
t/magicals/pid.t 1 1 100.00% 1
t/operators/arith.t 179 4 2.23% 149-150 152-153
t/operators/filetest.t 30 6 20.00% 2-3 5-6 23 26
t/pugsrun/06-dash-version.t 8 8 100.00% 1-8
kungfuftr r1916 # running tests with `make optimized; make test`
Corion autrijus: Still looking good here.
stevan Corion: madness abounds with Pugs :)
autrijus wagner.elixus.org/~autrijus/Perl6-P...2.0.tar.gz
if you are early into the previous test, please drop it
Corion autrijus: New version?
stevan autrijus: no problems here either
autrijus yes, likely to be final one.
jabbot pugs - 1917 - * pid.t skipped on win32 19:30
pugs - 1918 - * sporadic test fails reported by gaal a
gaal i love it how firefox names the download Perl6-Pugs-7.2.0.tar.gz :)
autrijus :D
Corion On my version, it was:
t/magicals/pid.t 1 1 100.00% 1
t/operators/filetest.t 30 6 20.00% 2-3 5-6 23 26
t/pugsrun/06-dash-version.t 8 8 100.00% 1-8
autrijus ok, I think I covered for all of them.
stevan mj41: what platform are you on?
autrijus as for filetest.t, because we don't have win32-specific forcetodo
I simply disabled all tests for filetest.t on win32.
stevan autrijus: no problems at all with Mac OS X 19:31
autrijus actually, you can just selective rerun that three first
stevan (but the test are not completely done)
Corion autrijus: Good Enough
autrijus pid, filetest, 06dashv
and report back whether they now Works.
Corion autrijus: Will rerun the three failing tests, plus the one of gaal
autrijus there's no arith.t in this release ;)
<- too tired to investigate msys
gaal *giggle*
it's a painful test to debug 19:32
autrijus that too.
autrijus waits for Corion/gaal to report
kungfuftr autrijus: r1916 is completely clean with optimized
autrijus wonderful!
looks like we have a solid release.
boogie server isc.perl.org 19:33
Corion operators/filetest.t crashes for me
gaal i just remembered that at some point, it reported the wrong coordinates to failing tests - but that may have been fixed
autrijus oh wow.
crashes!?
Corion (syntax error: $pugs is undefined)
*undeclared
autrijus argh.
Corion the other two are fine
gaal guess there's no filetest in this release either, huh? :p
Corion ... at least according to prove.pl
autrijus it's just a typo ;)
but hey, why not 19:34
Corion Yeah - pugs -Iblib6\lib reports the same failure :)
autrijus yeah, and it's fixed in trunk
Corion wipes his flight-dir 19:35
stevan is still testing ....
Corion starts another smoke run
kungfuftr filetest.t fails on freebsd
test 6 and 26 19:36
Corion kungfuftr: Likely because of the aforementioned syntax error :)
gaal still compiling, here :/
kungfuftr ah!
Corion kungfuftr: Oh - if it doesn't completely fail, then there is something bad...
lightstep what is `cannot find package plugins'? (on perl Makefile.PL)? no hs-plugins on this system?
autrijus right. that's fine 19:37
theorbtwo Correct. The next line should warn you about that.
autrijus Corion: it appears that it tries to -w pugs
and fails
anyway it's not in the release.
autrijus takes this No Mercy approach
ninereasons filetest.t doesn't fail on my bsd, r1920
stevan filetest.t is ok on my Mac 19:38
theorbtwo We couldn't figure out a real platform-independent way to shut it up... so we didn't, because it's better then hanging on half of our user base during Makefile.PL time.
Corion Maybe umask differences
kungfuftr ninereasons: r1920 still does here
gaal wishes for a haskell ccache 19:39
Corion ccache ?
gaal ah, tests just started
Corion, a hack to make c builds faster
Corion gaal: I think GHC is already pretty smart about compiling ;)
gaal it stores .o files in a cache, keyed by their MD5s
ninereasons I make realclean first, but surely that's not a difference, kungfuftr ?
jabbot pugs - 1919 - * so much for cargo culting 19:40
pugs - 1920 - * cargo cult too
gaal yes, but not when i rm -rf the old preflight dir :)
kungfuftr ninereasons: nope, just tests have changed since my last build
Corion gaal: Aaah ;)
gaal which i kinda want to do you know, as one more measure against paper bags.
kungfuftr ninereasons: yeah, pugs is 1916 here 19:41
stevan *sigh* we have too many tests ...
gaal kff, 1920 is the latest r
Corion stevan: Feel free to reorganize them in a way that exposes new bugs in the early-run tests :)
gaal ...so far so good on msys... 19:42
kungfuftr gaal: i know... but between 1916 and 1920, only tests have changed
stevan Corion: but how will we know where the next bug is ?
Corion Heh - quoting typo in t/magicals/pid.t - double quotes where single quotes should be used :)
nothingmuch Steve_p: darcs pull if you're interested
kungfuftr rebuilds anyway
nothingmuch i haven't merged yet, i'm on a blitz
theorbtwo I suspect the harnesses should shuffle, and run in -j2 mode.
Corion stevan: Ah, that I don't know. But maybe for regressions :))
nothingmuch i'll do it in 3 minutes, when I'm done with one more thing 19:43
gaal 'use less bugs' or 'use bugs <predictable>'
Corion theorbtwo: Hmm - -j2 would be interesting indeed, except that we don't have File::Temp yet
(some of my tests use sleep(), so -j2 would benefit from that)
nothingmuch Corion: temp? what for?
stevan Corion: actually we might want to look into a t/skip_todo file which avoid tests which we absolutely know are TODO
nothingmuch Corion: not only that, there's some of us with multiple CPU boxes
Corion nothingmuch: Race conditions. Currently all tempfiles are hardcoded
nothingmuch eris is 2 cpu
safta is 2 cpu HT 19:44
theorbtwo lilith is hyperthreaded.
Corion nothingmuch: My P4 box pretends to be 2 CPUs as well
nothingmuch Corion: then they should be different per test case
Corion nothingmuch: Which is why I want File::Temp
nothingmuch safta pretends it's 4 cpu
Corion: i mean hard coded diff
or we could do File::Temp
gaal t/operators/filetest fails here, predictably, in my pf2 package...
nothingmuch we could start by making it brutish
autrijus wagner.elixus.org/~autrijus/Perl6-P...2.0.tar.gz 19:45
Corion nothingmuch: Sure, but that will lead to stupid copying and inexplainable errors. That's why I want File::Temp :)
nothingmuch lets assume that the probability of collision with $$+rand+time+rev is ok
autrijus err, sorry: wagner.elixus.org/~autrijus/dist/Pe...2.0.tar.gz
nothingmuch and then we could really do it, once we have a proper stat
Corion nothingmuch: Hah. $*PID is constant on Win32.
theorbtwo assumes those are ~~?
Er, ~.
gaal oh - it's out?
nothingmuch huraah!
Corion nothingmuch: Never, never! use stat! There are (POSIX) calls for Tempfiles, and Win32 has them too (or via POSIX)
I want proper tempfiles :)
autrijus sanity check that dist/ tarball?
make sure the two headache tests are not there, the version etc passes when run individually, etc 19:46
nothingmuch Corion: well, i don't know of the proper ones, i just use File::Temp
autrijus using a built pugs
Corion Ooops - t/operators/arith.t failed due to a syntax error - $pugs undeclared.
nothingmuch i thought it was needed, afterwords
Corion nothingmuch: File::Temp uses the proper ones :)
nothingmuch to make sure the file you created was what you wanted, or something
autrijus Corion: fixed already
nothingmuch well, sure, whatever =)
Corion autrijus++ # fixing tests before I find failures
ingy hola
stevan hey ingy, care to preflight? 19:47
autrijus anyway, I think it's ready to fly.
wagner.elixus.org/~autrijus/dist/Pe...2.0.tar.gz
SIZE (Perl6-Pugs-6.2.0.tar.gz) = 642482
MD5 (Perl6-Pugs-6.2.0.tar.gz) = 8d5438d49db872ffe2394fd4995d335b
stevan autrijus: t/pugsrun/06-dash-version.t 8 8 100.00% 1-8
thats all
Failed 1/203 test scripts, 99.51% okay. 8/4246 subtests failed, 99.81% okay.
autrijus stevan: right, that's with the tmp/ one? 19:48
anyway it's fixed :)
stevan the first one you asked to preflight :)
theorbtwo Hm, the current -dcore-lint error happens both with and without optimized. Reporting it on #haskell.
autrijus yup.
ok, that's fine, posting on p6c now.
nothingmuch 4000... wow
stevan 4246 actually :P
nothingmuch remembers when his tests accounted for more than 10% of the suite 19:49
yeah yeah, i'm not blind
gaal sanity check looks good here
autrijus okay, fly!
jabbot pugs - 1921 - * This be Pugs, version 6.2.0. 19:50
gaal cheers
Corion r1919 looks good her
e
Yayyy
autrijus yay!
Corion Fly my Pugs! Fly!
ninereasons *applause*
castaway cheers
theorbtwo jabbot, nopaste
jabbot theorbtwo: Does talking about this bother you?
stevan whose got the champagne!!! 19:51
ingy pugs++
autrijus and again, our summarizer missed us...
mj41 s/todo_//g foreach test;
r1600 - 1308/4135 subtests failed, 68.37% okay.
r1700 - 1166/4222 subtests failed, 72.38% okay.
r1800 - 1252/4214 subtests failed, 70.29% okay.
r1900 - 893/4240 subtests failed, 78.94% okay.
gaal perlbot nopaste
perlbot Paste your code here and #<channel> will be able to view it: sial.org/pbot/<channel>
stevan watches as Corion's Pug falls several stories to the ground and splats
lightstep so few tests added over time? 19:52
19:52 bradb|out is now known as bradb
stevan I wonder how many commits we had in the past 48 hours? 19:53
kungfuftr can't remember where the churn graph is 19:54
stevan stevan% svn log -r "{2005-04-11}:HEAD" | grep '^r[0-9]'| cut -d'|' -f2 | sort | uniq -c | sort
1 Ovid
1 asavige
1 cognominal
1 kcwu
1 luqui
1 tobez
rindolf autrijus: I see you've added pugs to CIA.
stevan 1 xern
2 ingy
4 gaal
4 ninereasons 19:55
4 theorbtwo
5 corion
5 dvergin
5 gcomnz
16 Stevan
16 iblech
102 autrijus
that there is a lot of commits
autrijus rindolf: I did not. maybe someone else did.
lightstep too bad the last summary was weekly, not fortnightly
autrijus that's fine :) 19:56
posted to p6c.
same post posted to journal.
feel free to cross post the good news elsewhere. 19:57
anywhere is fine... even slashdot :)
now... I really should sleep.
thanks everybody. :))))
autrijus waves &
stevan nite autrijus
theorbtwo G'night, autrijus. 19:58
gaal good night!
kungfuftr full clean build here anyways 19:59
=0) 20:00
Corion Slashdot accepts every crap today. Maybe we should offer them to pay if they mention Pugs every week :)
theorbtwo ponders what useful he could do with the haskell bindings, but without user-visible objects. 20:02
I wonder if I can do objects on the level I'd want already.
Ugh. 20:07
./pugs -e'require_haskell("SHA1__0_0_1.o"); say sha1("")' 20:08
pugs: internal error: stg_ap_p_ret Please report this as a bug to [email@hidden.address] or www.sourceforge.net/projects/ghc/
kungfuftr gets pugs running under svk 20:09
ninereasons why did your build succeed this time, I wonder, kungfuftr ? 20:15
kungfuftr ninereasons: no idea, is very odd
bad cleanup from the test first time round? 20:16
nothingmuch oddd
very very odd
ninereasons btw, kungfuftr , have you succeeded in getting plugins installed (sorry to be a broken record about this) 20:17
gaal 9r: did you 'make register'? 20:18
ninereasons it won't build for me, gaal
gaal ah
what platgform?
ninereasons it doesn't seem to like what it gets from autoconf
freebsd 4.11
gaal it won'r build for me on windows either :( but on linux it does work. 20:19
kungfuftr ninereasons: nope, there's no freebsd port for it, so don't get on teh server i'm afraid
ninereasons (autoconf259 actually - maybe that's the issue)
yes, gaal, debian built and installed as did slackware
gaal slackware! my first distro 20:20
ninereasons i'm sentimentally attached to slackware.
gaal remembers going off to a distant university as a kid to download tens of 5 1/4" diskettes
ninereasons but debian makes me go "wow" 20:21
so many cool ideas.
gaal the coolest thing about debian is that mostly, it just works
ninereasons yes. I was introduced to it through knoppix
just works. 20:22
theorbtwo A close second is that when it doesn't just work, it doesn't get in your way when trying to fix it.
gaal true dat
ninereasons funny perl installation, though :) 20:23
Aankhen`` G'night.
gaal 9r: apt-get install dh-make-perl, if you have more than one box.
ninereasons noted, gaal. 20:24
gaal though nowadays i just say to hell with it and use cpanplus
nothingmuch mutt is slow
nothingmuch needs to repatch for headercache
kungfuftr ninereasons: come across hs-plugins in ports anywhere? 20:25
ninereasons no, kungfuftr .
kungfuftr, it's strange though, because 'don' develops on *bsd, I'm told
kungfuftr anyone updated MANFIFEST at all? 20:26
stevan kungfuftr: its part of the release process 20:28
kungfuftr ah, k 20:30
mj41 g'night
Corion theorbtwo: You could implement File::Temp via your external bindings, or SQLite - both could be object-less exports, simply passing through the handles/strings (I think) 20:42
nothingmuch Steve_p: darcs pull 20:47
kungfuftr: did you see Test::TAP::Model?
kungfuftr nothingmuch: nope... where? 20:48
nothingmuch darcs get nothingmuch.woobling.org/Test-Tap-Model
kungfuftr don't use darcs
nothingmuch wget -r then
kungfuftr nothingmuch: looks intresting 20:50
nothingmuch will now do Test::TAP::Visualize::HTMLMatrix 20:51
based on stevan's thing
btw, stevan, gaal, theorbtwo, autrijus, ingy et al: do you mind if i just put copyright "The Pugs Hackers"?
IMHO this code belongs to us all, and not only to us who worked on it
kungfuftr nothingmuch: i've had some ideas in how to improve the html generation tbh 20:52
nothingmuch i'm going to use a real templating thingy
for starters
Corion Pe-tal ! Pe-tal ! Pe-tal ! (although I admit that HTML::Template is easier done than Petal :) ) 20:53
nothingmuch yeah, i'm for petal
especially since Test::TAP::Model has a nice OOP interface on top of the hash structures 20:54
Corion "easier done" in Pugs
theorbtwo I'd prefer that copyright be stated as belonging to something with some legal existance: a finite and enumerated list of people.
lightstep to whom is pugs copyrighted?
theorbtwo Autrijus.
nothingmuch theorbtwo: perl foundation? autrijus? people?
i'm not a part of the perl foundation
kungfuftr pugscode.org ? 20:55
theorbtwo Er, pugs is copyright autrijus.
Corion BTW, Haskell School of Expression is one of the recommended books, right?
nothingmuch in fact, i feel somewhat alienated by it, for this hunk of code
theorbtwo You do, nm?
ninereasons my copy of "The Haskel School of Expression" is signed by the author
(although, not addressed to me)
nothingmuch a bit
personally i don't think they are actively helping
luqui and pmichaud do
but aside from that there has been no active input except on their level
that is, we do stuff
then we run into trouble 20:56
and we go ask
Corion Hmmm - the reviews say that it's a good second book ... Does it address the type system/monads in a good way ?
nothingmuch: Ah, who cares about TPF?
nothingmuch Corion: exactly
Corion ... or do they hold the (c)opyright on Pugs?
nothingmuch i think they don't
autrijus does
Corion nothingmuch: They didn't as much drop the ball, they saw it falling and never moved in the general direction IMO.
kungfuftr nothingmuch: "Copyright pugscode.org"?
theorbtwo "Pugs is Copyright 2005 by Autrijus Tang. All Rights Reserved." (From the README.) 20:57
Corion nothingmuch: Ah, he got TPF grant(s), so he feels more attached to them I think
nothingmuch theorbtwo: what should our code be, then?
theorbtwo Copyright all of us.
nothingmuch Corion: i spent work time today on this stuff, because work is interested
theorbtwo Anyone who has copyrighted stuff in it.
nothingmuch that doesn't mean the code is copyright hyperroll
theorbtwo Er, copyrightable.
clkao 5/win25
nothingmuch because i only refactored
oh my
does anyone have the capacity to svn blame | awk > copyright efficiently? 20:58
Corion Who cares about the copyright if the license is BSD/artistic ? I don't mind leaving the copyright with whoever as long as I can use/modify/distribute the code as I wish
nothingmuch Corion: he who holds the copyright decides the license
theorbtwo nm: That's one reason I try to keep track of AUTHORS.
ninereasons it's a very Windows friendly book, Corion: I assume you approve
theorbtwo Also, he who has copyright has standing to sue for violations of the license.
Corion nothingmuch: The license already is BSD/Artistic/dual, so I'm OK with that 20:59
ninereasons: Ah, well - I don't mind windows-friendly books :)
nothingmuch svn blame | cut -d: -f1 | sort | uniq | <map via AUTHORS>?
lightstep nothingmuch, it's gpl, so it can never be non-gpl
nothingmuch theorbtwo: i'll do it later, methinks
lightstep: i don't really know the details
but the point is, that copyrights are an important attribution of control 21:00
it's not that i don't trust that TPF will use it correctly,
but this code is ours and should be marked as such
theorbtwo: what about contributors whose code has been indented, for example?
theorbtwo nm: Don't use blame, use log. 21:01
nothingmuch good point
odd 21:02
svn log -q util/testgraph.pl | awk '/^r/ { print $3}' | sort | uniq
doesn't include stevan
lightstep "You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License." (from LICENSE/GPL) 21:03
theorbtwo nm: Did you expect it to?
nothingmuch IIRC he wrote the original
theorbtwo I think I did.
But looking at the svn log out, all the patches are very short. 21:04
nothingmuch ah, sorry =)
theorbtwo++
i was sure it was his all along =)
theorbtwo Oh, the "lines" thing is the number of lines in the changelog entry, not the number of lines in the patch. 21:05
pjcj ! 21:06
... ww
nothingmuch AUTHORS doesn't contain the rt.openfoundry.org id, does it?
egrep -i 'autrijus|corion|gaal|nothingmuch|putter|rootmj|theorbtwo' AUTHORS
even worse without -i
pjcj quit
Corion pjcj: You're not in the vi window! :) 21:07
theorbtwo I try to have it contain it, or something pretty obvious from it.
nothingmuch theorbtwo: are you willing to handle patching the copyright stuff?
Corion: that's not a vi command
theorbtwo is going to have to pick up Yet Another VCS for this, isn't he?
Tomorrow.
pjcj Corion: no - I was trying to install Apache::Filter ;-)
nothingmuch theorbtwo: or wget -r 21:08
and patch
theorbtwo Nah, I can pick up darcs no problem. 21:09
debian++
nothingmuch yeah, it's easy =)
Corion goes to bed. Have a good next 8 hours! I will be watching you via the logs :)
theorbtwo 5 Stevan
5 autrijus
5 putter
5 rootmj
10 corion
20 kungfuftr
30 nothingmuch
65 theorbtwo
70 gaal
theorbtwo goes to a different bed. 21:10
kungfuftr theorbtwo: wassat?
theorbtwo Number of checkins to the util testy stuff. 21:11
kungfuftr ah
nothingmuch oops
kungfuftr: darcs pull
i was pushing to the wrong repo
theorbtwo I should probably check for checkins to the old util/categorize_tests.pl before it got renamed over.
nothingmuch this has docs, and tests, and whatnot
doesn't it retain history? 21:12
theorbtwo It does.
nothingmuch so...?
theorbtwo But the file isn't there anymore, so the quick stats I just gathered don't include it.
I'll check more carefully later. 21:13
nothingmuch oh well
theorbtwo & # really. Monitor off in 3... 2... 1...
nothingmuch perl_subversion: 0 21:14
pugs_versnum: 6.2.0
so much for releng ;-)
jabbot pugs - 1922 - unTODOed test - worksforme on Win32, ple 21:20
Juerd Argh 21:31
After two days I find out why I can't svn up pleac anymore. It's s/pleac/cookbook/ now...
gcom yeah, fortunately i noticed that somehow before trying too many times 21:32
stevan nothingmuch: ping 21:46
nothingmuch pong
stevan the original catalog tests was mine
but I showed it to you and you ran with it 21:47
nothingmuch that i remember
i didn't remember testgraph.pl
btw, want to SEE?
stevan after that all i did was contribute ideas and such
yeah testgraph.pl I did the original HTML layout idea
but again,.. you all ran with it
(which is a good thing) 21:48
nothingmuch: I stay away from SEE
it tries to take over all my files and I cant seem to stop it
TextMate is my new favorite editor
nothingmuch somehow gets along with it
vim vim vim!
vim for the masses!
vim with molasses!
stevan prefers to have a GUI 21:49
nothingmuch gvim =) 21:50
stevan to be honest (and I know this is blasphemy in these halls) I dont like vi and any vi descendent
nothingmuch textmate looks pretty funky
stevan (modal interfaces)--
nothingmuch but i'm really hooked bad on vim
stevan: people who think like vi think like vi
that's all there's to it
i get along with it because i think modally 21:51
anyway, you're more than welcome to work with my coe
amend, or whatever
stevan nothingmuch: I would like too,.. but I am thinking I need to pay some attention to $work
nothingmuch beh!
stevan I have been doing the bare min for about a month now :)
nothingmuch hah 21:52
stevan luckily my boss is cool with all this stuff
but I have another project starting in a week or so,.. and I need to prepare for it
stevan is dreading going back to perl 5 land in some ways 21:53
nothingmuch i always thought you were the boss at II
stevan nothingmuch: LOL
nope
nothingmuch =)
stevan we are only 3 people, and we work pretty independently
so in a way I am my own boss
but its not my company
nothingmuch oh well =) 21:54
stevan nothingmuch: even if I was,.. i still gotta pay the bills :)
nothingmuch bills shmills
stevan unlike autrijus my client is not willing to run perl5 yet :) 21:55
nothingmuch pugs needs you
stevan :)
have no fear,.. I will not go anywhere
you couldn't get rid of me if you tried
nothingmuch: hey,.. have you figured out how you are getting to YAPC::NA? 21:56
nothingmuch no $$$
so no, i haven't
stevan hmmm, well if you can manage the plane cost,.. you can stay with me free 21:57
nothingmuch plane will be around $700 round trip
stevan you will just have to hide in the closet when the maid comes in
try and get $work to pay for it 21:58
tell em its vital to the business model
nothingmuch work is not in great financial conds
and they know it ain't ;)
again, small co
stevan what about mom and dad? you got a birthday coming up or something?
nothingmuch i'm richer than them =) 21:59
stevan nothingmuch: I am running out of ideas here :) 22:00
we are back to selling your body parts
nothingmuch damn 22:02
YAML is not being nice
22:07 metaperl__ is now known as metaperl_
jabbot pugs - 1923 - More pleac^Wcookbook comments, XXX marke 22:09
nothingmuch it calls the sub in scalar context! 22:17
ack! 22:18
run away run away!
stevan off for dinner,... & 22:26
jabbot pugs - 1924 - test a few escaped characters in strings 22:29
stevan nothingmuch: ping 22:55
jabbot pugs - 1925 - test stringified array 22:59
stevan nothingmuch: check your email 23:01
hey Ovid
bye ovid 23:02
Ovid Hi stevan
Bye steven
stevan goes back upstairs for dinner :)
Ovid s/steven/stevan/
stevan Ovid: you are not the first person to mess up the spelling and surely not the last :)
gcom Juerd: ping 23:07
Ovid: you around? 23:11
Ovid Yes, but working. 23:12
gcom k, i want to hit you up later to talk about pod in Cookbook examples
obviously it would have to be somewhat clean-room to avoid copyright violation
not even sure if it's wanted/needed though 23:13
i'll probably be online later or tomorrow though, no rush on that one
Ovid Good point. I was wondering about this issue. Should we touch base with O'Reilly?
gcom yeah, i don't know, i feel like venturing toward filling out the pods, cause questions coming up in comments seem to be asking for it 23:14
i don't own a perl cookbook though, so perhaps it's safe for me to venture forth (used to own one a long time ago)
Ovid I know a few folks who work there. Let me shoot an email and see what they say. 23:15
gcom cool
thanks
Ovid np
gcom later then, i gotta get back to work too :) 23:16
Ovid Email sent. 23:17
Limbic_Region hey Ovid - what's shaking? 23:19
Ovid Not much. Working right now.
Limbic_Region forgets you are 3 hours behind 23:20
Ovid works late, too.
Though I would prefer to be 5 hours ahead ...
shapr Come to Sweden! 23:21
Ovid God, I'd love to. So close to all of my family, but so far from a job :(
shapr Your family is in Sweden? 23:22
g'day mugwump
mugwump morning!
mugwump is slowly making it back into his home TZ
Ovid No. My dad lives in Germany, one brother's in London and then another brother and sister are a couple of hours north of London. Essentially, all of the family I have any serious contact with live on that side of the pond. 23:23
That's one of the reasons I miss living in Amsterdam. Family.
shapr Wow, your family migrates more than mine.
Ovid No. My father just slept around a lot. My sister in the UK is three weeks younger than I am. 23:24
shapr oh
Ovid :)
shapr I have no idea how to respond to that statement.
Ovid No problem. At this point, it's rather amusing.
mugwump does it run in the family? ;-)
Ovid (though I doubt my mother thought so at the time)
Of course not! I tell all of my women that I'm seeing others :) 23:25
shapr laughs
mugwump I don't like to volunteer the information. But if they ask, well, they get what they deserve for their suspicion 23:29
shapr I'm just monogamous. Having tried polyamory, I decided monogamy was for me. 23:30
mugwump It's certainly a more pragmatic choice 23:31
nice combigism btw (polyarmory), you should add that to the dicshunary.com 23:32
shapr snickers 23:33
A telepathic journal is totally mind-blogging!
Ovid Heh
Odin- shapr: Please remain where you are. The Pun Police are on their way. 23:36
shapr sniggers
The Punisher approaches! 23:37
Odin- notifies the SWAT.
Limbic_Region is too lazy for more than 1 woman, which suits his wife just fine 23:40
Ovid My best friend and I have been in a relationship with the same woman for years. Kind of freaks a few people out. 23:42
shapr question is, is your best friend that woman?
Ovid Heh, no.
Limbic_Region I was in a situation where one gf told me she didn't care if I was with the other gf as long as I was with her and the other gf said the same thing about the first 23:43
23:43 khisanth_ is now known as Khisanth
Limbic_Region everyman's dream right - not mine 23:44
too much work/hassle
shapr That's most of my problem with polyamory as well.
polylogistics make category theory look so simple.
Limbic_Region Ovid - doesn't really freak me out but makes me wonder who has what plans for the future 23:45
Ovid Limbic_Region: not much I can say about that for privacy reasons, but she's decided to move in with him and it looks like they'll be buying a house together. It's a good thing. 23:46
Limbic_Region *shrug* - who am I to say anything as long as all parties are happy
Odin- notes that while you're complaining of having too many women, others complain of lack of them. 23:47
shapr Yeah, I agree with that.
Odin- Why is it that nobody seems to hit the right balance? ;>
Ovid I'm hardly complaining about too many women, actually. I was kidding about my above comment about "telling them". I have one woman with whom I am serious, but she's now my best friend's primary partner so I don't see as much of her as I used to. Other than that, I date. 23:48
justatheory has too many woman. 23:49
Ovid Well, you're about to in a few weeks. 23:50
justatheory Yow! So true.
s/woman/women/
Ovid notes that no one would have any idea of the context of that comment.
Odin- Indeed.
justatheory Let them guess. :-)
Ovid :) 23:51
shapr hiya mark 23:58

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!