pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | FBSD_4.11:R1124(120/2913) win2k:R1124(110/2913)
Set by kungfuftr on 25 March 2005.
jabbot pugs - 1184 - Added tests for anonymous roles and dest 00:17
pugs - 1183 - Added clone and constructor tests.
iblech Hi, just wanted to say that a) i must totally sleep now :) and b) I'll be away the next week (skiing) 00:21
jabbot pugs - 1185 - Added tests for properties. 00:27
iblech pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | FBSD_4.11:r1124(120/2913) win2k:r1124(110/2913) Linux:r1185(111/3191) 00:29
kolibrie anyone tried 'export TEST_LOG_FILE=test.log; make test; vi test.log'? 00:48
theorbtwo Allo. 00:49
kolibrie groetjes 00:50
theorbtwo kolibrie: Haven't tried it -- there's a test harness that outputs YAML, though, which is probably useful for more or less the same stuff. 00:52
autrjus, re the TODO list generator: have you tried looking at the testgraph thing? 00:53
kolibrie theorbtwo: is YAML the kind of stuff META.yml is made of?
theorbtwo (The dark green and dark red is TODO: green is "good", red is "bad". (Yeah, typically Western.))
Yes.
kolibrie which module is generating that? 00:54
theorbtwo YAML.pm, naturally. 00:55
kolibrie I mean, how is it harnessed to the tests? 00:56
theorbtwo perl util/yaml_harness.pl
kolibrie I'll look at it
theorbtwo That'll generate a tests.yml file.
kolibrie it looks like util/yaml_harness.pl reads the $*OUT from the test scripts and generates YAML from it 01:01
I'm wondering if I should just change the output of my 'test.log' to spit out YAML 01:02
theorbtwo Something like that.
kolibrie or maybe allow it to spit out several different formats
autrijus greetings from YAPC::Taipei!
kolibrie woohoo!
theorbtwo Salutations!
autrijus I just refactored hash to take strings as keys (by default for now) 01:03
kolibrie not quite sure what that means - translation? 01:04
theorbtwo It didn't before?
autrijus it didn't 01:05
kolibrie theorbtwo: not sure what autrijus did
autrijus my %hash = (1 => 2); %hash<1>
doesn't work
because 1 is Int and <1> means {'1'}
theorbtwo Ah.
autrijus try it. :)
theorbtwo Ah. 01:06
kolibrie so now %hash<1> will say 2?
theorbtwo And, I hope, the opp will also start working: 01:07
./pugs -e 'my %foo=("1" => 2); say %foo{1}'
autrijus right. 01:08
YAPC broadcast -- rtsp://140.110.17.6/yapc.sdp 01:09
kolibrie in what language?
autrijus this hour is chinese (clkao)
after that it's all english
with chinese interpreters
kolibrie my chinese is non-existent
theorbtwo When are you on? 01:10
autrijus tomorrow last session 01:11
# yapc.elixus.org/schedule.html
90 minutes, "Perl 6 The Ultimate"
(times are in GMT+8)
kolibrie what software plays .sdp? Haven't run across that before. 01:12
autrijus mplayer or quicktime
kolibrie thanks 01:13
theorbtwo Sadly, gxine does not.
kolibrie totem popped up when I tried, but then said it couldn't do it 01:14
autrijus :-/
vlc maybe?
theorbtwo Hmm, (g)mplayer seems to have problems of some sorts as well, but it looks like it's simply not seeing any data from it.
bd_ I got mplayer to play it on the second try 01:15
kolibrie vlc just spit out a whole bunch of junk on the command-line
stevan, you around? 01:22
hcchien vlc doesn't work for me.
autrijus I want clkao's slides :(~ 01:24
autrijus can't see any word
Khisanth need new glasses? :) 01:25
autrijus I don't have glasses :)
Khisanth ah well there is your problem!
autrijus XD 01:26
kolibrie binoculars come in handy
Khisanth field of vision is too narrow
autrijus the speakers should just upload their slides :) 01:27
kolibrie would be nice
hcchien I just ask the speakers upload the slides before their talk. :) 01:28
kolibrie autrijus: good luck tomorrow, in case I don't bump into you before then 01:33
autrijus :D
luqui ditto 01:51
to the good luck bit
and s/in case I don't bump into you before then//
autrijus =)
wagner.elixus.org/~clkao/runapp-intro 02:27
stevan autrijus: next release is 6.0.13? or were you going to go to 6.0.14? 02:38
autrijus 6.0.13. I have no problem with 13, thanks :)
stevan ok :)
isnt 666 good luck in china as well?
autrijus right 02:39
stevan actually both my bosses are chinese-americans
autrijus nice
stevan actually they own the company
:P
and I am the only other employee :) 02:40
autrijus nice
stevan small consultancies ++
ingy hi stevan 03:05
stevan hey ingy
ingy install problems?
stevan yeah I noticed that the Test.pm Makefile.PL installed into my /Library/Perl/5.8.1 as well as the /Library/Perl/6 03:06
I noticed this because I copied that Makefile.PL for the FileSpec one 03:07
and prompty overwrote my perl5 File::Spec install :)
_metaperl how do I init my svn repository on rt.openfoundry.org 03:15
autrijus _metaperl: what is your project name?
_metaperl terry@Abulafia:~/perl/hax$ svn import DBIx-DBH svn.openfoundry.org/dbixdb\
h/ -m 'initial import'
svn: PROPFIND request failed on '/dbixdbh'
svn:
Could not open the requested SVN filesystem
dbixdbh
the project name is dbixdbh 03:16
autrijus uh
you just use it
_metaperl but my svn import just failed...
autrijus try again?
it takes some time to initiate I think 03:17
_metaperl yes, now it worked... but I never explicitly initted it
stevan autrijus: how far along is the parrot and haskell backends? 03:18
still pretty primative?
autrijus very primitive 03:19
stevan ok
autrijus but I'll focus on haskell backend
and port all Eval.hs over
so we have on the fly compilation
and won't need to maintain two piece of logic
the parrot backend should also benefit from that. 03:20
stevan cool
_metaperl hi Ovid 03:23
autrijus hey Ovid
Ovid Hi people.
stevan hey ovid
Ovid Hi stevan. Out of curiosity, do you know if anyone is using the Class::Traits module? 03:24
stevan nothingmuch tried to use it once
but it wasnt really what he needed 03:25
other than that nope
I enjoyed writing it :)
Ovid Hrmph. Bummer. We have some things we might be using it for in Bricolage 2, so I was wondering if anyone else had success with it.
stevan I have fiddled with it, but it's a weird way to think of OO 03:26
and so I never found a good fit for it
and nothingmuch's idea was pretty insane, he basically tried to use it for cracked out mixins 03:27
Ovid ? Traits should be a nice replacement for mixins. What was he doing?
stevan mixins :)
Ovid Heh :) 03:28
Traits would have been perfect, then. I think mixins are only marginally better than the Interface crap they came up with in Java.
stevan he really needed to use multiple inheritence and mixins, and thats not really what Traits are
he needed traits that didnt flatten into the class basically 03:29
nothingmuch is crazy :)
Ovid Ah, OK. That would be an issue.
The "not flattening" part. Not the "crazy" part :) 03:30
stevan I would be happy to resurrect that project though
it was fun
ingy: what is the status of the Kwid stuff? I need it for the Changelog? 03:31
cm hi 03:33
autrijus stevan: svn log not detailed enough? 03:34
stevan autrijus: in some places yes, in some places no, and I have not been following the kwid stuff much though
autrijus nod. I think Perl5-Kwid can gen html 03:35
stevan ok
autrijus and there is also a Kwid Formatter in Spork (outside pugs)
which can gen Kwiki, S5 and other slides using Kwid
stevan yes I saw that
would you classify the inline stuff as experimental as well? 03:41
autrijus sure.
stevan or is that pretty far along ?
autrijus no, just color it experimental
the point of .13 is to push out experimental stuff so people can make them robust in .14 :) 03:42
stevan ok
are we dropping all 6.2.2 support after 6.0.13? 03:44
autrijus we are officially dropping 6.2 support on 6.2.
stevan ok
autrijus before that, there may be features that 6.2 won't run
stevan BTW - iblech ported 17 modules and wrote over 200 oo tests 03:45
iblech++
what does pugs --extern do? 03:46
nothingmuch good morning
autrijus stevan: it generates the haskell wrapper
stevan good morning nothingmuch
autrijus for a inline module
nothingmuch actually 03:47
good night
stevan nothingmuch: we are all on internet time here
its night for me, morning for you,.. I dont know what for autrijus
autrijus 11:47 noon
nothingmuch drinks his tea, and goes to read
well, it is morning, physically
just not mentally
for autrijus it should be around noonish
stevan 10:47pm here
nothingmuch 5:47 am 03:48
so anyway, today i hope to finish the linking of results to the cross index 03:49
stevan nice
nothingmuch and we have to generate index.html files per dir
depends on if i have to go to my luthier or not 03:50
well, we'll see
anywho, ciao
cm 10:50 pm, mhm
stevan you jamming to hard or something?
nothingmuch was semi drunk up to around 3 hours ago
a bit hoarse (good music)
my neck kind of hurts, from the cape 03:51
and my arm does too, from playing with the fairy wand
stevan ah yes isreali halloween right?
nothingmuch yup
we are celebrating the death of some persian-age hitler
mostly by eating cookies
named 'ears of haman' 03:52
a metaphore for the real ones that were allegedly cut off
it's also considered a 'mitzva', that is, a sort of good thing to do
stevan thought christian holidays were weird 03:53
nothingmuch as perscribed by god
to get drunk
en.wikipedia.org/wiki/Purim
en.wikipedia.org/wiki/Hamantasch 03:54
zzz &
stevan metaperl: you around? 03:56
_metaperl here 03:57
stevan do you happen to remember what numeric functions you impemented? 03:58
_metaperl well they are in the same place in Prim.hs so Icould find them
stevan so I dont have to dig really deep into SVN
_metaperl oh, you want to write tests?
hold on
stevan actually its for the changelog, but I will write tests too if they are needed
_metaperl atan2 03:59
cos, sin, sqrt
and exp 04:00
stevan excellent,.. thanks
_metaperl sure 04:01
stevan autrijus: r1186 has the new Changelog in it 04:35
autrijus: you might need to add some details that I missed
ingy: add any Kwid stuff too it as well
jabbot pugs - 1186 - adding ChangeLog for 6.0.13 04:37
stevan ok I will TODO the tests tomorrow, night all & 04:41
jabbot pugs - 1187 - * ARGS is the new ARGV 05:17
autrijus stevan++ 05:22
Khisanth what did the V stand for anyway? 05:36
bd_ Vector maybe? 05:37
Darren_Duncan 'Values' I think 05:38
It comes, I think, from the C days when the main() got 2 arguments ...
Khisanth well yeah
Darren_Duncan ARGC (count) and ARGV (values)
Perl didn't need the count, but kept the Values name anyway 05:39
Khisanth could be vector though :)
Darren_Duncan perhaps
vector is another name for array, isn't it?
Khisanth vector is a rather confusing term!
Darren_Duncan If you know math, it makes sense 05:40
'vector' is a term in mathematics that is a list of related values that act like one value
Khisanth if you are using Java or C++ you could be refering to the specific classes :)
Darren_Duncan eg, a direction in 2+D space has 2+ elements in the vector, one per dimension, and all together they give a single indication of the direction 05:41
Schwern Victory
Visitors
Darren_Duncan the math term 'matrix' is like a two-dimensional array, or a grid
Schwern ..._ 05:42
Darren_Duncan now, about my porting progress ... 05:43
I made a promise to have a perl-6ification up today
so to meet that promise, I'm going to leave the perl 5 updates half-finished and do them later
though I originally planned to do them first
the very first commit of the port will also lack a SYNOPSIS, but that should be added the next day 05:44
as a separate decision, my announcement to dbi-users will be made after the port, so the announcement can include that a perl6 version is available 05:45
working ... 05:46
autrijus tmoertel++ # crazy folk 05:58
in one day!
# community.moertel.com/ss/space/star...05-03-25/1
tmoertel blames the espresso 05:59
autrijus ooh just like me
tmoertel: will you compile ruby to parrot after 50 days?
or compile to haskell... or compile to pugsast
tmoertel: you know, with template haskell, you can just focus on the evaluator and it will magically become a compiler :D
tmoertel autrijus: no, I want to implement only the fun stuff :) 06:00
autrijus aw, I assure you all of it is fun :)
<- starts pugs only to model junctions
tmoertel <- has parsimoniously allocated tuits
autrijus you can get some imaginary tuits
or quarternion tuits 06:01
tmoertel hmmm... sqrt(-tuit); yes, that might work...
autrijus Kwid is the new Changelog format
hash string keys and scw's END{} fix are now in 06:06
Schwern++
err,
scw++
jabbot pugs - 1190 - * scw contributed a fix to make END {... 06:07
pugs - 1189 - * kwidify changelog
pugs - 1188 - * mad golfer should not have my chinese
gaal hi 06:10
autrijus hey 06:12
Schwern scw != Schwern 06:13
autrijus except my autocompleter brain cells only does two character lookahead
and you sort first :)
gaal heh (lucky i have a double-a deal in my name) 06:14
autrijus, re the command line, shall we do it haskell then? 06:16
autrijus you mean the parser? 06:17
sure, that'd be nice.
write a
gaal yes
autrijus parseArgs :: [String] -> Arguments
or rather
parseOpts :: [String] -> Options
and declare an option struct for things we care about?
gaal okay; the only problem is things like p5 -0 that can be specified a few times and basically have effect only until they are redefined 06:18
ie they have side effects
on the code that accumulates as they are parsed :/
autrijus that's fine. 06:19
you don't need to do it right now :)
gaal so this Options can contain part of a one-liner
autrijus yeah.
or evan a function
optSource :: IO String
gaal imagine that :)
autrijus which, when run, returns a string
either from readFile or just "return str" 06:20
gaal hmm, but there are things that are environment / more or less global, not part of the source String 06:22
autrijus sure, then add more opt stuff 06:23
gaal i'll take a look at how the current library does this, i have a lot of learning to do 06:25
autrijus that's fine :)
feel free to ask around on #haskell
or google for haskell-cafe archives about opt parsing
gaal ah, good, good. though perl probably outweirds most of what's foudn there :) 06:26
good lead though.
jabbot pugs - 1191 - fix enum test. 06:27
gaal &
jabbot pugs - 1192 - * asavige pointed out that nonblocky dec 07:57
autrijus obra's talk is about to begin 08:10
I'm the interpreter :)
autrijus> _ 08:11
Dabian ajeau! 08:13
Schwern Can you JIT obra? 08:14
Darren_Duncan has the talk started yet? 08:24
anyway, quick question ... 08:25
is Pugs no longer using a MANIFEST?
apparently not ... although it still has a MANIFEST.SKIP 08:26
anyway, the first SQL::Routine commit has been done ... 08:27
the translation is not complete, and I will resume it tomorrow
jabbot pugs - 1194 - SQL::Routine first commit - partial tran
pugs - 1193 - Locale::KeyedText - fixed a few minor Pe
Darren_Duncan meanwhile, I delivered something of it by the day I promised, Friday the 25th 08:28
GOOD LUCK on the talk, Obra and Autrijus
I look forward to the report tomorrow
08:45 ruuku is now known as luqui
crysflame autrijus++ # autrijus> _ 08:48
luqui autrijus> _ ?
scw luqui: 'autrijus>' is the prompt :p 08:49
hcchien interpreter. :)
luqui and the _ is? 08:50
scw cursor?
luqui uh
whatever
scw or maybe `everthing` since he's crazy on Haskell :pp
luqui I was thinking: 08:51
for 1... { die if $_ > autrijus }
that would give the halting computer a pain
hcchien hmm...., ingy may announce FreePAN project later. :) 09:23
Alias_ eep
I'd hoped to talk to him about it first
There's some... problems... with the concept
hcchien you can ping him now. 09:24
ingy hello 09:25
autrijus Alias_: I've voiced your concerns yesterday.
during ingy's b-day party
Alias_ hahahah
oopsi
I've got a few more to add
ingy autrijus: was I concerned?
autrijus type away
ingy: you were drunk
hcchien haha
ingy I had one beer
Alias_ My main new concern is this idea that the dist will exactly match what is in svn
autrijus you had insufficient sleep
gawd I made larry write S28 09:26
Alias_ My dists practically never match what is in my repository
autrijus www.nntp.perl.org/group/perl.perl6.compiler/460
Alias_ In fact, my layout is totally different
autrijus luqui: S28.pod? XD
Alias_ I have a package build script to "compile" a release
ingy Alias_: the dist will match your MANIFEST likely
at the time you tag a release
Alias_ MANIFEST? 09:27
I have no such think
thing
It's built at package compile time automatically
ingy I must not be understanding you
obra Alias_: _why_ do you do this?
ingy sorry I am very tired right now
obra what's the purpose?
Alias_ obra: A variety of reasons...
obra: Firstly, so all my devel modules are in one huge /lib 09:28
autrijus I'm cool with MANIFEST not in trunk.
(ingy is cool with that too)
Alias_ So I can work on improvements to several interlinked modules at once
Secondly for quality control
obra oh. so the problem is that svn doesn't have views yet ;)
for quality control?
Alias_ My build script auto-adds 99_pod.t and does various other sanity checking
autrijus nor does svk
but svk will :)
clkao: will svk have views?
Alias_ obra: Once Test::Inline 2 is done it will also automatically compile the inline tests
clkao it's half implemented. and some new fancy api broke that branch. but it is fixable 09:29
Alias_ In the actual directory for the dist, I generally have two files only. Makefile.PL and Changes
The rest is built
Next possible problem, how will Freepan handle unpack/repack? 09:30
autrijus aw, "twigil"
$@% are "sigils"
*?^ are "twigils"
ingy nice
Alias_ Having the release manager/indexer do things to people's packages
obra Alias_: will the new test::inline be 100% backwards compat
Alias_ obra: As clone as I can get to it 09:31
obra Alias_: it sounds like you'll want to do "releases" to FreePAN
Alias_ obra: It will drop a few very redundant things
obra worries that it'll break RT
Alias_ obra: The new Test::Inline has much better support for big projects
ingy Alias_: you don't have to use the autodist feature. Likely there will be support for you to build your own dists.
Alias_ obra: You can compile tests for the entire lib in one hit
My biggest T:Inline usage involves 59 test scripts and 5000 SLOC 09:32
ingy or maybe FreePAN just leaves exports tarballs of releases to CPAN, or something
TMTOWTDI
Jouke waves at the Y::T attendees 09:34
autrijus yo Jouke!
obra Alias_: Why do I care that I can compile?
Alias_ obra: Because the only thing that pod2test does is compile inline tests into test scripts?
obra er right. I don't want to check those into my repo 09:35
Alias_ Maybe bad use of "compile"
obra untill I ditch test::inlinethough I've needed a custom pod2test
Alias_ The downside of T::Inline 2 is it currently has a little dependency bloat 09:36
I hope to resolve that as I get closer to something more community friendly
obra when do you plan to release?
Alias_ When I get close enough...
It works fine for me in my uses ATM... the rest is just a gift to everyone, done as I have the time 09:37
But I won't do a "real" release until it's good enough
s/good/compatible/ 09:38
ingy and now I speak for FreePAN... 09:54
autrijus: will you translate?
autrijus sure!
ingy Alias_: I will give you commit access to the project :)
Alias_ ingy: It scares me too much at the moment :) 09:55
ingy Alias_: everything around here scares me
hcchien FreePan talk now 10:02
jabbot pugs - 1195 - Added examples/japh/madgolfer.p6 -- is t 10:17
Alias_ While we are on the subject of cons 10:19
ali.as/OSCON
jabbot pugs - 1196 - r2839@speights: samv | 2005-03-26 18:3 10:47
pugs - 1197 - cleanup/unification of command line hand 11:17
11:31 integral is now known as thaum 11:32 thaum is now known as integral 11:53 saorge` is now known as saorge 12:32 tsee_ is now known as tsee
Corion btw - Win2k:r1197(105/3116) 12:59
13:24 Aankh|Clone is now known as Aankhen`` 13:57 Aankh|Clone is now known as Aankhen`` 14:07 cnhackTNT|away is now known as cnhackTNT
scw Hello, I packed pugs into debian package :) 14:39
www.csie.ntu.edu.tw/~b92103/pugs
There are still several problems (path and something like that), I'll fix them tomorrow :) 14:40
14:49 cnhackTNT|away is now known as cnhackTNT
kungfuftr pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | FBSD_4.11:r1197(114/3115) win2k:r1124(110/2913) Linux:r1185(111/3191) 14:51
kungfuftr pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | FBSD_4.11:r1197(114/3115) win2k:r1197(105/3116) Linux:r1185(111/3191) 14:52
kungfuftr bah 14:52
integral mm, what's the two numbers? files failed / subtests failed? 14:56
malaire subtests failed / subtests total 14:57
integral can add OS X to that
obra hm. pugs has no sockets yet 15:06
malaire My testresults: Linux:r1197(105/3116) 15:07
integral hmm, 13 minutes, _slow_ ;-) 15:13
my results: MacOSX:r1197(105/3116)
pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | FBSD_4.11:r1197(114/3115) win2k:r1197(105/3116) Linux:r1197(105/3116) MacOSX:r1197(105/3116) 15:14
autrijus obra: Write Tests 15:22
test and ye shall receive features
ingy hola 15:27
autrijus: did you actually ever really get SHA1 working as advertised?
autrijus ingy: no, the NamedLoader thing needs another ocuple minutes 15:28
I will make it all work before (or during) my talk.
ingy aye
I want to move it into modules where it belongs 15:29
that will require major mods to the build system to install all the core .o files
So they can be built outside of pugs 15:30
and then we can move modules into freepan
I need to spend another day fighting MakeMaker
nothingmuch morning 15:35
autrijus ingy: you want to move SHA1 to modules? 15:41
ingy: like, you think it's not possible to make it install properly tomorrow? 15:42
s/install/make/
that is fine.
you can move at will
ingy autrijus: I don't want to move until later
autrijus k.
ingy let's try to make it work where it is tommorrow
that would be nice 15:43
autrijus k.
elmex woho 17:13
autrijus ? 17:32
elmex autrijus++ 17:33
your cool ;)
autrijus thanks. :) 17:34
jabbot pugs - 1200 - * mandel.p6 now produces the correct ima 17:37
pugs - 1199 - add my name.
pugs - 1198 - fix one test.
elmex bbl 17:39
kungfuftr autrijus: who owns jabbot? 17:43
autrijus gugod 17:45
stevan autrijus: I have TODOed all the tests
I am going to svn up it now
I added # unTODOme to the end of all the todo-ed tests 17:46
do some shell script wizard maybe can whip something up to automate this
jabbot pugs - 1201 - * switch to the C-line #line comment at 17:47
stevan and when all the tests run, we have 3198
autrijus thank you. 17:48
stevan ok r1202 - all the tests TODOed
autrijus I've just implemented Leo's fixes.
stevan no problem :)
autrijus perl5 is now:
stevan for Parrot stuff?
autrijus real 0m6.213s
user 0m5.695s
pugs is now: 17:49
real 0m5.652s
user 0m5.225s
yeah.
stevan :)
ok I am gonna be offline for the day, enjoy yourself
BTW - kwid changelog ,.. very nice 17:50
autrijus yeah, much more readable
thanks stevan, see ya!
kungfuftr autrijus: where's a good place to dive into the haskell code? 17:53
autrijus kungfuftr: depends on your interests...
AST.hs is the definition of everything
Prim.hs is most rewarding 17:54
kungfuftr autrijus: paranoia is my intrest
=0)
autrijus Eval.hs is the heavy weight.
so, it's all good :)
jabbot pugs - 1204 - * Prepare for release. 17:57
pugs - 1203 - * switched both mandel.p5 and mandel.p6
pugs - 1202 - TODOed all the tests in preparation for
kungfuftr bah, i need a faster machine
autrijus zzz & 17:58
elmex hi 18:15
jabbot pugs - 1205 - * add newline at end of file 18:17
pugs - 1206 - untodo one test, it's already ok. 18:37
kungfuftr pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | FBSD_4.11:r1206(10/3197) win2k:r1197(105/3116) Linux:r1197(105/3116) MacOSX:r1197(105/3116) 18:47
jabbot pugs - 1207 - fix examples test. '&>' will go backgoun 19:07
19:49 _metaperl is now known as metaperl 19:51 Aankh|Clone is now known as Aankhen``
elmex hi 19:54
boyz
where to find a real perl6 spec? 20:09
gaal t/Synopsys
kungfuftr elmex: dev.perl.org/perl6/ also handy 20:10
elmex gaal: the Synopsys are a bit ... 'fuzzy' ;0 20:14
gaal they are concise; they're considered the (incomplete) spec. perhaps the exegeses and apocalypes might come in handy? 20:15
Darren_Duncan Anyone know about how many hours I have left to squeeze in more commits before the 6.0.13 release cutoff? 20:18
vladtz and break a few test along the way :-) 20:19
Darren_Duncan all my commits in the next few hours are in /modules, so shouldn't break anything 20:20
they are just semi-critical to making the ports look a lot better 20:21
vladtz well it was just good to see all tests passed again after several days of finding out how to compile for i386 on a athlon-64... 20:24
kungfuftr pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | FBSD_4.11:r1207(0/3197) win2k:r1197(105/3116) Linux:r1197(105/3116) MacOSX:r1197(105/3116) 20:36
kungfuftr NB: while 0 tests fail, the goto test still doesn't compile correctly 20:36
pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | FBSD_4.11:r1207(0/3197) win2k:r1197(105/3116) Linux:r1197(105/3116) MacOSX:r1197(105/3116) **[793 tests TODO]** 20:42
gaal it doesn't? 20:42
i thought i'd fixed it last night 20:43
kungfuftr 2 secs
gaal compiles fine here
kungfuftr gaal: paste.husk.org/2890 20:44
gaal sure you svn up'ed? can you look at the file and see if most of it is commented out? 20:45
(from line 17 on) 20:46
kungfuftr 2 secs
yar... conflict from last svn up... goi
gaal we're no trying to adhere to 'all tests must run' policy - which means parsefails are tagged with a todo_fail() or a fail(), then commented out. 20:47
if you see a test blowing up, please fix it :) 20:48
kungfuftr heh, if i could... not enough haskellknowledge... =0/
gaal i meant the test :) though hey, fixing the code would be nice too :) 20:49
kungfuftr heh
would love to fix code... but has spirraled
vladtz kungfuftr: yes I know, lots of todo's 20:52
enough for now, bye
kungfuftr all tests pass on FreeBSD 4.11... yay! =0) 20:59
gaal: there's going to be a release very soon, aye? 21:23
anyone? 21:34
Darren_Duncan it should be on the 27th, Autrijus' time
which is later today the 26th, my time
I need a few more hours to get stuff in first 21:35
kungfuftr k, will commit an interim MANIFEST
Darren_Duncan why was the old MANIFEST removed?
It can be regenerated along with any commit using 'make manifest'
kungfuftr being updated too much iirc
Darren_Duncan since Autrijus may get up and do his thing at any time, I am meanwhile doing smaller commits less often, to lessen risk 21:37
jabbot pugs - 1208 - large Locale::KeyedText documentation up
Darren_Duncan more often I mean
kungfuftr yar, i think he's off to bed now
Darren_Duncan yes 21:38
its early morning in Taiwan right now
autrijus er. 22:47
don't do a interim MANIFEST. :)
I can gen one locally.
btw, hi. 22:48
kungfuftr np
noticed wasn't in svn anymore, so didn't
autrijus :) 22:49
kungfuftr can be sensible at time 22:50
Darren_Duncan hello Autrijus!
kungfuftr autrijus: go to bed!
Darren_Duncan question: in how many hours is the cut-off for getting /modules stuff into 6.0.13?
didn't he just get up?
kungfuftr autrijus: btw, is there anything in place for .xs stuff at the moment
Darren_Duncan: 4 hrs sleep does not count 22:51
autrijus Darren_Duncan: I don't know, how many you need?
Darren_Duncan as many as I can get ... or I suppose 3-4 hours 22:52
autrijus that's fine.
it's 7am here.
I'll do release at noon.
Darren_Duncan I can give new estimates as time goes by
autrijus sounds like a plan?
Darren_Duncan sounds ok so far
I'll let you know closer to that if I'm running short of time
autrijus I need to do interpretation and my real talk afternoon
sure =) 22:53
will you be committing new files so I need to regen manifest?
Darren_Duncan meanwhile, I'm doing smaller commits more often, to reduce risk
kungfuftr Darren_Duncan: what platform you running on?
Darren_Duncan I'm on Mac OS X 10.3.8 currently, 10.4 when it comes out 22:54
autrijus k, cool!
Darren_Duncan rumors say mid-April
kungfuftr Darren_Duncan: can you update the end of the topic line?
Darren_Duncan I never tried to do that ... can a non-admin do it? 22:55
autrijus all can.
kungfuftr yup
there's no +t on the channel
Darren_Duncan I'm using the Colloquay client
quy
kungfuftr Darren_Duncan: you can paste in chan, someone else will update
Darren_Duncan I'll explore the menus in case there's something about the topic line
is there something you want me to put in the topic line for you? 22:56
kungfuftr autrijus: an .xs stuff at all? or easy was to port .xs based modules?
Darren_Duncan: MacOSX:r1197(105/3116)
<platform>:<revision>(<failed>/<total>) 22:57
Darren_Duncan do you want me to try a rebuild?
kungfuftr yar, from latest revision (1208)
Darren_Duncan wait a few minutes
kungfuftr np 22:58
Darren_Duncan starting 'make' now ...
starting 'make test' now 23:04
autrijus kungfuftr: there will be inline support for C real soon
External/C.hs
kungfuftr autrijus: nice 23:05
autrijus: anyone thought of tackling a DBI port?
bollocks... bigger than I initially though... hhhmmm... 23:06
autrijus er, port DBI::PurePerl
that's all you need
kungfuftr ah 23:07
Darren_Duncan makes sense to me
make test is still running ... fyi, I'm still on GHC 6.2, not 6.4, if that matters 23:08
kungfuftr autrijus: ah! thanks, will have a peek in the morning
autrijus Darren_Duncan: that's good.
although if you go up to GHC 6.4
you'll have a much quicker test turnaround :)
kungfuftr autrijus: ghc 6.4 in ports yet? 23:09
Darren_Duncan because its faster, or it skips more tests?
pjcj autrijus: Would I be right in thinking that if I were to walk the AST and hook into reduce I would be able to produce a p6 coverage tool? 23:10
Or maybe there's a better way?
autrijus kungfuftr: yup 23:12
kungfuftr has another peek
autrijus pjcj: yes and yes. 23:13
pjcj: if you outline the things you need
then I can probably add hooks for you ast part of my Eval.hs refactoring
in a way that doesn't hurt any performance if not used.
send to p6c :)
Darren_Duncan okay, there is good news and bad news re make test 23:14
pjcj OK, I'll do that. When are you doing the refactoring? ie, do I need to do it now?
autrijus no, today is 6.0.13
refactoring is next week
then 6.0.14
and may or may not be a 6.0.15
Darren_Duncan good news is, 'All tests successful, files=144, tests=3198'
autrijus then 6.2.0.
pjcj OK, I'll have a think and send off some mail tomorrow then.
autrijus danke
happy thunking
pjcj but first, sleep, I think 23:15
Darren_Duncan then it appeared to try running another test suite and died
citing file "t/1-basic.t"
will pastebot it
pasteling "Darren_Duncan" at 24.108.164.7 pasted "result of 'make test' on Mac OS X -- good news, bad news" (15 lines, 1.2K) at sial.org/pbot/8656 23:17
kungfuftr Darren_Duncan: i assume it's a test for ext/Test/t/1-basic.t
autrijus ingy will be around soon.
is the "make install" overwriting perl5's Test.pm bug fixed? 23:18
I'll check both with him.
kungfuftr it's because pugs isn't installed in that path
autrijus it shouldn't need to.
it should respect HARNESS_PERL 23:19
Darren_Duncan kungfuftr, probably, and I pasted the details anyway
anyway, because of the death, I didn't get any (N/M) result
kungfuftr autrijus: i get similar here
kcwu after make clean, src/External/Haskell/*.o havn't be removed .. 23:20
Darren_Duncan Note that I did the initial Makefile with the system Perl 5.8.1 but that should be acceptable
Although I did get a "Warning: prerequisite ExtUtils::MakeMaker 6.15 not found. We have 6.12."
autrijus Darren_Duncan: try svn up and again 23:21
Darren_Duncan okay
now doing 'make' again 23:22
kungfuftr does the same
Darren_Duncan now 'make test' again
autrijus kcwu: fixed. 23:23
r1210
elmex perl6 is dead
obra long live perl6
kungfuftr elmex: no no, it'll be here by christmas
elmex i don't belive in miracles 23:24
Darren_Duncan I'm on 1209; should I kill the test and up again?
obra wrote a perl6 script last week
autrijus you don't need to believe when you can actually know.
Darren_Duncan: no need to
Darren_Duncan ok
obra has no faith.
obra has something better
kungfuftr obra: leet hacker perl skills? 23:25
autrijus prescient ability?
obra visible proof
obra waves at his proof
kungfuftr obra: tatoo of a camel?
autrijus particles back 23:26
jabbot pugs - 1210 - * kcwu reported that we should clean src 23:27
pugs - 1209 - * attempt a fix at the make test bug dar
kcwu does make tags works for you? i got "hasktags: src: openFile: inappropriate type (is a directory)"
autrijus kcwu: fix please :) 23:28
kcwu autrijus: ok, just make sure it's not my own problem :) 23:29
Darren_Duncan autrijus, it now has no errors and says all tests successful 23:31
kcwu err, seems that vim doesn't like tags file generated by hasktags :( 23:35
jabbot pugs - 1212 - hasktags only likes files, not directory 23:37
pugs - 1211 - minor LKT 5v1.03 sync update, ignorable
stevan autrijus: any way to make split support regexp quickly? 23:42
kungfuftr stevan: you can do split with pure perl5 regexes 23:44
stevan kungfuftr: how :) 23:45
kungfuftr stevan: my @foo = $string =~ /<insert voodoo regex here>/; 23:46
stevan :)
jabbot pugs - 1213 - * tags file should depend on source file 23:47
stevan ok, hopefully I can make the regexp Voodoo work in pugs :)
kungfuftr it requires doing some zero lookahead assertion and use of the g modifier
stevan kungfuftr: my regexp fu is only so-so
I am trying to do this: split /([A-Z]<\s+|[A-Z]<|\s+>|>)/, $text 23:48
any thoughts on how to convert it? 23:49
kungfuftr 2 secs, i'll have a think
stevan cool
kungfuftr stevan: can you give me a copy of the text?
stevan yeah 23:50
perlbot nopaste
perlbot Paste your code here and #<channel> will be able to view it: sial.org/pbot/<channel>
pasteling "stevan" at 67.186.136.119 pasted "split -> to -> //g" (10 lines, 328B) at sial.org/pbot/8658
stevan kungfuftr: BTW - I am doing it in perl5 right now 23:52
to make sure my logic is right
kungfuftr yar, will do same here 23:53

Powered by Cro and the Raku® Programming Language.

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

Thank you!