pugscode.org <<Overview Journal Logs>> | r1773/winxp: 587/4256 smoke/win2k (occasional): xrl.us/fqum (10/4279) | pugs.kwiki.org
Set by Corion on 13 April 2005.
kungfuftr yay! 02:17
bah, where's nothingmuch when ye want him 02:47
crysflame 08:47 -!- nothingmuch [~nothingmu@212.143.92.226] has quit [] 04:21
21:22 now
autrijus hi. 05:24
wolverian: Array.isa("List").
$a = <x y>; $a.ref; # this should be "Array". 05:25
so that's a bug.
Array is a mutable IType; List is an immutable IType, that's all
Hash is also mutable; Array and Hash represents different mutable views on List.
so they both .isa lis
list
wolverian right. thanks. 05:26
autrijus since that's a bug... please add a test :D
wolverian sure. by the way, is <a b c>.kv failing a todo or not?
I'm unclear on the todoity of things.
autrijus wolverian: <a b c> is a List not an Array 05:27
so according to Rod's draft of s29 it shall fail.
wolverian lists don't have .kv? hm.
autrijus # www.rodadams.net/Perl/S29.html#perl6::lists
wolverian right. that makes sense, if Hash.isa(List)
autrijus I see no .kv there.
if you'd like to have it... tell Rod 05:28
wolverian it'd be logical, since a list can be indexed similarly like an array
autrijus maybe. in my mind though, list isn't keyed 05:29
although you can certainly write <a b c>[0]
wolverian that's what I'm basing my feeling on.
autrijus you can also write
<a b c d>{'a'}
wolverian hmm. 05:30
autrijus and neither is really keyed to the list
it's just dynamic polymorphic interface
wolverian in that case, .kv makes no sense.
I'll remove the todo. :)
I mean, the test.
autrijus =)
wolverian where should the .isa tests go?
autrijus I have no idea. improvise :) 05:32
I need to run now to $wr0k
wolverian will do!
autrijus see you in a bit
wolverian have fun.
t/builtins/list/* look like array tests :) 05:33
autrijus refactor mercilessly then. :)
autrijus waves &
mj $a = <x y>; $a.ref; # this should be "Array". 05:47
but <a b c> is a List not an Array, right?
wolverian I don't know. autrijus, help? :) 05:48
mj autrijus said that, IMO, I have to study :-) before push @larry, $mj. 05:50
wolverian svn is giving me hard time. oh well, have to learn this sometime. 05:52
gaal wolverian, still there? 06:25
wolverian yes.
gaal there's an accidental test for the List tihng in undef.t already
(if's failing, of course) 06:26
wolverian argh.
gaal no no, it's okay
it's all right if we happen to have duplicate tests
wolverian right. didn't think to look there :)
gaal in undef.t it's just a sanity check of some scaffolding before the "real" test
my gut feeling is that <a b>.ref should be List, but i mat be wrong! 06:28
hey nm
brb, need coffee
wolverian yes, it should be. :)
nothingmuch beh
these antihistamines are really fucking me up
wolverian my computer has suddenly decided to get cranky. X doesn't want to use DRI and my sound doesn't work :()
s,),, # argh. 06:29
nothingmuch ugg
not good news
i hate fixing DRI
wolverian it worked on the last reboot..
nothingmuch in fact, i hate fixing it so much that i only ever finished doing it once
cvs-dri on mach-64 ati card
wolverian and this is ubuntu, I haven't had to touch any X related configuration file at all
nothingmuch 3 years ago
wolverian and it just worked! but now it suddenly stopped working. bleh.
nothingmuch ubuntu sounds funky 06:30
wolverian it is nice.
nothingmuch thinks it might be good for workplace
people will need a few linux workstations
wolverian hmm. so. sub foo (Array $foo) { ... } foo <a b c>; # works? 06:35
I could try.
doesn't. 06:36
06:39 castaway_ is now known as castaway
castaway mornin 06:40
nothingmuch morning castaway 06:42
castaway hey..
did I kill SEE? :)
wolverian relogging to x fixed DRI and sound. hmph. 06:43
nothingmuch castaway: maybe, i dunno 06:48
castaway It's not there anymore, anyhoo 06:49
nothingmuch i'll try to open it
beh
someone rebooted
no windowserver session
sowwy 06:50
castaway O
Ok
hoowa why pugs doesn't support .last method? 07:01
rindolf nothingmuch: ahoy there, matey! 07:23
nothingmuch ahoy 07:24
how may i be of assistance dear sire?
wolverian hrm. does perl6 have instance method references? 07:39
theorbtwo wolverian: Doesn't need something specific for them. 07:49
They're just a specific case of .assuming.
wolverian hmm, good point. that's a tad longer to type, though. :) 07:51
masak what's the syntax for changing an array in pugs?
i'm trying to do this (perl 5 code):
$a = [(1,2,3),(4,5,0)];
$a->[1][-1] = 6;
but it doesn't work
nothingmuch that's not p5, methinks 07:52
masak hm
maybe i should try it in p5 first
wolverian my pugs still doesn't like nested structures.
masak it doesn't?
nothingmuch it will make @$a = (1, 2, 3, 4, 5, 0);
which is not what you meant, i think 07:53
masak not what i meant
i want to use nesting
nothingmuch try using a greater list comma:
$a = [ 1, 2, 3 ; 4, 5, 0 ];
masak ah, of coruse
thx
nothingmuch i don't know if it will work 07:54
wolverian it doesn't here.
masak not here either
nothingmuch ok 07:55
masak i'm trying to write tic-tac-toe
i could do it without nesting
nothingmuch so try $a [ [ 1 .. 3 ], [ 1, 2, 0 ] ]
masak ok
nothingmuch tic tac toe would be wonderful bound
$diag[0] := $matrix[0][0] 07:56
$diag[1] := $matrix[1][1]
$diag[2] := $matrix[2][2];
then you just iterate @diags, @rows, @columns
and then see if they are all the same
wolverian iterate $diag, you mean :)
nothingmuch no,@diags is ($diag1, $diag2)
wolverian sure! 07:57
nothingmuch if @diag[0] >>eq<< 'x'
wolverian been up all night, so I think it's time to sleep. it's only 11am.
masak i'm trying to modify the matrix by
$matrix[1][1] = 3
but it says
"Can't modify constant item"
nothingmuch that's odd 07:58
we ought to tell autrijus
it sounds more like binding complexity and context issues than subscripting or multidim
masak always when i test things i turn up oddness :)
nothingmuch that means you're a good tester, masak 07:59
masak you'd better keep me away from testing the vital parts of the code :)
nothingmuch btw, can you commit a test for that case?
masak ok
nothingmuch why?
we want the bugs to shine =)
masak it's more like i cause bugs by finding them. quantum physics :)
nothingmuch heh 08:00
masak i would be a good tester if i were actively looking for bugs
right now i'm only trying to write a game
ingy hola 08:01
masak hi ingy
nothingmuch hola ingy
i need releng advice
testgraph.css, template.html
where should they go?
ingy I am reading about doxygen and getting good ideas for Perldoc
nothingmuch doxygen is rather nice
ingy CRAZY IDEAS!! 08:02
nothingmuch but smart comments are not %POD savvy
ingy sorry
nothingmuch: they could be
nothingmuch then they should be =pod foo
not comments
ingy that's why I'm meeting with Larry in Toronto :)
nothingmuch which raises the issue for doc on the tail of a line again 08:03
masak this should work, right?
-bash-3.00$ pugs -e 'my $a = [[2]]; $a[0][0] = 6; say $a'
Can't modify constant item
Val VUndef
ingy damn i think today was one of the best days of my life 08:07
masak why?
ingy it was just a good one
masak :)
nothingmuch details! 08:08
ingy a lot of cool things happened
nothingmuch masak: in theory
masak: commit a test =)
masak ingy: you should have best days of your life more often, i recommend it
nothingmuch: ok
ingy masak: I recommend it too
nothingmuch i think it's not copying 2, since it's not directly assigned 08:09
rindolf nothingmuch: what do you think of better-scm.berlios.de/ ? 08:10
ingy nothingmuch: what do you think about comments that get preprocessed into regular pod (but with smarts based on position etc) just be before compilation?
nothingmuch where's darcs?
rindolf ingy: hi there.
ingy hi schlomi
nothingmuch ingy: i think comments should be accessible either way
rindolf ingy: it's "Shlomi" (without a C)
nothingmuch you don't want to know what Devel::FIXME does 08:11
ingy ah
nothingmuch but i'll tell you anyway
ingy ha
nothingmuch it puts a subref in @INC
sets a local lock
calls require again
ingy cool
nothingmuch this time the hook fails since the lock is in effect
ingy yep
nothingmuch the "normal" @INC finds the module
and then you go back to the callback
where you can find it's file in %INC 08:12
so it opens that, and looks for comments with a regex
which is really silly, since it could all just be done with a well defined hook for "new module was required"
and the comment data tree exposed
parsing perl with a regex is stupid
i could have done it with source filters, or at least Filter::Simple's regex 08:13
but it just seemed like too much work, and getting a fake FIXME from a string that contains "# FIXME foo" is not that bAD
ingy :)
masak i have a problem with my test, it aborts on the nested change attempt $a[0][0] = 6; 08:14
nothingmuch rindolf: why isn't darcs listed?
because arch is typically tla
masak and i even put it in an eval {}
nothingmuch or bazaar/arx? or maybe they're under arch?
masak: not eval {}
masak ok
rindolf nothingmuch: because it wasn't added yet. I have to add it.
nothingmuch: bazaar/bazaar-ng and arx fall under the Arch category.
nothingmuch my $a = [[2]]; eval_is('$a[0][0] = 6; $a[0][0]', 6);
ingy I lost 15 lbs in 8 days 08:15
nothingmuch i don'y see tla under there
oh my
do you have cancer or something?
ingy I just need to lose 15 more
masak nothingmuch: thx
nothingmuch where would you put svk?
ingy only 8 more days!
rindolf Why does everyone and his mother start their own VCS project?
nothingmuch why are you trying to lose 30 pounds in 16 days?
rindolf nothingmuch: it's in the comparison for time being.
nothingmuch rindolf: darwinism is one of the side effects of opensourcce 08:16
ingy because I want to to weigh under 190
nothingmuch how much is 190?
ingy 94 or so?
rindolf perlbot: 190 lbs = kg?
perlbot: 190 lbs
nothingmuch hah
perlbot know too much 08:17
ingy perlbot: change 190 lbs to kg
rindolf Barbaric Americans. Pounds...
perlbot: convert 190 lbs to kg
nothingmuch rindolf: i also think that the front page should have:
ingy pounds rindolf into submission
nothingmuch FREE (speech|beer|NA)
distributed/centralized 08:18
those are the top level distinctions i make
rindolf nothingmuch: where should it have it?
ingy 2.2 lbs = 1 kg I think
nothingmuch right under the name, maybe
ingy perlbot: 190 / 2.2
rindolf 1lb = 0.46 kg or so.
perlbot: math 190 / 2.2
perlbot 86.3636363636364
pjcj nothingmuch: did you try using CORE::GLOBAL::require ? 08:19
masak perlbot++
nothingmuch btw, if you're listing bitkeeper, why not p4 and vss? p4 is free (Beer) for a two user, two machine license for opensource devel, i think
pjcj: yes, it looked easier to implement a hook in @INC than to reimplement require
because it is sort of weird
accepting barewords and such
rindolf nothingmuch: I'm not listing VSS because it's even lamer than CVS.
nothingmuch rindolf: so? people should know to be weary of it
rindolf nothingmuch: As for perforce, maybe I should add it.
ingy I should patent my new diet
rindolf nothingmuch: I'm not listing CVS either. 08:20
nothingmuch ingy: what's your new diet?
rindolf: but the page says "cvs is bad use this:"
wolverian svk! darcs! bazaar! bazaar-ng!
ingy nothingmuch takes the bait
well...
you go to Taiwan
come home
nothingmuch i'm just suggesting "cvs is bad, use this, but don't use this, it's as bad as cvs"
ingy $gf dumps you
nothingmuch do you get really fat in Taiwan?
from all that beer?
ah, the last step is taken care of 08:21
ingy no I lost a pound maybe in Taiwan
nothingmuch quite some time ago
ingy then you sleep a lot and don't eat
nothingmuch can't sleep too much
ingy and then you start losing weight
and you say cool!
masak ingy: $gf dump you because you lost a pound in taiwan?
nothingmuch makes me depressed, but furthermore, it makes me feel really bad
ingy and then you eat like a bowl of rice a day 08:22
nothingmuch rindolf: also, most VCSs should, IMHO have their major strong points and bad points listed
ingy masak: no $gf dumped me because she's an idiot
(of course)
rindolf nothingmuch: possibly. Have you looked at the comparison yet?
nothingmuch arch: capable, complete, GPLed, tla is complicated, arx, bazaar are new
i have
masak ingy: of course 08:23
nothingmuch monotone: capable, new, crypgraphic, doesn't use gpg for crypto, slow
rindolf A page with a VCS systems to avoid would actually be a good idea.
ingy but after a couple days I just want to go up to $gf and get down on my knees and...
rindolf nothingmuch: and weird hash values as versions.
ingy thank her passionately for leaving! 08:24
nothingmuch well, i don't really know monotone, so i can't say more
rindolf ingy: :-)
nothingmuch ingy: so why are you sleeping a lot?
ingy life is like 100 times better
nothingmuch: because when you barely eat you need more sleep
nothingmuch ah, i thought it was the other way around 08:25
masak ingy++
nothingmuch when you sleep too much, you barely eat
masak i fasted for nine days a couple of months ago
ingy yeah that too
masak it was nice
ingy masak++
nothingmuch the antihistamines are having that effect on me, now
ingy yeah, it's totally doable
nothingmuch sleep too much
so i eat about 40%
less
but then again, i don't burn all that food either 08:26
ingy also mugwump told me i was fat in Taiwan
which is about the only hing he said to me that really made sense
nothingmuch heh
mj 21/4310 subtests failed, 99.51% okay.
masak i've committed a test, how long before the perlbot thingy announces it? 08:27
ingy tommorrow I have a bike race though
I will likely need to eat a little more
rindolf ingy: motorbike, or bicycle?
masak ingy: sounds like a good idea
ingy bicycle
mj r1953 - 973/4310 subtests failed, 77.42% okay.
rindolf ingy: ah good. /me likes biking.
ingy I'm the fattest guy in the peleton 08:28
once i get below 190 I'll be flying though
mj i hate new miranda irc :-(
ingy 190 is about my optimal race weight
rindolf ingy: I never pictured you as a fat guy. 08:29
masak ingy: i think i'll try your rice/no $gf diet
nothingmuch aren't you losing a crapload of muscle tissue with this amazing new diet?
mj s/todo_// ... r1953 - 973/4310 subtests failed, 77.42% okay.
ingy masak: the secret is to pretend you are depressed
masak ingy: how?
ingy after you start losing weight
you'll feel good
masak i'm so happy
ingy about yourself
so then you have to fake it 08:30
nothingmuch mj: todo_ how much is t/oo?
ingy that it really sucks that you got dumped
masak ingy: hm. sounds dangerous
what if you start to believe it?
ingy I get on the scale about 25 times a day 08:31
masak: well I have a goal weight
Going for 160 would be dangerous
nothingmuch ingy++; # amusing
masak ingy: so after you reach your goal weight you'll stop simulating depression? 08:32
castaway doesnt think pretending that is such a useful idea
ingy anyway I'm hungry, better catch some zzzzzzzzz
nothingmuch hah
rindolf castaway++
perlbot: karma castaway
perlbot Karma for castaway: 5
jabbot rindolf: castaway has neutral karma
masak ingy: i want to eat rice but i don't want to sacrifice happiness
rindolf jabbot: I didn't ask you.
jabbot rindolf: Oh, I?
castaway Just eat what you want to eat, and be who you are :) 08:33
ingy but masak I just had the best day of my life
it's not all that bad ;)
nothingmuch ... fat, and ugly, and unattractive! yah! /me can vouch for that
rindolf ingy: eat according to the stomach and not according to the eyes.
ingy nightnight all
castaway night ingy
masak ingy: the best depressed day of your life? :)
nothingmuch i'm happy, /and/ chubby
rindolf ingy: bye
masak night ingy
rindolf nothingmuch: you're not fat.
nothingmuch rindolf: not fat, like obese 08:34
but the mass that I do have is not very healthy mass
mj wiki.kn.vutbr.cz/mj/index.cgi?Pugs%...0-%20tests and links
nothingmuch has no money to buy the bike he wants
masak how long before the bot thingy announces my patch? how does it work? do i have to do something?
nothingmuch but that shall be rectified in due time
rindolf nothingmuch: don't you have two jobs?
nothingmuch one, now
but i just bought a double bass
castaway it does it by itself, when it notices
nothingmuch and before that i bought a camera
rindolf nothingmuch: I see.
nothingmuch and before that a laptop
and a computer 08:35
rindolf nothingmuch: and an iPod.
nothingmuch and ....
castaway hugs nothingmuch
nothingmuch an ipod was way before
i painted a house for that
rindolf nothingmuch: do you have a savings account?
Or an insurance?
nothingmuch savings account was dipped into, to cover the bass
and what would insurance help me?
rindolf nothingmuch: no, I meant, in case something bad happens to you. 08:36
So you won't spend all your money on accessories.
nothingmuch what should i insure?
the bass should be insured, as soon as the forms are faxed over 08:37
rindolf nothingmuch: is it a bass guitar?
castaway (savings, what are those?)
nothingmuch rindolf: bass guitars aren't that expensive
rindolf is listening to Frankie Goes to Hollywood - Relax 08:38
nothingmuch xrl.us/frxn <-- double bass
rindolf nothingmuch: oh you mean a contra-bass. One of those huge violins.
nothingmuch it's not in the violin family, it's a viol (traditionally) 08:39
and it's called contra bass in german and hebrew
but upright bass or double bass in english
but yes, that's what i mean =)
rindolf nothingmuch: why do you need such an instrument for?
nothingmuch i don't know, it's just pretty
you, like to keep in your room
it takes up about 20% of the volume 08:40
but i can live with that
rindolf nothingmuch: you don't play it?
nothingmuch ofcourse i do, don't be silly
rindolf Heh.
boogie hi to all! is it possible to make a wish related to pugs here? :)
nothingmuch boogie: ofcourse
kungfuftr nothingmuch: yullo! 08:41
masak "roses are flowers
viols are bass
what is it good for?
put it in your ear"
nothingmuch hola kungfuftr
masak: seek help =)
kungfuftr nothingmuch: i have a new template and css file for you
masak :)
boogie i'm interested in pugs on the web, and "read" support is needed fo it. ;)
nothingmuch darcs send, kungfuftr )
kungfuftr nothingmuch: no darcs here
nothingmuch s/\)$/=)/
aaah! no darcs!
ok, so send a unidiff, 08:42
or nopaste it somewhere
or, i dunno
kungfuftr kungfuftr.com/template-tg.html and kungfuftr.com/tg.css
(i have a slightly newer version, but that's at home
nothingmuch ok
what does it do differently?
xhtml 1.1? 08:43
i see funny tags
and I see css saying display: table-cell, etc 08:44
but i can't say if that's the whole difference
(btw, how backwards compatible is this?)
kungfuftr not hugely, tbh
anything that's CSS2 and xhtml will support it 08:45
nothingmuch ok
i'll apply it soon
kungfuftr nice thing is that it gives the ability not only to view it, but to parse it as XML too
nothingmuch yeah
although if you're sane you just usse Test::TAP::Model =)
kungfuftr yar, need to change the code to not use classnames as selectors at some point too 08:46
boogie nothingmuch: so is there somebody who could implement "read"? :))
nothingmuch like read($fh)? 08:47
with byte counts?
boogie yes, read($fh, $string, $length);
nothingmuch i can try 08:48
but in about 30 minutes
nothingmuch has to do some C++ hacking
=(
boogie personally, I would do it on a socket, but on normal STDIN would be usefull, too.
thanks a lot!
mj my $b = [ <a b> ]; 08:57
say $b[0].ref ~ $b[1].ref ~ $b[2].ref;
StrStrScalar
undef == Scalar ?
and 08:59
my $b = [ { <a b> }, { a } ];
say $b[]().ref;
and 09:02
my $c = [ { <a b> } ];
say 'c ' ~ $c[]().ref ~ $c().ref;
second my $b ... pugs crasses ... third 'c ListList' 09:03
:-) || :-( ,,, i can't reproduce crash 09:12
my $a = undef; say $a.ref; # Scalar ... seems ok 09:16
so question is 09:17
my $a = [ { 1 } ]; say $a().ref; # Inf
my $b = [ { 1 }, { 2 } ]; say $b().ref; # pugs: cannot cast from VList [VCode...
09:20 b6s_ is now known as b6s_utf8
autrijus mj: make them into tests! 09:30
mj: do you have committer bit?
brb... 09:32
mj mj := rootmj; # committer bit found 09:33
time to go home, maybe tomorrow
masak hi autrijus 09:34
i found a deep array modification bug when trying to write tic-tac-toe 09:37
-bash-3.00$ pugs -e 'my $a = [[2]]; $a[0][0] = 6; say $a' 09:38
Can't modify constant item
Val VUndef
kungfuftr masak: looks possibly like an auto-quoting bug (ie: is trying to subscript instead of assigning value '2') 09:55
masak kungfuftr: i don't understand but i don't think it matters... i'm only a tester :) 09:59
masak has to learn haskell soon
nothingmuch autrijus: ping 10:08
tiw hi nothingmuch 10:13
nothingmuch hola
que tiw?
tiw is dereference implemented?
nothingmuch i have no clue
i'm out of date
tiw aha, me too 10:14
nothingmuch goes to drive dad 10:20
10:23 b6s_ is now known as b6s
machack666 is this expected behavior: pugs -e 'say (3).ref' --> Int 12:24
pugs -e 'say 3.ref' --> parse error (unexpected 'r')
stevan boogie: what are you looking to do? 12:35
boogie: have you seen the CGI and LWP::Simple in Pugs already? 12:36
nothingmuch: ping
pugs -e 'say ref(undef())' # Scalar 12:38
I think that makes sense :)
boogie stevan: I would like to "hack" a web server. 12:39
stevan: just a small for playing around
stevan boogie: I was thinking of doing that too :) 12:40
boogie: have you see examples/network ?
boogie stevan: Yes, on the perl-compiler mailing list as "Andras Barthazi" I had some mail
stevan boogie: ahhh, welcome :) 12:41
boogie: did you get the commiter invite yet? I just sent it to you
boogie Yes, I've started reading it now. :) 12:42
nothingmuch pong 12:44
cognominal in haskell, one can use a function name as an operator, say: x 'mod' 2 12:57
Is there anything like this in Perl6?
stevan &infix:<my-op> 12:59
or something like that (i think)
cognominal thx: that means that one don't even needs the quotes unlike haskell 13:05
autrijus x `mod` 2 13:09
and yes:
&infix:<mod> ::= &mod;
or if you are feeling haskellish 13:10
&infix:<`mod`> ::= &mod;
hrm. does :x-on-by-default applies to :perl5 also? 13:11
I'll assume so 13:12
stevan autrijus: I would think so 13:14
masak stevan: there's a comitter invite? what is it?
autrijus cool. expect full flags coming
stevan masak: are you a committer?
masak yes
stevan autrijus++
the invite is just the mail sent from OSSF 13:15
masak ah, ok
i seem to remember it vaguely... parts of it were in korean :) 13:16
stevan chinese :)
masak right :) i still haven't managed to conf fedora core 3 to show any of those correctly :( 13:17
masak wants chinese characters 13:18
stevan masak: thats why I like OSX ,.. it all works out of the box
masak osx++
stevan has several asian fonts installed,.. but can read none of them
kungfuftr nothingmuch: you like? 13:19
stevan has anyone told perl-qa about Test::TAP::Model? 13:21
kungfuftr don't think so, have a feeling it's not quite ready just yet
stevan kungfuftr: I disagree,.. the official module is not ready, but the Pugs version is up and working 13:22
we could show them that
kungfuftr up and working?
autrijus sanity-check:
stevan the smoke tests you run
autrijus { rxRegex = mkRegexWithPCRE (encodeUTF8 str) $
[ flag_i `implies` pcreCaseless
, pcreExtended -- Always assume /x
, pcreDotall -- . matchs \n
, pcreDollarEndonly -- $ does not match \n$
]
i.e. even if you write :p5
kungfuftr stevan: current smoke stuff i'm running is what's in util/
autrijus it adds //sx for you
and cause $ to not match the eol \n 13:23
thus agreeing with p6 default behaviour
stevan kungfuftr: yeah that stuff
autrijus sane? (you can always turn it on/off inside the rx)
stevan autrijus: I am not a p5 rexexp guru, and I like the p6 ones better, so I am not a good judge
kungfuftr stevan: that stuff isn't Test-TAP-Model, it's it's predecessor
stevan but yes I think its sane :)
autrijus :D
ok. I'll do that and if people are hurt, they'll yell 13:24
stevan autrijus: but many people (including my family) question *my* sanity , so I am not a good judge of that either
autrijus another option is to suddenly allow perl5 modifiers after :p5
i.e.
:p5:s:m///
stevan kungfuftr: I am just thinking that we can say "test::TAP::*" is on the way,.. and here is the early version
autrijus: are you shortening it to :p5:? 13:25
autrijus that may create another insanity mess though, so I tend not to go there
stevan: e05 says both :p5 and :perl5
but s05 does not mention :perl5
stevan autrijus: p5: is much nicer
autrijus I think I'll allow both.
stevan sounds good to me
kungfuftr stevan: i'd wait, tbh... tighten it down a little
stevan kungfuftr: ok with me
autrijus strangely we almost always agree.
stevan autrijus: scary isnt it 13:26
perl 6 makes so much more sense to me
I love perl(5), but before this project started I was feeling the Ruby/Python/Java pull
autrijus indeed. 13:27
stevan craving multi-subs/methods, type, etc
kungfuftr new template stuff++ # it's XML, but yet you can view it directly in firefox/<sane browsers>
autrijus and I guess CPAN was keeping you afloat as well
stevan likes XML actually
kungfuftr nothingmuch: you going to be around at 7pm BST?
autrijus likes XML with XML::Twig only
stevan autrijus: that and the 13K of perl5 code which keeps our company afloat :) 13:28
nothingmuch BST? what's that?
stevan XML::SAX++
autrijus 13kloc is not terribly large :)
kungfuftr british standard time, 1 hour ahead of GMT
nothingmuch i'm leaving for my bass lesson in about an hour and a half
autrijus I imagine porting it to p6 will be ~2wk of work
nothingmuch then i am going to a friend
stevan autrijus: exactly :)
nothingmuch should be back by midnight
which is 22 gmt
21 bst, appearantly
stevan but porting it to Ruby/Python etc is not possible
nothingmuch if daylight savings is in synch 13:29
kungfuftr ah! k, well i should have more improved templates for you then
stevan (it depends on mod_perl, DBI, etc)
nothingmuch is gmtime 13:30 or 14:30?
one can never trust computers when it comes to israeli DST, which is determined pretty much at random every year
kungfuftr: goody =) 13:30
kungfuftr nothingmuch: use subversion!!
since darcs is broken on *bsd currently
nothingmuch kungfuftr: =(
kungfuftr svk++
nothingmuch my home box is disconnected too often for svn to be useful
and i haven't really used svk enough 13:31
i might sometime
kungfuftr perfect usage for svk
nothingmuch but i *really* like darcs
kungfuftr it's very nice, bloody similar to svn
nothingmuch and i have 0 time right now
kungfuftr np! have fun
nothingmuch especially with antihistemines ruining my sleep cycle
boogie I'm trying to run module tests, and I got errors. Somebody can tell me, what' wrong? 13:32
nothingmuch boogie: what are the errors?
perlbot nopaste
perlbot Paste your code here and #<channel> will be able to view it: sial.org/pbot/<channel>
boogie This is the messege from the *.t files: /.../boogie/work/pugss/pugs/blib/script/pugs: No such file or directory
nothingmuch btw, i didn't get to read(), i'm being suffocated today 13:33
how are you running the *.t? prove? make test?
hmm
are you make installing?
./pugs t/foo.t?
autrijus ok, regex patch is in.
boogie make test in the pugs/ext/CGI/lib directory
autrijus enjoy
nothingmuch 's never done that
can't really say 13:34
autrijus: releng issues?
boogie: try HARNESS_PERL="/full/path/to/pugs"
(the executable)
autrijus nothingmuch: what releng issues? util/pugs_config.h not regenerated by default?
nothingmuch no, what boogie is experiencing
autrijus oh
nothingmuch and then PERL6LIB="/ful/path/to/ext/Test/lib"
export these two variables 13:35
autrijus the tests are meant to be run from toplevel
not from the ext/ itself
nothingmuch and then try prove t/ in the dir
autrijus that's all
i.e. known issue. fixes welcom
e
nothingmuch Test::TAP::Visualize::HTMLMatrix is being tested to detah
autrijus I want that :D
boogie nothingmuch: read() is not a big problem, but it would be nice it would work. I you'll have some time, please try to hack in. ;)
nothingmuch we have a home grown C++ unit framework
autrijus nothingmuch: will you backport it to pugs?
nothingmuch it's also using TAP
stevan boogie: inside ext/CGI/ run this: prove -I lib/ t/*.t
that should work
nothingmuch since it catches SEGV, divide by zero, and all other typically very fatal things it's the first use of Bail out i've seen 13:36
so that is supported
autrijus: it will stay p5, because i want it to use a templating toolkit
but i will make testgraph.pl use it
boogie autrijus: thanks for the tip :)
nothingmuch Real Soon Nowā„¢
btw, you can darcs get it 13:37
autrijus nothingmuch: no, I mean port pugs to use that
port the existing util/ stuff
so, exactly what you just said :)
Real Soon Now++
nothingmuch i'll do that eventually 13:38
that was the initial plan anyway
autrijus nothingmuch++
nothingmuch btw, kungfuftr: as far as I'm concerned, if you want write access to my public repo, and a shell on my box to use darcs, it can be arranged
kungfuftr oh, got vim? 13:39
nothingmuch is there any other editor?
=D
nothingmuch just tought the new sysadmin some vim
with a 3d matrix metaphor
kungfuftr wait a sec
nothingmuch and he basically said "cooooool"
kungfuftr autrijus: can svk mirror darc repostiories? 13:40
autrijus kungfuftr: no. want to help?
kungfuftr doh
autrijus VCP::src::darcs
nothingmuch planned on doing that
=(
but then again, i didn't learn SVK either
autrijus or just a simple darcs2svn
and we can reuse that
stevan needs to try out SVK again
nothingmuch there's darcs2svn
the other way is a bit harder
kungfuftr nothingmuch: i've never used darcs, but hey, i can learn
nothingmuch it's very simple 13:41
so i doubt there'll be any trouble
kungfuftr yeah, sounds like a plan then
nothingmuch kungfuftr: there was a small Petal error in my template, btw
i fixed it now
kungfuftr ah, might be changed code, etc
nothingmuch what login would you like?
kungfuftr kungfuftr/scott 13:43
either or
nothingmuch www.zanorg.com/pitbull/shakingpope.swf 13:48
kungfuftr nothingmuch: ta! 14:02
cognominal wky so much hate. this guy was about the only one fighting against all sort of slavery invented in the sake of liberalism. This flash game is fun, though. 14:17
nothingmuch i don't think he was bad
it's all good fun
cognominal :)
nothingmuch i think our own religeon-monkies are much worse 14:18
they are corrupt, vile, hateful, and manupulative 14:19
stevan cognominal: I dont think it's hate,... a little mean spirited maybe
and you have to realize that to some people (ex-catholics) he symbolizes a lot of negative things
cognominal I hated instutionalized religion, his marial cult was almost paganist and his postion of contraception was backward. but he was probably the best pope catholicsm could have.
stevan cognominal: yes, but that is not saying much :) 14:20
nothingmuch doesn't know enough to enter this argument
cognominal and if they elect a pope in the southern hemisphere, this can shake things.
stevan the contraception position alone can be used to blame much of the AIDS crisis in africa
I personally am hoping for a non-european pope 14:21
by which I mean an african or south american
that alone will force so many people to deal with their own personal latent racism
nothingmuch that could be cool
stevan and I think that is a good thing :)
nothingmuch stevan++ 14:22
stevan nothingmuch++
but then again I am a bitter ex-catholic, so I should probably not talk anymore about this :)
cognominal I had a friend, a born provocative, who went to a moonist church and said he could not accept tha the leader of a church cannot be white :) 14:23
ok, back to learning haskell 14:24
stevan cognominal: the reign of the white european will hopefully end soon,.. we (I am one) have not done a very good job IMO
cognominal stevan: just see the AUTHORS in pugs :)
stevan cognominal: yup :)
Pugs is mucho internationale 14:25
stevan just makes up languages sometimes
autrijus o/~ l'internationale / unites the sentient races o/~
cognominal and most good research papers in the states are by indian or chinese. This say much about the future of the states
stevan tries to light an american flag on fire .... :P 14:26
cognominal I will denounce you as a terrorist 14:27
nothingmuch update AUTHORS
stevan damn lighter is not working
nothingmuch bought a zipp
o
that usually works
autrijus mmm more right-to-left love
nothingmuch not that i smoke that often
stevan cognominal: bah! I've been on The List for years 14:28
it all started with that trip to the Soviet embassy in Mexico city with my buddy Lee Harvey
cognominal come to Europe, we are decadent, but at least we know about it.
stevan cognominal: I used to live in Germany, I miss it actually 14:29
stevan 's father worked for the US Govt. for 20 years
cognominal I must say I like European cities in that states, like SF or NY. 14:30
stevan I used to live in NYC too :)
but rents got too much and my wife had twins and we needed room :)
rgs stevan: so maybe you know about the ^Hatalien landing field in nevada ? 14:31
maps.google.com/maps?ll=37.627187,-...&hl=en
s/atalien/alien/ :)
stevan rgs: not that branch of the Govt :) 14:32
rgs :)
stevan (although that would have been really cool)
nothingmuch mom? where's dad? is he coming home?
stevan actually he used to investigate fraud in govt. contracting,.. so I was raised to be very anti-corporate (which makes job hunting hard)
nothingmuch no, Stevan, he is working very hard on something
stevan ground control to stevan's dad...... 14:33
nothingmuch in fact, he worked so hard, that the gov't decided to reward him, and give him a sort of, vacation
stevan LOL
I'm half martian :P
I actually tend to not believe in govt. conspiracies because I know how f*cked up and confused the govt. is and know that they could not pull anything so complex off 14:36
cognominal I was searching that quote. America is the only country that went from barbarism to decadence without civilization in between.? Oscar Wilde. Hopefully american in the perl communauty are very nice
nothingmuch that's why the MIB excuse is there
conspiracies above the govt
stevan cognominal: thats a great quote
nothingmuch to make sure it's done right ;-)
stevan hears a knock on his door and hides under his desk 14:37
nothingmuch oh crap, there's my old house! 14:39
maps.google.com/maps?ll=37.627187,-...&hl=en
no wait
how do you take a link to a google map?
theorbtwo Click the "link to this page" link in the upper-right corner. 14:40
nothingmuch oops
theorbtwo++
stevan there is my old building: maps.google.com/maps?q=1619+3rd+ave...&hl=en 14:42
google maps ++
theorbtwo Sadly, there is horrible resolution around where I used to live.
stevan theorbtwo: yeah the same with my current house
theorbtwo Oddly, there's a seam along the exact borders of Massachusetts. 14:43
clkao win23
stevan ok I am off to $work, later all 14:53
nothingmuch beh, back 15:00
boogie Hmm. I'm not sure, but it seems to me, that something broken with the current pugs version. When I'm trying to run tests of the CGI modul, it gives no error, but going to sleep during testing. 15:07
Maybe regexp? Can somebody check it? Revision 1962 is OK. 15:08
jabbot pugs - 1960 - Refactored test locations after S29 15:56
pugs - 1961 - Add identity tests for arrays and hashes
pugs - 1962 - test for changing of nested array
pugs - 1963 - * :p5 allows as well as :perl5 for regex
pugs - 1964 - * die "" confuses harness, so use exit()
pugs - 1965 - Gaal's name in AUTHORS can also have heb
pugs - 1966 - * Reverted to "is" instead of "isnt"
pugs - 1967 - <nothingmuch> $a = [ 1, 2, 3 ; 4, 5, 0 ]
pugs - 1968 - utf-8 URL decoding support with some tes
pugs - 1969 - Converted tabs to spaces and fixed inden
theorbtwo wonders if it's too early to start thinking about a gtk binding via gtk2hs. 16:11
crysflame hello stack of commits 16:17
gaal theorbtwo, not to early 16:36
me want me want 16:37
though we do need objects.
theorbtwo Exactly my wonderment. 16:40
...esp because there does not seem to be a callback in place for "a method got called on us" in the IType system. 16:41
I could do a lot of faking, but I'd prefer to not do that now, since it'd just mean a rewrite latter down the chute. 16:42
autrijus cwest: oh. you're here. 16:48
autrijus switches channel
jabbot pugs - 1970 - Upgraded HTML::Entities and discovered t 16:55
stevan kungfuftr: do you have that URL for nothingmuch's Test::TAP darcs repo?
boogie: you need to add yourself to the AUTHORS file, and the CGI AUTHORS section 17:01
autrijus ...and thus make hangman.p6 more challenging to win 17:04
...which is a Good Thing
stevan ;)
stevan is off to have lunch
boogie stevan: ok 17:08
autrijus: :D
ninereasons is there an svn command to delete everything with status "?", as listed by "svn status" ?
autrijus svn status | grep ^\\? | xargs rm 17:10
or, easier:
rm -rf *; svn up
:)
ninereasons I've done the second one, often enough :) 17:11
kungfuftr bbiam 17:22
jabbot pugs - 1971 - * :i :m :s :x now all work in :P5 mode. 17:25
autrijus stevan: larry overruled us 17:26
so, no /x by default
crysflame :( 17:36
autrijus that means you won't have to rewrite your old regex when porting over tho. 17:38
ingy hola 17:44
autrijus yo ingy. 17:45
I'm journaling, about to sleep
jabbot pugs - 1972 - ":perl5" modifier now spelled ":Perl5"
gaal autrijus: bug in the changelog/latest journal entry. [1..10][0] != 0 :) 17:54
ingy :) 17:55
autrijus ... 17:56
it's assignment
assignment!
crysflame heh
gaal haha
masak :)
gaal with autovivification even 17:57
well, potentially.
autrijus yup 17:58
gaal my $x; $x[2][1][0]="liftof"; $x.perl.say 18:00
autrijus that looks good 18:01
gaal yes
except i should spell it "liftoff". :)
theorbtwo Allo, autrijus, the newly-natively-spelled-in-AUTHORS gaal, others. 18:02
gaal urrr?
wolverian hello.
sleeping through the day is the best thing ever.
is there any patch limit for adding oneself to AUTHORS? 18:03
gaal aha! that nothingmuch, less pizza for him :)
autrijus journal up. g'nite!
wolverian good night. :)
gaal night autrijus
kungfuftr moo 18:05
theorbtwo notices that Autrijus never answered my question about gtk. 18:11
autrijus theorbtwo: it's never too early. go for it. 18:13
'nite! &
kungfuftr autrijus: know anything about web access on pasta.woobling.org? 18:15
theorbtwo G'night.
castaway it belongs to nothingmuch 18:16
kungfuftr yeah, to to find a way to get http access to my account 18:17
castaway http access?
kungfuftr ah 18:20
ninereasons i'm irritated that I misread iblech's bug report 18:22
castaway in the channel? 18:23
ninereasons no, a committed pugsbug
i seem to be having more trouble with english than with perl 18:24
not a good sign at my age.
castaway awww
theorbtwo wonders what that age is. 18:25
jabbot pugs - 1973 - Updated qotw-regular-24.p6 to work aroun
pugs - 1974 - The context inside of hash and array con
pugs - 1975 - * P5 flags are now spelled as :P5<imsgx>
ninereasons ancient, by pugs norms.
castaway I think you're getting forgetful in your young age, love 18:26
theorbtwo Getting?
castaway ok, are ;)
ninereasons I think i might be the oldest here, and perhaps the least experienced as well - at 48
theorbtwo Ah.
castaway was close, said 47 ;)
you must look younger, ninereasons ;) 18:27
ninereasons www.perlmonks.org/?node_id=44862
castaway ooh, black&white.. ancient ,) 18:28
ninereasons :)
theorbtwo 9: Larry is older then you. 18:30
ninereasons yes, but he doesn't hang around here. lots of young people in #perl6 18:32
ninereasons hates the sound it makes when he says "young people"
castaway heh
kungfuftr yay!
castaway yay? 18:33
ninereasons access to pasta?
kungfufter, celebrating all alone? where's the fun? 18:35
castaway Corion! 18:36
Corion castaway! 18:37
Corion would really like it if the people changing Pugs --version would also change the tests they break. 18:38
kungfuftr ninereasons: heh... well if i can remember how to set up my own local perllib, everything will be hokeydokey 18:40
Corion wonders why tests show as broken in the test graph, but work when run from the command line ... 18:41
ninereasons Corion, a harness thing, maybe?
Corion ninereasons: Might be, yes ... 18:42
jabbot pugs - 1976 - map dies with an untrappable error messa 18:55
pugs - 1977 - Oops, description of previous checkin ma
pugs - 1978 - Fixed --version tests failing
pugs - 1979 - Fixed changed filename of fp-p6
kungfuftr ninereasons: pasta.woobling.org/~kungfuftr/Test-...mple.xhtml 19:03
have a looky at the source
q
nothingmuch good evening 19:08
castaway 4nm!
oops s/4//
nothingmuch 3cast34y!
theorbtwo ~~0
stevan nothingmuch: mah main man
nothingmuch stevan: their auxillery woman! 19:09
kungfuftr nothingmuch: yullo
nothingmuch hola kungfuftr
i am back early, as you can see 19:10
kungfuftr hacking now
nothingmuch as we say in hebrew, someone banged up a tap for me
Corion kungfuftr: Looks nice indeed
kungfuftr nothingmuch: pasta.woobling.org/~kungfuftr/Test-...mple.xhtml (improved version)
nothingmuch (tap, like a water tap)
ooh, purrrdy
what about spreading them out nicely?
kungfuftr oh? that not werking for you? 19:11
nothingmuch nope
kungfuftr hhhmmm... odd
Corion Spreads nicely over 100% of the body for me
nothingmuch i'll try to see why
Corion (where conventional lotion only covers 95% of my body)
nothingmuch i'm also not seeing bgcolor for testrunpercent 19:12
Corion: no visual metaphors please
kungfuftr nothingmuch: which browser?
nothingmuch sort of 19:13
safari
which ~~ konqueror
kungfuftr ah, khtml--
nothingmuch i think maybe the display: table stuff is giving it crap 19:15
kungfuftr yeah, safari isn't CSS2 compliant yet 19:16
nothingmuch oi crapxor
kungfuftr nothingmuch: any idea what the .nottest class is for? 19:30
nothingmuch yes
diag from before test output
but it's not in template.html yet
i was doing edge cases on bail out behavior today
stevan hey I was thinking 19:32
I dont like YAML
(no offense INGY)
how flexible is the underlying Test::Harness collector
nothingmuch that's why Test::TAP::Model is agnostic
stevan can I write an XML::SAX collector? 19:33
nothingmuch it is a subclass of Test::Harness::Straps
stevan nothingmuch++
nothingmuch that's pretty easy
override _handle_test in Test::TAP::Model
stevan gaal++ # cause it was his cracked out idea to start with
nothingmuch to generate events instead
stevan cool I will look into that later tonight maybe
nothingmuch (actually i am shamelessly taking credit for that, it's just that i never got around to it, and he did ;-)
theorbtwo needs a project.
kungfuftr nothingmuch: you might need to darcs get 19:34
stevan theorbtwo: you like XML::SAX?
nothingmuch i'll do it momentarily
stevan really likes XML::SAX
theorbtwo Not purticularly.
stevan theorbtwo: you'll love it :)
none of this tree building crap,.. pure stream/event based processing
very clean, very nice
theorbtwo I think the stuff in util/ works fine, and am not purticularly intersted in rewriting the stuff that's already there without making it appreciblally better.
theorbtwo finds print to be a very nice clean way of generating HTML. 19:35
stevan theorbtwo: well at least for Test::TAP::Model I dont like having YAML as a dependency because it seems to have issues
kungfuftr theorbtwo: should have a peek at the testgraph i did
nothingmuch stevan: i prefer Storable for it because it's much faster
theorbtwo: sinner 19:36
theorbtwo YAML is diffable, though, which can be nice.
nothingmuch i see serializing modules this way:
stevan so is XML
nothingmuch YAML is for humans
Data::Dumper is for debugging
Storable is for storing
kungfuftr how often does anyone _actually_ look at yaml that's used for serialisation? 19:37
stevan XML is streamable too,.. YAML is not
nothingmuch YAML is also nice for config files
YAML is supposed to be streamable 19:38
stevan nothingmuch: it is not
theorbtwo Also, XML knows how to deal with charsets.
stevan you have to build the whole thing at once
theorbtwo XML is understood by people outside of the perl world.
stevan before you write it
theorbtwo: naw yur cookin!
theorbtwo In any case, like I said, I think there are no major problems with the util/ stuff as it exists, and am not purticularly interested in a rewrite that does not have substantitive new features that interest me. 19:39
nothingmuch i think doclinker should be generailzed into a module
stevan its less for util/ and more for Test::TAP::Model
nothingmuch debugs 19:40
theorbtwo I think the stuff in util could be cp'd and released as a bunch of scripts without major changes.
...and I'd prefer the end-result doesn't require much more then it already does.
stevan theorbtwo: aside from the doclinker,.. thats what Test::TAP::Model is right? 19:41
theorbtwo As far as I could tell from what I looked at the other day, Test::TAP::Model was a rewrite with a big dependency or three. 19:42
stevan what deps?
theorbtwo Er, wait.
Test::TAP::Model isn't what I mean.
I mean Test::TAP::Visualizer::HTMLMatrix
stevan yes
nothingmuch Test::TAP::Model? no deps 19:43
Test::TAP::Visualize::HTMLMatrix needs Petal
that's all
nothingmuch prefers dependancies over flakiness
theorbtwo doesn't think his code is flakey.
nothingmuch there were many cases it didn't handle
and it was pretty hard to read
and harder to modif
stevan nothingmuch: is ::HTMLMatrix structured that we can use any template module?
nothingmuch IMHO
stevan: in theory yes
it's basically just marrying Test::TAP::Model with Petal 19:44
in not too many lines of code
theorbtwo Oh, you should have said something; I would have put in another sub or two.
stevan Test::TAP::Visualizer::HTML::Template , etc etc etc
nothingmuch TT should be tres simple
stevan TT can be a pain to install sometimes
ideally it can support all forms of templating,... even print()
plugins man,... plugins 19:45
nothingmuch too much for 0.01
stevan wussy
theorbtwo shrugs.
nothingmuch i was concentrating on Test::TAP::Model
which is where the logic is at
theorbtwo You're welcome to do whatever you like with my code, just credit me and I'm fine.
nothingmuch theorbtwo: you shall be
worry not 19:46
masak says good night 19:55
nothingmuch night
jabbot pugs - 1980 - LWP::Simple test fix
stevan nothingmuch: what time is it there for you? 19:57
nothingmuch 22:57
Limbic_Region chip about? 19:58
seen chip?
jabbot Limbic_Region: chip was seen on Fri Apr 8 19:06:05 2005 GMT
kungfuftr nothingmuch: there be no more HTML!! 20:01
nothingmuch aaah! what have you done?
what have you done with it's body? 20:02
kungfuftr it's still here, the rest is all xml now though (apart from links)
should be some more tidying up and that's be about it 20:04
nothingmuch tries very hard to think of another ref to brazil, but just cann't
kungfuftr & # food
nothingmuch ok
how about khtml?
kungfuftr not css2 compliant, gecko and opera are the closest tbh 20:05
nothingmuch so we just drop support? i'm not sure that's a good idea
Corion khtml ... KDE is for losers anyway! :-)))
nothingmuch khtml is used by safari too 20:06
and that's me
and i will use the module
baaah!
20:06 co3 is now known as co2
kungfuftr nothingmuch: screenshot? 20:06
nothingmuch sure
ooh, testgraph.pl wrangled
that was easy
as soon as i found out that _init was not calling SUPER::_init 20:07
gaal--; # not very careful
nothingmuch.woobling.org/badkhtml.jpg 20:09
kungfuftr hhhmmm... i sorta had to put a hack in for that... 2 secs, see if i can get it in another way 20:10
Corion nothingmuch: Looks like a VU meter - funny too :) 20:11
nothingmuch vu meter?
Corion nothingmuch: Loudness / volume indicator
nothingmuch ah
Corion www.google.com/search?&q=vu%20m...id=firefox
(google displays two images directly on the results page for me) 20:12
"Google, the Emacs of Webservices" -- seen on slashdot
... but it doesn't show any images for "hardcore porn" on the search results page :)
nothingmuch when we have google os
that will auto spell check your syscalls
then i'll agree with that quote
kungfuftr nothingmuch: try now
nothingmuch blammo! 20:13
kungfuftr moo?
nothingmuch however: <testpercent style="background-color: #00ff00">100.00%</testpercent>
the style="" is a no op
btw, i would like to keep status=failed etc at 1 char 20:14
and make it reaaaaallly concise
because this stuff *grows*
kungfuftr bad idea, tbh. it's like having 15,000 lines of code with 2 character variable names 20:15
nothingmuch: so it working for you now?
nothingmuch the table is working
kungfuftr yay!
nothingmuch but the % column is all white
kungfuftr screen shot? 20:16
nothingmuch the style="background-color: #xxxxxx" stuff is not working
the old screenshot applies
the reason i want 1 char classes/attrs is that this makes it shrink by several kilobytes
the std testgraph.pl html is around 1mb 20:17
Corion 800k for me
nothingmuch btw, i think there should be a black border on the left side of the whole table
and that the test "status" should be centered 20:18
and that "test file" should have colspan=2
but i don't insist
i would like some convincing though
kungfuftr nothingmuch: colspan is impossible unfortunately
nothingmuch ah
kungfuftr not sure what's going on with the testpercent stuff 20:19
think there's another possibility though
nothingmuch =/ 20:20
theorbtwo: did you ever get around to hooking svn log -> authors -> coopyright? 20:21
kungfuftr nothingmuch: can you give me the rgb values for 0,10,20,30,40,50,60,70,80,90,100 percent? 20:22
nothingmuch the value is not hard coded
see Test::TAP::Visualize::HTMLMatrix::Colorful
my current thoughts are: XML for display might not be ready, if browser support is lacking 20:23
perhaps if it ain't broke we shouldn't fix it?
i would prefer, given current status of KHTML to backport the change
keep your patch around, but work on it
and in due time revert the patch that converts it to regular HTML
kungfuftr hhhmmm... 20:24
nothingmuch return sprintf("#ff%02x%02x", $l + ((255 - $l) * $ratio), $l-20);
$l is at 100
it gives it a slightly softer tone
kungfuftr bollocks that i don't have safari locally to test
nothingmuch true that 20:25
kungfuftr nothingmuch: try now 20:30
nothingmuch nope 20:31
same old
kungfuftr should be a failure in the list, right?
nothingmuch yes
there is
but the 100% isn't green
and the 90+% isn't yellow
the matrix cells /are/ green and red 20:32
kungfuftr yeah, is here... very very weird
safari's not even obeying !important
nothingmuch *sigh*
technology sucks so much
kungfuftr 2 secs 20:33
crysflame safari normally obeys !important? that's odd 20:37
gaal yo.
nothingmuch hola gaal
kungfuftr ah, safari rendering bug
quirksmode++
2 mins 20:39
gaal ooh, we have a workable new smoke system? 20:40
shiny shiny, shiny new smoke system
new smoke system, in the dark
nothingmuch not really 20:41
oh my gawwwwed
layout completely screwey
gaal with red and green, tooltips that do delight 'er
nothingmuch safari is annoying
gaal come on baby, smoke for me!
nothingmuch kungfuftr: i think that it's too early for XML
stevan gaal++ # for the VU ref 20:42
kungfuftr nothingmuch: yeah, most places don't support it...
i think safari's too early for XML
=0)
nothingmuch too bad
well, stevan is planning on making an XML backend anyway
i think
or is it just SAX?
kungfuftr nothingmuch: why safari? 20:43
stevan nothingmuch: yup,... but just the backend Test::Harness stuff
nothingmuch why do i use safari?
kungfuftr yar... understand mac... but why not something gecko based?
nothingmuch stevan: what kungfuftr did (pasta.woobling.org/~kungfuftr/Test-...ple.xhtml) was try to get it to both be displayable and minable at the same time
my puter is hella slow
400mhz
firefox is much less responsive than safari
kungfuftr never found that, and i use both 20:44
stevan minable?
nothingmuch much less stupidity... it works so well in hebrew!
stevan: it's XML that makes sense
gaal VU ref is even better than VUndef, i think.
kungfuftr stevan: have a look at the source code of that xhtml page
gaal needs water 20:45
stevan kungfuftr: very nice
works for me in Safari
nothingmuch stevan: do the percentages on the right side have color?
stevan no
jabbot pugs - 1981 - Updates, clarifications, and expansions
nothingmuch i think we should stick with plain old HTML
(xhtml ;-) 20:46
so it's not just my box
beh
stevan I tend to agree with nothingmuch on this
kungfuftr will play tomorrow on the mac at work
stevan at least for now
nothingmuch but keep the changes somewhere
kungfuftr: can you commit the change?
stevan *cough* plugins *cough*
nothingmuch so that i can backport?
stevan: patches welcome ;-)
stevan some organizations may want this
nothingmuch i want to do basic releng today
stevan nothingmuch: don't make me do it!!!! I'm just crazy enough!!!!
kungfuftr nothingmuch: commit, how? 20:47
nothingmuch kungfuftr: darcs whatsnew
gaal heh, nothingmuch, you should port Object::Meta::plugin to perl6 :)
nothingmuch darcs record
gaal: i think there will be no need
stevan gaal: would he need too?
nothingmuch for a while at least
stevan $object.meta.* should do enough I would think
nothingmuch roles (per object) do most of what it does
stevan: you *so* don't want to know what Object::Meta::Plugin does
as for plugging in templates:
stevan gaal: if nothingmuch sets up that bit torrent server he promised too (poke poke) then I got some nice VU stuff to share 20:48
nothingmuch i don't want to do it yet because the petal <-> Test::TAP::Model bridge is very thin at the moment
i'll do that now
gaal stevan: in Dichotomy Deset Island Land, would you bring John Cale or Lou Reed with you? 20:49
(well, their ALBUMS I mean. :)
stevan Lou Reed
gaal was it a hard choice?
stevan but thats because I really love the "New York" Album
not really. 20:50
gaal: and you?
nothingmuch kungfuftr: so what's next?
gaal I'd pick John Cale, but I'll miss Berlin. And Transformer. And Magic and Loss. :/
stevan Transformer is a great album
gaal yes.
I don't know the New York one! 20:51
stevan gaal: even more of a reason for nothingmuch to get the server up
gaal hehheh
nothingmuch one minute please, i'm reading up on how
stevan it came out in the early 90s,.. when Lou was in his "Rock and Roll should be pure" phase
gaal i don't think I'll be listening to a lot of music soon though :(
you mean purely beer? 20:52
nothingmuch gaal: want to borrow my ipod?
stevan gaal: they dont allow music in the big house
nothingmuch it might get stolen
but then again, i don't use it much anymore
stevan 2 guitars, bass and drums,.. thats all
nothingmuch haven't commuted by bus in a long while
stevan gaal: soap on a rope,... get some
gaal ?
stevan it's an american jail thing 20:53
dont drop the soap in the showers
gaal thinks he doesn't wanna know :)
stevan gaal: you dont :)
gaal++ # again for standing up for what you believe in
gaal when i lived in the uk there was this commercial, eh, crap i forgot who it was with, some singer 20:54
stevan I know if they did that in the States,.. I would have been a canadian long ago
gaal he gets into trouble (it's a *series* of commercials) and in one of them he ends up in jail
there's a bit with the soap there too, though ha-ha-understated. 20:55
kungfuftr nothingmuch: ah ha... i know a fix
gaal nm: an ipod will last about thirty seconds there. but thanks :)
stevan gaal: you could trade it for protection from the big guys 20:56
gaal s/will/would
stevan or cigarettes
gaal it's not that kind of prison
really won't be that bad. won't be fun, of course
stevan gaal: can you bring books?
nothingmuch pasta.woobling.org:7080 20:57
gaal yes. i killed a tree the other day printing p6b
stevan ah yeah you said that
gaal plus i have: david lodge, richard russo, mikhail bulgakov, uh the guy who wrote fight club, and some others. 20:58
nm, is any of the new matrix stuff p6? i haven't had time to look at it 20:59
crysflame p6b?
nothingmuch gaal: no 21:00
just your code in OOP
and kungfuftr is working on the XH?TML output
gaal "my" code, hah. that's mostly petdance's code, etc.
nothingmuch nevermind
=D
gaal are we not doing it in p6 because pugs isn't there yet, or because we want this to be used in p5land? 21:01
(is there open| in pugs?)
nothingmuch both, methinks 21:02
gaal how's the upgrade path for tests of existing p5 moduels to p5? should we bother porting Test::*? From a clean slate POV probably it'd be best if we worked on a very strong single Test.pm, but i'm not sure how happy existing p5 authors would be when their hard work doesn't work any more 21:05
i mean, porting your moduel == fun
porting tests == less so, probably
nothingmuch no clue 21:06
gaal or at least, less necessarily
nothingmuch i think i agree on strong single Test.pm
but not 100%
sometimes there are many ways to do one things
is_deeply should be gone
gaal well i sure don't want 15 separate modules - at least, not 15 separate distribs
nothingmuch Test::Deep is the way for deep 21:07
and not every one needs it
Test::Exception should be core
Test::NoWarnings is odd
the Test::Distribution family is weird
it's functionality should be in Test::Exceptions
i think it should be STD
but rethought, interface wise
gaal sure, refactoring++
stevan I agree,.. Test::Exceptions should go into Test.pm when we have CATCH
nothingmuch Test::LongString is silly
stevan i think chromatic was going to start on the Test::Builder port
nothingmuch Test::Simple shouldn't exist 21:08
stevan I love Test::Deep
nothingmuch Test::MockObject should stay (but get my patch ;-)
gaal but i don't like how you have to install five modules just to test this nice thing you just d/led from cpan
kungfuftr nothingmuch: try now
nothingmuch Test::Unit should be one and only
woot!
stevan Test::Simple would have been Test.pm if the name wasnt alreayd there
kungfuftr nothingmuch: they all working? 21:09
nothingmuch it looks odd though
but we do have color
kungfuftr yeah, needs a little tweaking
nothingmuch italic
and boxed
twice
gaal any need to separate test::more and test::simple any, er, more?
nothingmuch Test::Plan is nice
gaal maybe test:simple can be a facade
nothingmuch Test::More is Test
Test::Simple is a subset 21:10
gaal yes
nothingmuch Test::Plan should be in Test too
gaal so expose it just as a facade
but don't dupe code
what's test plan do, remind me?
nothingmuch it's got an idiotic name
Corion pugscode.org <<Overview Journal Logs>> | r1773/winxp: 587/4256 smoke/win2k (occasional): xrl.us/fqum (25/4321 , r1980) | pugs.kwiki.org 21:10
nothingmuch but basically it's a quick way to do the 21:10
gaal ah ah skip all
yes i remember
nothingmuch BEGIN { eval { require Module::I::Need }; $@? plan skip_all : plan tests => 5 } 21:11
i would have called it Test::Predicates or something
but it should be core anyway
gaal yeah. skip needs to be made stronger, sure
exactly
nothingmuch Test::Block also helps skipping
by telling in advance how many blocks are per block
and then you say 'skip rest' and it gets the number right
Test::strict should be Test::Distribution 21:12
Test::Output should be in Test
gaal i'm still not 100% convinced i like having to specify a plan number automatically
Corion Somebody should port "release" / Module::Release!
gaal err
kungfuftr nothingmuch: okay, should look better now
gaal manually
kungfuftr stevan: look okay for you?
nothingmuch but it should overrid eonly when it's in effect 21:13
Corion Specifying a plan number is important, especially as long as Pugs / Perl can silently quit your test.
gaal yes, just i hate counting
i wish the computer would do it for me
nothingmuch reload nothingmuch.woobling.org/badkhtml.jpg
gaal: predeclaring is useful
gaal i think i had this discussion before, nm was it with you?
nothingmuch Test::Block just makes you do less arithmetic
no, not with me
but plans are optional, so it doesn't matter 21:14
this is for skipping
gaal like how i saw someone being clever, and declaring all his tests as closures and pushing them in an array
nothingmuch *cough*
gaal planning by @tests
then running them one by one
well look on the bright side
nothingmuch i did that twidce
MPEG::Audio::Frame did calc of plan
but not with closures
gaal if you get a parse error, it will die before you planned 21:15
nothingmuch and Object::Meta::Plugin did not use Test::Builder at all
gaal so the harness knows
the editor should perhaps have an "update plan" feature, if we don't go the clever route. 21:16
run the test, extract "looks like you only ran 2 tests out of 42" data, etc. 21:17
that's probably doable in perltest.vim
kungfuftr nothingmuch: now? 21:20
nothingmuch almost
now the border is OK 21:21
but there is a light grey
the green is only as high as the text
kungfuftr bah...
nothingmuch i think this is too workaroundish
you're working hard
and without much benefit
kungfuftr not hugely, i do this sorta stuff for kicks
ingy fyi YAML is very streamable
YAML.pm does not currently have a streamable interface 21:22
YAML and YAML.pm are very different things :)
kungfuftr nothingmuch: right... now? 21:27
nothingmuch between the test case lists 21:28
text is not vertical aligned, but all else is OK
oh wait
we have some odd 1px white boundries
but only after the second row
..ML--
kungfuftr bah... ffs safari sucks
nothingmuch wait, no, CAML not mm
--
HTML--
and ancestors, and derivs too
woot, YAPC::NA will be on my birthday 21:30
i think i'll try for a talk, maybe
vertical align ok 21:31
but we're back with whites
on top and bottom of percent cell
thin borders gone 21:32
mldonkey is misbehaving
cognominal ouch 21:40
pugs> my $b = my $a = 1 21:41
*** Error: No compatible subroutine found: &my
nothingmuch hah
cognominal pugs> my $a = 1
undef
my should be a function like in perl5?
kungfuftr nothingmuch: moo 21:45
nothingmuch is?
si?
kungfuftr that any better? 21:46
nothingmuch nope, the vertical-align is back
but the white lines are gone
kungfuftr vertical-align the only issue? 21:47
nothingmuch yep
kungfuftr yay
nothingmuch no wait, i didn't notice 21:49
but there's a margin between the cells
about 3px
clickxchange.com/graphics_smart/131...062312.gif 21:50
what *wont* they advertise these days?
are you depressed? click here, and we'll show you a 5 step guide towards feeling better
pasta:6969/ 22:01
gaal: ping
damnit! 22:06
Limbic_Region Janet - I love you 22:11
what - no RHPS fans 22:12
Limbic_Region does the Time Warp by himself
kungfuftr nothingmuch: et voila 22:19
if that don't work feck it
nothingmuch ack
borders screwed up
cells small
border-spacing between test cases is 0 22:20
vertical align of percent cell is wrong
safari--
kungfuftr++
kungfuftr bah, bugger it... would require a restructure of the dom 22:22
nothingmuch i think this is too much
the old mantra
if it ain't broke don't fix it
i think it applies here
kungfuftr it's not that something's not broke, it something that would have provided two interfaces 22:23
there's always xslt i suppose... *blink*
nothingmuch nonononono!
you can make Test::TAP::Visualize::XSLT if you want
i'm not touching that with a 10" swede
kungfuftr is there a TAP xml schema? 22:24
nothingmuch not that I know of
kungfuftr nothingmuch: oh yeah, you should also try to capture time taken for test 22:25
nothingmuch kungfuftr: test case?
kungfuftr nothingmuch: moo? 22:26
nothingmuch what do you want stamped? 22:27
it can all be
kungfuftr each sub test 22:28
and tesfile i suppose
bugger, all of it i suppose
nothingmuch subtest timing is a simple case of adding the data in the _handle_test stuff 22:34
btw, can you port your css to the original HTML?
or should i do that?
it's much prettier than mine
(when it works ;-) 22:35
Ovid Anyone know if "temp $var" is working yet in Pugs? I assume it's not. 22:53
jabbot pugs - 1982 - Could not make the method calls work in 23:35