pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | FBSD_4.11:r1302(183/3383) win2k:r1324(187/3474) Linux:r1302(193/3383) MacOSX:r1197(105/3116)
Set by Corion on 29 March 2005.
jabbot pugs - 1332 - Crummy test fixed 00:07
theorbtwo ping. 00:20
jabbot, seen theorbtwo 00:27
jabbot theorbtwo: theorbtwo was seen on Wed Mar 30 08:20:20 2005
theorbtwo Aaah, it's alive! 00:38
Limbic_Region now if only it would learn to put the toilet seat down when it was done 00:39
jabbot pugs - 1333 - Perldoc::Sender utility pass tests 01:07
01:11 theorbtw1 is now known as theorbtwo
jabbot pugs - 1334 - * set mime 02:07
autrijus hi lambdacamels! 02:14
journal is up; power/network outage yesterday :-/
theorbtwo Allo, autrijus!
autrijus how's pugs? 02:15
autrijus starts backlogging
stevan hey autrijus
autrijus yo :) 02:16
stevan hey theorbtwo
autrijus 15-hor-sleep++
hour, even
finally out from this post-conference syndrome tarpit
theorbtwo Allo, stevan! 02:17
stevan I just spent the entire day in meetings with the client and driving 02:18
no fun at all
Khisanth autrijus: so when is the next tarpit scheduled? :p
autrijus Khisanth: yapc::na I think
theorbtwo autrijus: Before or after, and how long?
autrijus theorbtwo: before, 4~5 days, to the zeroth approximation 02:20
theorbtwo Great!
I've been trying to schedule my vacation, and have been missing data... 02:21
autrijus :) 02:22
currently only the location is unsure
the date should be just that
theorbtwo Cool.
Any idea who I should ask questions about the rooming to? 02:23
autrijus I'm supposed to write to ask John Macdonald 02:24
john at perlwolf dot com
theorbtwo nods.
autrijus he offers hosting at a place 40 mins drive from toronto
but not sure whether we have net or not there 02:25
theorbtwo Oh, I was hoping to be in the dorms.
autrijus that is also possible.
stevan theorbtwo: I think they have some kind of dorm/summer residence available
but maybe only for the conf days
theorbtwo Yeah, but I'm not real clear that it'd be possible to be in the dorms before/after the conf. 02:26
autrijus etc. as I'm now in hotspring mode (meaning, spotty network), I'd be absolutely thrilled if you can help me figure out an optimal location with rdice and john and other YAPC::NA crew :)
theorbtwo I suspect the dorms have 802.11, or if not, somebody in the dorms will make them do so for the duration.
obra bets one could find a nice, inexpensive b&b nearby
stevan theorbtwo: you can hide out on the common rooms and just hide when the RAs come around
theorbtwo Also, my sister thought it unlikely that I could get a double with Jess, because of gender issues. 02:27
I expect she's wrong, but I rather dislike the possibility.
And budget is a large concern for me on this.
stevan 89chestnut.com/ 02:28
thats what they are linking to on the site
theorbtwo Right; they link to 89chestnut.com as the place where it's being held, and give rates.
...and say that dorms will probably also be available, for a lot less. 02:29
stevan I thought 89chesnut was the "dorms"
theorbtwo I'm not sure.
stevan reads the fine print this time 02:30
theorbtwo From YAPC last year, I think I want to stay in the dorms this year.
stevan yapc.org/America/accommodations-2005.shtml
theorbtwo (Last year I stayed with somebody who lived in-town.)
stevan oh well, I am off to watch some TV, see you all later 02:31
theorbtwo has never experinced dorm life, much less geek-dorm life. 02:33
chip Dorms are like prison cells, only without the food service. 02:38
theorbtwo It's only about a week. 02:39
And it's a chance to clustergeek with a lot of other people.
Remind myself that I'm smarter then a bunch of them, hopefully. 02:40
chip theorbtwo: YAPC is a wonderful place to confirm self-image, good or bad. 02:43
theorbtwo Hm, that's too bad, as I want my self-image improved, not confirmed! ;) 02:44
Darren_Duncan question: If a Perl module is distributed with Perl itself, such as Test::More or Test, is it worthwhile to mention said module in the Makefile.PL as a dependency, or better to not mention it there? (Assume that the same Makefile already specifies a minimum Perl version that is known to bundle it.) 04:17
theorbtwo I consider it worthwhile, but it messes with the ppm packager. 04:18
Darren_Duncan I haven't looked around, but what is the standard CPAN approach, considering Test::More is a defacto standard now
fyi, I'm doing a coordinated v5/v6 update that sees the v6 using 'Test' the same way that v5 uses 'Test::More' 04:19
(or technically the reverse, since v6 used Test first)
would you say it is better to leave the dependency mention out so to help those with ppm? 04:20
I think I will exclude any Test::More/Test mention from Makefile.pl then 04:23
Khisanth thinks it would be better if ppm was improved upon instead :)
theorbtwo So do I. 04:24
Darren_Duncan still, my main question is, what do most CPAN modules do?
distros
theorbtwo I don't think it'll happen, though, unless somebody takes it over from AS.
...which will only happen with kicking and screaming.
(Since newer versions of PPM try to get you to buy extra stuff from AS.) 04:26
Darren_Duncan I'm switching to Test::More etc in an effort to standardize, relative to rolling my own (which only takes about 5-10 lines)
therefore, I prefer to adopt everything related to it as is standard/recommended practice
theorbtwo I have no idea what standard pratice is; I only have a couple of modules, and they don't use MakeMaker. 04:27
Darren_Duncan theorbtwo, can users still do the standard 'make', 'make test' etc with your modules? 04:28
theorbtwo Yeah, I ship with a Makefile.PL as well as a Build.PL 04:29
OTOH, AFAIK, my modules have no users.
Darren_Duncan I don't really know if my modules have users 04:30
I get the occasional email about them, but thats about it
Khisanth no report from testers? 04:31
Darren_Duncan I get automated CPAN tester reports for all my distros, but that doesn't mean anyone is actually using them 04:32
theorbtwo Some reports from testers, Khisanth, but I'm not sure any of those are actually /users/, or if they just dl'd to test it... I hear some poeple do that.
Darren_Duncan right now I'm just looking at the random cpan module to see how they handle test::more etc 04:33
actually, I'll just do the simplest thing and not mention Test::More in code except in the *.t files themselves 04:35
my Perl 5 stuff requires 5.008001 and above anyway, and that includes the module, so I implicitly state the requirement anyway 04:36
jabbot pugs - 1335 - adding Makefile.PL.local and the perlpod 04:37
Darren_Duncan and likewise, my Perl 6 stuff simply mentions Test.pm (a Test::more clone) only in the *.t file, and thats bundled with Pugs
Khisanth is there some policy that basicallys states "once a module is in core, it can never be taken out"? 04:38
theorbtwo Yes, Khis.
Except in the 5->6 jump. 04:39
Darren_Duncan I think that's just what tends to happen
yes, in Perl 6 everything should be dropped by default
theorbtwo too.
Darren_Duncan and modules will only be bundled on a case by case basis, I think should happen
stevan Darren_Duncan: what are you wondering exactly? re: test::more 04:40
Darren_Duncan certainly, doing it just because the perl 5 versions were bundled is a baaaaaad idea
theorbtwo Larry has suggested that we stick everything from p5's core into p6's core but in a namespace that encourages you not to use them.
Darren_Duncan Stevan, yes
theorbtwo (So the translator can just translate them like that.)
Darren_Duncan I'm following your suggestion of using Test::more in my perl 5 distros
as a side effect, it helps me make my perl 5 and 6 versions the most alike to each other
stevan but are you wondering whether to put Test::More in the Makefile.PL? 04:41
Darren_Duncan the more similar they are, the easier to maintain two parallel versions
yes
stevan I think most people do
Darren_Duncan that's partly what I wanted to know
what most people do, and what best practice is
stevan My feeling is that if it is a dependency, it should be in the Makefile.PL 04:42
Darren_Duncan question: when will we be able to write our Perl 6 module Makefile.PL in Perl 6?
stevan regardless of whether it is just for tests or not
Khisanth Darren_Duncan: now? :)
stevan Darren_Duncan: when someone ports the approriate modules :) 04:43
Khisanth MM is not ported yet?
Darren_Duncan Pugs::Makemaker is a perl 5 module
stevan there is a Module::Install::Pugs, but that is perl5
Khisanth or Module::Build
stevan sorry, Darren_Duncan your right Pugs::MakeMaker
Darren_Duncan okay, here's the thing ... If my Perl 6 Makefile.pl mentions perl 6 modules in its PREREQ_PM, will pugs::Makemaker do the right thing or look for perl 5 modules? 04:44
stevan Khisanth: my guess is that Pugs does not currently have enough of a feature set to support it
Darren_Duncan: are you thinking of releasing the Perl6 verson on CPAN?
Darren_Duncan eventually
stevan ok 04:45
well hopefully by then the issue with be resolved
Darren_Duncan right now it's premature
stevan until then,.. perl6 depends upon perl5 :)
Darren_Duncan another question: is it useful at all for any /modules in Pugs to include a Makefile.PL?
stevan Darren_Duncan: not right now 04:46
although I am pretty sure make will never try to run it anyway
Darren_Duncan the sample module did it, so I followed suit
maybe I'll just take those out
jabbot pugs - 1336 - first commit of CGI::Pugs; why? I have n 04:47
stevan Darren_Duncan: can't hurt to leave it in 04:48
up to you really
Darren_Duncan I suppose that if it won't be executed, it doesn't matter if the contents don't make sense per se
such as perl 6 requirements in a perl 5 makefile]
here's what I'll do for now ... 04:50
I will keep the Perl 6 makefile, but it won't state any dependencies
It can just be used to generate the likes of META.yml from, having info like dist name and version 04:51
jdv79 stevan, OT but have you used IOC for anything publicly visible?
looks intersting
stevan jdv79: nothing public, I do mostly intranet work 04:52
but it just got launched into production last week
jdv79 solid so far?
stevan been running without issue for the past 3 months of development
jdv79 nice nice, I will get back to you about DBD::Mock eventually - been out of it lately 04:53
stevan jdv79: not a problem, I have been doing Pugs so much lately I have been neglecting my perl5 stuff :)
jdv79 NO! 04:54
its for the greater good though - anything for p6 is cool with me:)
stevan yes, p6 is very cool :) 04:55
Darren_Duncan for internal consistency, I'll change my Makefile.PL to Perl 6 code, pretending that Pugs::MakeMaker is a Perl 6 module 04:56
then all files under /modules are perl 6
05:03 theorbtw1 is now known as theorbwto
stevan thinks theorbtwo might be having a personality crisis :P 05:04
theorbwto Just multiple machines, one of which tends to have it's bluetooth thiny fall out of the USB socket, and one of which was rebooted to (attempt to) play some NWN. 05:05
hattmoward That boy couldn't hold a TCP session if it were taped to the wire.
;)
theorbwto Oh, hi, hatt. 05:06
hattmoward Hello, Mr. WTO
Darren_Duncan okay, I committed my Makefile changes 05:07
only under my own modules, in case other people don't like it happening with theirs 05:08
but I can do it for the others if it is desired
theorbwto Oh!
Darren_Duncan my /modules stuff is now 100% perl 6 in intent
theorbwto You were talking about the makefiles for the p6 modules, not for the p5 modules?
Darren_Duncan I was talking about both 05:09
theorbwto Oh, right, you want them as close as possible.
Darren_Duncan correct
in the case of pugs, it looks silly to have perl 5 code under /modules anyway
Khisanth theorbwto: tried duct tape?
Darren_Duncan CPAN serves that purpose, it hosts both
stevan past my bedtime, night all 05:10
Darren_Duncan my next CPAN uploads of my perl 5 modules will use Test::More and say so in the Makefile
theorbwto Khisanth, haven't, but I'm not sure that'd so much work...
Khisanth I meant for holding the bluetooth antenna in place 05:11
theorbwto I knew that. 05:12
But sometimes I want to take it out.
Darren_Duncan so, any objections to my changing the Makefile.PL of all other /modules to Perl 6?
said change would assume a perl 6 port of Pugs::MakeMaker would be out to be used
actually, I'll post that matter on the ML instead for discussion, re making /modules/*/Makefile.PL into Perl 6 code, and I'll leave the other distros alone 05:15
jabbot pugs - 1338 - CGI scripts can be tested on the command 05:17
pugs - 1337 - Makefile.PL is Perl 6 now for LKT and SR
sdtr443w OK back to that readline detection problem while building PUGS... 05:24
Darren_Duncan okay, the posting to p6c is done
re Makefiles
sdtr443w I had to do some mucking around to get this to build on my SuSE box. Is the Perl 6 newsgroup the best place to document all the steps? 05:37
theorbwto perl6-compiler
liyuray window log
gaal morning! 05:38
theorbwto Mornin, gaal! 05:39
gaal getting some tea, brb
sdtr443w I should be specific -- I'm talking about PUGS 05:40
Building it seems to get so hectic that is causes me to lose my connection to the machine. Right now I'm trying a nohup run in a hope that it won't kill itself when I get disconnected. 05:43
gaal sdt.*, do you know about screen? 05:45
sdtr443w I guess not
Sometimes I wonder how I got this far into Perl -- given my lack of UNIX
Oh hey it worked this time. Jesus is Lord or something.
Finally! A pugs executeable! 05:46
gaal www.gnu.org/software/screen/ - incidentally it solves the nohup problem but is also a great tool for other things
congratulations :)
05:47 theorbwto is now known as jesus
sdtr443w Ahh I just typed screen on the command-line and now it's doing strange, strange things! 05:47
jesus Thank you, my son, but I didn't have anything to do with it.
05:47 jesus is now known as theorbtwo
theorbtwo Hm, somebody has the username jesus registered with nickserv. 05:48
sdtr443w Terry Schiavo
gaal Darren_Duncan, still around? 05:50
Darren_Duncan yes 05:51
gaal you really want a port of MM? :)
Darren_Duncan I was working on my P5 test suites, upgrading to test::more
gall, in theory
gaal
sdtr443w Is there a Perl 6 compilers newsgroup? I don't think my newsgroup server carries it
I have the internals and the language groups
Darren_Duncan perl 6 compilers is an email list
justatheory checks his gall
gaal Better port every candidate p5 module to Module::Build first 05:52
sdtr443w Oh snap
Darren_Duncan it is the same as language and internals
gaal and then port *that* to p6 :)
sdtr443w oh
Darren_Duncan whatever, I just think that the Makefile.PL for perl 6 modules should be written in perl 6
gaal Makefile.PL should *die*! :) 05:53
Darren_Duncan there just needs a solution that is easy to use and standardized
gaal and is cheap and we need it fast, right? :)
anyway, MM is probably not it
theorbtwo I'm all for Module::Build taking that place.
In any case, something that does not require as much outside of perl. 05:54
(IE make.)
Darren_Duncan if such a module exists in perl 6, I'm all for updating the /modules ports to use it by default
gaal it doesn't yet, but perhaps someone could be talked into porting it (== Module::Build)
or coming up with something entirely better! 05:55
Darren_Duncan meanwhile, I'll invoke Pugs::Makemaker ... the nonexistant perl 6 version
new and better is good
that's what perl 6 is all about
gaal in that case may i suggest you use a different name
Darren_Duncan and things associated with it can learn from the example
gaal because it shouldn't be coupled to pugs
and it shouldn't bow down to mm's legacy 05:56
Darren_Duncan want to suggest a placeholder name?
gaal Instill?
Instate?
Present?
Darren_Duncan how about 'TheGreatModuleInstaller'
theorbtwo I disagree with "tied to pugs", BTW.
Shouldn't even be tied to perl too much.
gaal Pugs::Present?
Darren_Duncan since its a placeholder, a level-1 name is probably best 05:57
gaal Build? :)
Darren_Duncan like your first 3 examples
sure, I could go with Build
theorbtwo I like Distro::Instill
Darren_Duncan here's another legacy thing, should the '.pl' be uppercased?
theorbtwo Possibly even Package::Instill
gaal i never liked it 05:58
theorbtwo No, but I like beginning with a cap letter.
Darren_Duncan doesn't "instill' mean something else
gaal didn't make much sense in incasitive filesystems, either.
theorbtwo (makes it sort toward the beginning.)
gaal _______Build
Darren_Duncan no
theorbtwo "instill" is a valid English word, but the meaning works decently well here.
I like package more then module; should be usable for installing things other then modules. 05:59
Darren_Duncan it is an english word, but I thought the meaning meant something different, such as "to put into"
or maybe "to put into" is appropreiate
gaal www.m-w.com/cgi-bin/dictionary?book...=0&y=0
problem with that word is you'll have people cursing you when their fingers type 'install' instead. 06:00
theorbtwo 1;
Darren_Duncan here's a separate but similar question ... what should we use for the !# lines in Perl 6 scripts and modules under /modules ?
theorbtwo #!/usr/bin/perl6
I think.
Darren_Duncan I"m currently use !#pugs
I mean #!pugs 06:01
theorbtwo But IIRC, Autrijus said it shouldn't be named perl6 until and unless Larry blesses it.
gaal anyway *i'm* pretty sure i want "that pre-installation thing you run" thing to do other things besides install
e.g. package a distro PAR
Darren_Duncan I also thought Perl 6 would simply end up being named 'perl' anyway
gaal #!pugs sounds good to me
anyway these are mostly modules and don't need a shebang... 06:02
Darren_Duncan so then we stay with #!pugs for now, and change it later when a perl binary exists?
back to the Makefile replacement ... 06:03
theorbtwo I would like the "standard" thing to be able to:
Darren_Duncan I've never used Build.pm ... does it work without 'make' entirely?
gaal i actually prefer Instate or Present (or Build :). Instill is cute and punny, but to cute.
yes.
theorbtwo Create .debs, create .rpms, install directly, create some package for installing on windows boxes, and be able to install it...
Khisanth .msi?
gaal theorbtwo, and also build. :) 06:04
Darren_Duncan Deliver ?
gaal :)
Darren_Duncan that's kind of generic, handling both 'install' and 'build' and 'packaging'
theorbtwo building is part of creating a binary package. 06:05
gaal But I never had a gripe against Module::Build. has anyone? Let's at least look for shortcomings before going off to do something else?
theorbtwo I like Deliver.
I also like Module::Build.
gaal so we can just port it, no?
Darren_Duncan and who says 'build' has to mean binary?
gaal adding the .deb .msi etc capabilities as required.
Darren_Duncan how complicated is Module::Build ... porting easy? 06:06
theorbtwo Sounds good to me.
gaal these things are never *easy*
system dependent, etc.
actually, i think they'll be made a little bettter with that fancy, what was it, Class::Path? that there was some talk about.
Khisanth gaal: don't forget .ppd :) 06:09
gaal that too
theorbtwo ppd? Should just die.
We should replace it with something that doesn't suck.
gaal par par par par par 06:10
gaal kinda likes par :)
theorbtwo AFAIK, PAR is almost completely not what we want.
sdtr443w Is there a source for a bunch of very, *very* simple Perl 6 programs to dabble in?
gaal a par *target* is one of th things i want.
theorbtwo But par of a module makes no sense, only par of an executable. 06:11
Khisanth Build par?
gaal sdt.*, see the t/ and modules/ directories.
sdtr443w I guess hanoi.p6 in the examples path of PUGS isn't so bad
Darren_Duncan would anyone object if I injected Module::Build into /modules so we can start porting it?
gaal of course it does! think deployment
sdtr443w Oh that's better
theorbtwo I am thinking deployment.
sdtr443w Once I figure this all out, I should submit an entry to the docs on how to get pugs going once it builds 06:12
gaal okay: you have a big project
theorbtwo I haven't used par, possibly it does something quite different then what I thought.
gaal part of it has dlls and stuff
XS and "things that require a compiler" stuff.
Khisanth ah you want to just Module::Build for things besides modules :) 06:13
Darren_Duncan since its thought that module::build is the best point of departure, I thought of putting it in to hack on ... however, I can skip the parts that are specific to XS
gaal you develop it, and want to give it to the customer
Khisanth something like PAR would be nice as a core module ... 06:14
gaal so you do it on your box, which is (say) linux 06:15
move over to your staging, which is the same solaris as the client's
add solaris' version of the shared objects to the par
run tests there too
Darren_Duncan okay, I'll ask the opposite question ... who DOES want Module::Build as-is ported? I won't start the process without a yes
gaal and throw the par at the client.
theorbtwo . 06:16
gaal next you want to send a fix
you find the bug, it's probably in your code
you make a par only of the files that have changed
Khisanth, here's a PAR target for M:B: perlmonks.org/?node_id=437941 06:17
don't have to go through the anoyying build process, and don't need to give a big package to the client
Khisanth gaal: there are more useful things to have before that :) 06:18
gaal just put the patch earlier in the search path
Darren_Duncan actually, nevermind the Build.pm port for now ... it has a huge number of dependencies
theorbtwo hmmms.
Khisanth well merging two par probably isn't too difficult ... 06:19
theorbtwo There was a fairly sizeable thread this on p6-[lc] a while ago.
Khisanth so you can give the client a patcher too :)
gaal Khisanth: there are many things to do and i suppose people will work on what they have foudn important for them in the past
to2, i'd be grateful for a link 06:20
PAR doesn't rhyme with JAR for nothing, y'all know. 06:21
Darren_Duncan their both related to pirates
gaal Arr.
Darren_Duncan Aaarrrr
Gyeaar 06:22
theorbtwo My machine is acting funny, so I doubt you'll get it from me.
Khisanth they are both zip files :) 06:25
gaal that's like saying "it's just XML" :)
theorbtwo ...only more so.
gaal i need to get going to work, bbiab 06:26
Khisanth hmm P 06:27
hmm PAR as an XML file?
Darren_Duncan ... and a lot of people joining at once 07:04
nothingmuch morning 07:22
Darren_Duncan morngi
jabbot pugs - 1339 - trivial LKT/SRT test update - nothing to 07:37
nothingmuch *sigh*... etherwake was not configured 07:46
no eris this morning
liyuray help load 08:21
castaway mornin 08:26
nm, you awake? 08:27
nothingmuch is awake, si 08:28
nothingmuch adds 'nm' to popup words
castaway *g* 08:29
can you confirm that the codingmonkeys subetha test thingy is dead? 08:30
nothingmuch nope, works for me
castaway hmm, telnetting to it doesnt, from here 08:31
(or attempting to connect with actual beep type protocol 08:32
I wonder if they banned our subnet..
wow, it works from perlmonk.org.. I bet they banned us somehow.. 08:37
barstewards
nothingmuch how will you fix it? 08:38
castaway I dunno yet..
nothingmuch perhaps you can use nylon on pm?
i can let you use nylon on my box at home, too
castaway nylon? 08:39
whats nylon? 08:40
I think we need to test more locally, else we'll just get other hosts banned too ,)
we got a mac mainboard off ebay, so maybe we'll have a mini test machine soon 08:41
nothingmuch sorry 08:57
nothingmuch is back
castaway no worries 08:58
nothingmuch nylon is a proxy
mac mainboard?
monkey.org/~marius/pages/?page=nylon
castaway motherboard, bunch of circuits, wires and such, y'know :)
nothingmuch it knows how to tunnel plainly, and also do socks
how much of it?
cpu + board, and that's it? 08:59
castaway I could just use ssh tunnelling
nothingmuch because macs are pretty propietry
nylon is like shell redirects for netcat
ssh knows to do more advanced stuff
castaway havent looked all that closely.. as far as I understood it, all it would need would be monitor, keyboard, mouse and atx supply
we have a monitor adapter
nothingmuch in terms of how things are tunnelled
ah, that's good
castaway I'm wondering how stable SEE is, for packet testing .. 09:01
I dont particularly want to get another host banned 09:02
Is yours reachable? ,) 09:03
oh, and probably a diskdrive of sorts ,) 09:04
nothingmuch it can work on another port, and be always on 09:08
my SEE is not open
like codingmonkeys
i'm at work
i can set one up when I get home, on eris
castaway nods
nothingmuch castaway: if you guys have nat punched, i can connect to you, though 09:26
Khisanth hrm has there been any updates to forms for perl6?
nothingmuch and capture tcpdump for you
Khisanth: forms are going away
they will be warpped in a module
let me find the synopsis for you
Khisanth I mean Form.pm
nothingmuch ah 09:27
i think it's not goint to change much
it's for compatibility
castaway nm, I think we do, but I'm working on client side, at the mo.. havent done much server-side yet
Khisanth I was just looking at search.cpan.org/~dconway/Perl6-Form-0.04/Form.pm and wondering if the {} stuff would cause problems
castaway (As, afaik, emacs cant listen on ports..)
We really need interaction
Khisanth since {} is also used for interpolating code 09:28
nothingmuch castaway: p'raps this stuff is better put in a library to be bound by emacs later?
castaway nothingmuch: that would require writing it in something other than elisp :) 09:29
nothingmuch that is exactly my agenda ;-)
nothingmuch wants it for vim
castaway heh
well, one thing at a time :)
nothingmuch which happens first? .tcshrc or .login? 09:30
nothingmuch really hopes rc
decay_ yep 09:33
castaway Hmm,m I wonder if clisp makes libraries ,)
emacs can use cl 09:34
decay_ btw: fun.drno.de/pics/english/vi-keyboard.jpg 09:45
jabbot pugs - 1341 - Rat 10:17
pugs - 1340 - add method and submethod
pugs - 1342 - some small fixes 10:37
11:00 metaperl__ is now known as metaperl_
integral pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | FBSD_4.11:r1302(183/3383) win2k:r1324(187/3474) Linux:r1302(193/3383) MacOSX:r1342(189/3478) 11:31
castaway wonders what those numbers are about. 11:35
malaire castaway: jabbot reports commit to pugs (also in here: rt.openfoundry.org/Foundry/Project/.../pugs/log/ ) 12:04
s/commit/commits/
integral castaway: the numbers are subtests failing out of the total
malaire you didn't mean jabbot's numbers then ;) 12:05
castaway I meant the ones in the topic 12:07
because it all looks like gibberish to someone who hasnt been sitting here the last month or so
malaire perhaps topic should include explanation in the end - "platform:revision(failed/total)" or something like that? 12:09
castaway it might be useful.. or a url which points to a more useful list of numbers .)
integral malaire: anyone can change the topic ;-)
12:11 Aankh|Clone is now known as Aankhen``
rgs notices CGI::pugs 12:23
castaway wossat?
rgs ask stevan :) 12:25
gaal_ good Time::HiRes::time 12:37
kungfuftr stevan: ping 12:58
stevan kungfuftr: hey 13:20
castaway: CGI::Pugs is a minimal CGI module for pugs 13:21
its a proof of concepts in a way
stevan got sick of writing/porting modules which wont work for months 13:22
I prefer immediate feedback :)
castaway I see 13:23
nifty
stevan castaway: it works, but only for GET and it does not encode/decode the strings
(need some features to implement that) 13:24
rgs you already have index and substr :p
stevan rgs: thats true
I could do it really really really old school
rgs in C ? :) 13:26
stevan assembler :)
stevan reaches for that IBM 360 assembler book he got on Ebay 13:27
stevan puts on a white lab coat and gets out his cardpunch 13:28
castaway *g*
kungfuftr stevan: yullo! you know your 'debug' pragma style module on cpan... is it just a custom loggin tool or does it actually work with the perl debugger? 13:30
stevan kungfuftr: it is a very basic custom logger
realllllly basic
its a really bad name, I should fix that
kungfuftr stevan: ah... k. damn... no easy profiling for me 13:33
stevan kungfuftr: basically I didnt know about Log::Log4Perl
kungfuftr ah
=0)
Limbic_Region stevan - IMO, it would be much nicer to go the CGI::Simple emulation route then it would be CGI.pm - CGI.pm is a fugly boheameth
castaway It works though :) 13:34
stevan Limbic_Region: I agree, I was really just experimenting last night
I am also not very familiar with CGI::Simple (or CGI for that matter)
rgs for cgi, minimalistic is better
stevan I use mod_perl handlers, and TT
rgs TT uses cgi::something to parse what mod_perl hands to it, anyway 13:35
Limbic_Region stevan - doing something (anything really) is better than what I am doing (nothing) - didn't mean to sound as though I was unappreciative - I am very much btw - thanks
stevan Limbic_Region: didnt sound that way at all :) 13:36
kungfuftr bah, yaml harness no work unless pugs is installed
gaal_ l4p, now that's somehting nice to port to p6
kff: ? i use it all the time and pugs was never installed on my system
(hi all :)
13:37 gaal_ is now known as gaal
stevan hey gaal 13:37
kungfuftr gaal: how?
(ie: command line usage) 13:38
gaal util/yaml_harness.pl 13:39
:)
stevan rgs: I agree, simple CGI is good
kungfuftr gaal: i just get lots of 'could not exec /usr/bin/pugs'
gaal urr!?!?
kungfuftr and can't locate Test.pm if i put a symlink in 13:40
gaal what does your HARNESS_PERL env var say?
try setting it to $PWD/pugs
kungfuftr gaal: still moans about missing library 13:42
gaal does prove work? prove -v t/builtin/goto.t for example 13:44
export PERL6LIB=ext/Test/lib too.
kungfuftr ta 13:45
seems to be running now 13:46
gaal cool. 13:47
scw I wrote an one-line shell script file: PERL6LIB=ext/Test/lib HARNESS_PERL=./pugs prove $*
rgs PERL6LIB=blib6/lib might be better. or shorter. 13:48
gaal prolly a good idea to put that in the wiki. pugs.kwiki.org/?FrequentlyAskedQuestions 13:50
stevan Limbic_Region: I just looked over CGI::Simple, very nice 13:52
much better than trying to decode CGI.pm :)
14:07 xerox_ is now known as xerox
Limbic_Region stevan - and to be honest, a lot of those methods are only there to be compatible with CGI.pm (IOW - there is still fat that could be trimmed) 14:16
stevan Limbic_Region: can you maybe write down a core set useful routines? I nevver use these modules, so I am just guessing 14:19
stevan has to run an errand, bbiab
14:20 xerox_ is now known as xerox
Limbic_Region stevan - maybe we should ask James Freeman (Tachyon) if he wants to get involved 14:28
I only ever use a few CGI methods but I don't do a lot of web programming myself
jabbot pugs - 1343 - * Parrot IMC compiler, parser, evaluator 14:37
autrijus hi :) 14:38
<- implemented a minimal IMC compiler/parser/prettyprinter in 1 hr
castaway IMC? 14:39
mugwump showers autrijus with confetti/champagne/jism
obra uh.
autrijus castaway: parrot intermediate code
castaway ah
nifty
rgs wow
autrijus++
autrijus gadt++
th++
scw autrijus++
autrijus # amazing technology
scw autrijus: Err, stupid problem. How can we help on IMC? 14:40
autrijus scw: well, by downloading parrot
build it, look at imcc/*
read docs/* and imcc/docs/*
and then get back to me :)
scw Ouch, so you didn't read them? :p 14:41
autrijus I did. you also need to read them if you want to help :)
scw Oh, sure :)
castaway bah, programming blind is so much more fun :)
autrijus castaway: true, but not as helpful :)
scw and perhaps more bug?
mugwump so, you're not including *that* in your 1 hour, are you? the truth comes out 14:42
autrijus mugwump: eh, what?
castaway true enough
autrijus no, I did not include the time I spent on reading PDDs, nor did I include the time I spend on reading TaPL, or the time I spent learning english.
if you count it that way, it took me 23 whole years
fancy that.
castaway *g* 14:43
mugwump gestation? :)
autrijus 24 then
stevan++ # wow you're crazy, man 14:44
(CGI::Pugs) 14:45
obra CGI::Pugs?
autrijus obra: yeah, real working one
pugs/ext/CGI-Pugs/
multi sub param returns Array is export { keys(%PARAMS) } 14:46
multi sub param (Str $key) returns Str is export { %PARAMS{$key} }
multi sub, man
rgs doesn't urldecode though.
rgs loves multi subs
autrijus it has this decode_url_data
obra it's a CGI lib for pugs?
autrijus obra: yeah
obra any reason it needs Pugs in the name? non core features? 14:47
autrijus there's no reason. we are at this "name it whatever" phase.
castaway
.oO( why does that sound so long winded?)
obra nods
kungfuftr gets an automated smoker going locally 14:48
autrijus kungfuftr++
kungfuftr autrijus: sorry bout dbi::pureperl btw... =0( 14:49
autrijus np :) 14:50
mmm cyclop smiley
kungfuftr nope... big nose smiley
autrijus ahh. 14:52
autrijus decides to crash for 15 mins or so
brb &
15:01 clkao_ is now known as clkao
nwc10_ Soooo, " their Makefile.PL are now written in Perl 6" .... 15:20
there's a plan to port MakeMaker to Perl 6 and have schwern's personal hell last another 20 years +? 15:21
:-)
nwc10 I found that I'm already on freenode :-)
theorbtwo No, there's a plan to port Module::Build and add anything else it should have. 15:22
kungfuftr rightio... automated smoker setup... yay 15:23
kungfuftr.com/pugs-smoke.html - run every hour on the half hour 15:24
15:27 Corion_ is now known as Corion
Corion Heh. Chat!Zillas auto-reconnect feature is nifty - it reconnects me after my machine wakes up from hibernating. Nice :) 15:29
... I should automate the reauth with NickServ too ... 15:30
chip *sigh* I like MakeMaker.
Corion chip: It's better than the alternatives :)
chip Perl 6 makes rewrites fun! 15:31
castaway thinks someone should port it .)
Corion castaway: I might, if just to spite the Module::Build crowd.
rgs Corion: that's evil :)
theorbtwo Don't spite us, just join us.
castaway heh 15:32
gaal yeah, you get to maintain it too Corion :)
chip Zombie::Build
Corion rgs: I have been less than enchanted by Module::Build, its development process and the attitude.
theorbtwo Find the things that module::build doesn't do, but should, and make it do it.
Corion: We're forking. Different development process, different attitude.
gaal alternatively, write sometihng completely new.
castaway what it doesnt do seemms to consist of "adding patches" which is hard to patch :)
Corion theorbtwo: That implies I am interested in developing it. I'm interested in using it.
theorbtwo: I hope so :) 15:33
rgs yay, flamewar.
chip theorbtwo: Forking implies a new name, unless the old fork is dead, in which case it's a change of management
which is it?
Corion rgs: No - it's a constant rant of mine :)
chip: I consider "both" old systems dead, except that EU::MM is the better variant of dead, as it is a clean skeleton instead of a rotting carcass
castaway *g*
theorbtwo chip, we haven't picked a new name yet. 15:34
chip theorbtwo: Have you considered building on Cons?
theorbtwo: All the dependency work has already been done
Corion theorbtwo: BTW, what do I need to create your funky HTML graphs of failing/succeeding tests ? Just run the Perl script? 15:36
theorbtwo Run the yaml harness to get a tests.yml, run my perl script with stdout redirected somewhere to make the HTML graph. 15:38
Corion theorbtwo: Thanks :)
theorbtwo (Both should be run from pugs' root directory -- that is, the one containg t.)
q[acme] there can only be one 15:44
theorbtwo We're the princes of the universe? 15:47
autrijus notes that Pugs uses the 3rd way. 15:48
Module::Install, that is :D 15:49
it's the middle path
kungfuftr bah, every half hour i mean
theorbtwo You weren't here when we had the discussion earlier, autrijus: did you use M::I because it was convient, or because you like it?
...and is it a good stepping-off point for making a system that does the right thing for perl6, has a clean code base, etc? 15:50
(AFACS, it fails pretty hard on having a clean code base, since it mostly delegates to EU::MM.) 15:51
Limbic_Region autrijus - did you get the /msg I sent you yesterday? 15:52
mugwump Every half hour? That's not quick enough, we need real time streaming!
gaal kff: better to listen over the commit feed and do it right after you automatically svn up 15:54
but if you wait a few days we'll write that for you. :)
autrijus Limbic_Region: about what? :) 15:56
Limbic_Region pdcawley's question about the Santa Claus golf
Limbic_Region goes to find it again
www.perlmonks.org/index.pl?node_id=443062 15:57
pdcawley used to do the p6 weekly summaries
Limbic_Region notes the time and wanders off to a meeting
Corion autrijus++ # if you count it that way, it took me 23 whole years
autrijus theorbtwo: uh, well, M::I can delegate to M::B (which also delegates to EU::MM 15:59
and yes, I used M::I as a stopgap
because, well, I wrote it ;)
theorbtwo I didn't realize that.
Corion theorbtwo: I notice that all diag() output belongs to the test before it, which is sensible, except that my tests output the program command line before executing it. Is there any way to indicate to which test a diag() output belongs ? (I guess no :) ) 16:00
theorbtwo At some point, we need to have a fairly long discussion about how this stuff should be done, because the current p5 way sucks in many ways.
Corion, not that I'm aware of.
Corion theorbtwo: Not that important anyway :) Switching order makes the diag useless (I output in case it hangs), and implementing such a feature introduces useless bloat (as I will be the only customer) 16:01
theorbtwo Not only that, but it's a matter for the TAP protocol layer. 16:02
stevan kungfuftr++ # for the auto-smoker 16:03
orba: I plan on changing the name from CGI::Pugs, it was late last night, so I was not in my right mind (which also explains why I wrote it too) 16:04
any suggestions for a better name for CGI::Pugs are welcome
Corion stevan: CGI ? 16:06
nothingmuch stevan: ping
Corion (let's leave HTML generation up to the templating modules)
... like HTML::FromAST
gaal Corion, re diag: TAP doens't suppotr that afaik
nothingmuch could you do a bit more complicated data input into testgraph.pl?
gaal but it doesn't mess up the smoke output
stevan nothingmuch: hello
Corion (for those who really crave procedural HTML creation)
nothingmuch i would like to representd consolidated results
either a test is green, that is, passes everywhere
stevan Corion: I dont want to replicate CGI, its too bloated
gaal ah actually in your case it might,a bit 16:07
nothingmuch or red, fails everywhere, or whatever
Corion gaal: No, it doesn't mess it up - it just displays the information one-off
stevan: Yes. Clean up CGI as it should be. Or call it CGI::Request :-)
Maybe there should be no CGI.pm6
gaal only declarative stuff. maybe we need anohther method that does the same as diag but decorates the output so it's clear it's a mesage about the next test.
stevan Corion: CGI - The Next Generation
theorbtwo nm, possibly that is a sepperate tool that takes a bunch of YAMLs and creates one summary YAML from them? 16:08
nothingmuch and if there is yellow, then pass/fail per platform
and yellow in case there is some disagreement
in any case, all diags are shown in the tooltip
Corion gaal: It doesn't matter that much. Extending TAP would be nice, but it's not important
nothingmuch theorbtwo: yes, the merging is done by something else
gaal to2: yes, i'll wotk on that tomorrow
unless nm beat me to it :)
nothingmuch i doubt i will
nothingmuch is sufferring from fatigue
work is getting to me 16:09
theorbtwo Oh: I was working on a better HTMLifier of the docs, which is presently half-done.
nothingmuch and i have a movie date tonight
if i'm energized when I return, then i could do it
gaal is suffering from too many weddings
nothingmuch but probably tomorrow
kungfuftr parsefailure? 16:10
hhhmmm...
ah... worked it out 16:13
gugod_ data.gugod.org/pugs.png # SVK::Churn result of pugs. Red = lines added , Green = lines removed 16:55
autrijus 20..k? 16:56
oh wow, 28k
gugod_ hmmm that doesn't looks right 17:00
gaal what's the y axis?
gugod_ lines of code
I'll do it again with some legend 17:01
Corion Yay. Seems like I finally got "-I foo" to work the same as "-Ifoo" :) 17:21
Regression tests++ # Forgot to fix some other command line parsing in the process 17:22
autrijus Corion++
Corion autrijus: But for the next incarnation of the command line parser, I'll turn to a proper Haskell parser I hope ...
... because the Pugs/Perl command line actually warrants it :)
autrijus clkao++ # core darcs patch algebra API landed on SVK 17:23
clkao grins
autrijus Corion: cool :) I may interest you in parsec :)
Corion autrijus: Yeah - I'll have to learn about Haskell types anyway ... But first I need to get this out of the door, do svn up and then another test run to update the Win32 stats :) 17:24
metaperl_ clkao, if you had to choose just one version control system, which would you choose? 17:25
autrijus cool :)
clkao cp/mv
autrijus svk cp and svk mv, of course?
metaperl_ lol
clkao orz
Limbic_Region stevan - my email to James (the author of CGI::Simple) was returned since he no longer uses the account (Spam), but I have resent to what I believe to be a good address 17:36
gugod_ data.gugod.org/pugs.png # updated 17:46
autrijus woot 17:47
dada 28k lines added in one go? 17:48
gugod_ that doesn't really look right :/ but svk result looks correct (/svk.png) 17:50
crysflame hi, gugod 17:51
autrijus zzz & :)
gugod_ hi crysflame 17:52
Corion If I add a new .hs file, what are the steps to take except svn add ArgParse.hs ? Do I manually update the Makefile / Makefile.PL or is there some generator ? 18:15
kungfuftr ah, home at last 18:16
PerlJam Corion: update the MANIFEST probably.
Corion PerlJam: Good idea!
PerlJam Corion: (and Makefile.PL)
I'm not certain about the relationship between Makefile.PL and the MANIFEST, but I know it carps at you if there are discrepancies
kungfuftr MANIFEST is only generated for releases 18:17
PerlJam ah, then updating Makefile.PL should be sufficient.
Corion Ah. Makefile.PL picks up all src/*.hs automagically
So svn add should be all there is to do ... AFTER a make test :)
kungfuftr yup, if you find all the output a pain you can run `make manifest`, just don't commit it
Juerd Hehehe. A flash intro says "Loaded Inf%" 18:21
Corion And what am I really glad that I wrote all those silly regression tests against -e , -V:foo and -I - by fixing one, I broke the other. Two times. :) 18:23
hlen hey dudes.. still no ==> and <== operators in pugs? 18:27
jabbot pugs - 1344 - '-I foo' now works the same as -Ifoo 18:31
stevan hlen: no that does not work yet 18:32
kungfuftr yay! auto-smoke is actually working properly from cron now 18:33
stevan kungfuftr++ 18:34
if you dont mind the traffic, put the link either on the wiki or in the topic here please
kungfuftr pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | Auto-smoke: kungfuftr.com/pugs-smoke.html | win2k:r1324(187/3474) Linux:r1302(193/3383) MacOSX:r1342(189/3478) 18:36
Corion kungfuftr: Have your script somewhere ? Because for a Win32 smoke, cron is not the best solution :) 18:36
kungfuftr Corion: i have a script, not very portable though
Corion kungfuftr: Ah, it's not in Perl but in shell I gather :) 18:37
Corion crosses his fingers.
kungfuftr nope, it's in perl okay... just calls lots of shell commands... =0)
Corion I've just tried to switch my DSL provider and hopefully they'll do so in the next few weeks without a hitch ... 18:38
kungfuftr: Hmm - depending on which shell commands, I might have them (cygwin) :)
kungfuftr Corion: 2 secs
Corion: paste.husk.org/2905?ln=on&tidy=on&hl=on 18:42
18:43 Aankh|Clone is now known as Aankhen``
Corion pugscode.org <<Overview Journal>> | pugs.kwiki.org | logged: xrl.us/e98m | Auto-smoke: kungfuftr.com/pugs-smoke.html | win2k:r1344(187/3478) Linux:r1302(193/3383) MacOSX:r1342(189/3478) 18:46
Corion kungfuftr: Your script doesn't use much of shell :) I have svn and (n)make, otherwise, I couldn't build Pugs :)) 18:47
kungfuftr: But thanks !!
kungfuftr np 18:57
might commit it in a bit 18:58
Corion kungfuftr: Do so! You could guess / use Config.pm to get at the correct make program to use, but maybe that's just overkill :) 19:01
*make = sub { system $Config{make}, @_ }; 19:02
Can anybody give me usage examples of the -C<backend> and/or the --external switches ? 19:04
wilx Hmm...
The Pugs Journal page lacks encoding specification thus the Tuesday March 29, 2005 entry and its MetaClass part displays badly by default for me... 19:06
Corion wilx: Do you mean autrijus' journal at use.perl.org ? 19:07
wilx Yup. 19:08
PerlJam was starting to boggle that pugs had already attained a journal in the same vein as TPJ ;)
wilx Hmm, and it was MetaCategory, not MetaClass.
Corion Grrr. Just found another bug in command line parsing :(
(but already wrote a regression test for it :) ) 19:09
vladtz wilx: in mozila: View\Character Enconding\UTF-8 will fix that for now. 19:10
wilx Yeah, I know.
vladtz Could have guessed...
wilx :)
19:12 domm_ is now known as domm
Corion WTF ? The docs claim that the sort is stable, but I get elements in reversed (or random?) order... Is there any way to inadverently reverse a list in Haskell ? I'm always chopping of the head of lists and only use foldl() ... 19:12
... or does foldl() reverse the list ? 19:13
kungfuftr erk... even string eval hack
s/even/evil/; 19:14
Corion Ah no - just personal stupidity ;) 19:16
jabbot pugs - 1345 - Addition of script to help run smoke gra 19:21
pugs - 1347 - Test reading a snippet from STDIN 19:41
pugs - 1346 - -l now emits a fake -e comment
wilx Hmm... 19:48
I seem to be having chicken/egg problem with Pugs.
make[1]: Entering directory `/d/WilX/tmp/pugs-svn/ext/Pugs-MakeMaker' 19:49
Perl v6.0.0 required--this is only v5.6.1, stopped at /d/WilX/tmp/pugs-svn/blib/lib/File/Spec.pm line 1.
BEGIN failed--compilation aborted at /d/WilX/tmp/pugs-svn/blib/lib/File/Spec.pm line 1.
kungfuftr wilx: HARNESS_PERL=./pugs and PERL6LIB=ext/Test/lib 19:50
wilx I have no previous pugs... 19:51
jabbot pugs - 1348 - Bah... string evals hurt baby rabbits!
Khisanth no minipugs? :) 19:55
wilx None that I know about :)
Khisanth or should that be micropugs, I keep mixing the two
Corion minipugs I think - but Pugs currently is mostly minipugs, as it doesn't have any dynamic linking facility 19:56
integral for perl5, micro- indicates it can be built without a fancy build system, and mini- indicates that it's not got the bells and whistles of a full perl 19:57
wilx Well, how do I build minipugs?
Corion wilx: No need to - you just need a Perl 5 to build Pugs (in my experience), and Haskell of course :)
PerlJam wilx: What was the exact sequence of steps that you performed? 19:58
wilx: forget that, did you do "perl Makefile.PL && make && make test" ?
Corion is ?unlink() currently broken ? It doesn't seem to remove files here :) 19:59
wilx I did make clean; perl Makefile.PL; make and the make borks. 20:01
PerlJam wilx: what revision?
wilx 1348
Windows. 20:02
Corion wilx: Are you sure you want to use make and not nmake ?
PerlJam ah. it's probably a windows-specific problem then.
Corion wilx: What compiler do you have/use ?
Oooo - Cartesian products sponsors Perl6 !!! my @tests = ( '-C' ~ any('Parrot','Pugs') ~ ' ' ~ any('-e1',$*PROGRAM_NAME ) # generates all combinations :) 20:03
wilx Mingw, GNU make, GHC 6.4 built with that Nmake.
Ooops.
make
s/Nmake//
Corion wilx: Hmmm - that sounds consistent. What Perl do you use ? Mingw Perl ? 20:04
wilx Yup.
Corion wilx: Hmmm... How does your make die ? I use an MSVC7 Perl with nmake and GHC6.2 (binary from haskell.org) and all is fine
theorbtwo BTW, Corion, 6.2 is no longer supported. (It might work anyway, but we don't support it.) 20:05
Makefile.PL should probably warn you of that.
Corion theorbtwo: Ugh :) So I gotta upgrade soonish :) But it still works :))
jdv79 shouldn't the new version of CGI be under modules/ ? 20:06
Corion Heh. @tests is 3, but for @tests -> $test iterates over all 12 combinations. Is there a way to create the cartesian product in @tests without having to iterate over it? 20:07
Hmmm - my bug lies elsewhere, as usual :) 20:08
jabbot pugs - 1349 - Added test for -C compiling option 20:11
rgs stevan: your next mission is to port blosxom to pugs. 20:14
Limbic_Region rgs - I believe stevan mentioned several times that CGI isn't his gig ;-) 20:16
rgs Limbic_Region: well, I might try it, then.
Limbic_Region I have sent an email to James Freeman, the author of CGI::Simple asking if he is interested in getting involved
rgs pugs currently lacks string substitutions, right ? 20:17
kungfuftr yup 20:21
Corion Some of the examples do not compile, for example pugs -CParrot examples/junctions/any-any2.p6 - I leave them as failing tests in my test for the -C switch 20:23
jabbot pugs - 1350 - Compile examples with Parrot and Pugs, T 20:31
kungfuftr rightio, hope the new css for the test graph is a little more readable and less straining on the eye 20:35
theorbtwo Oh? 20:36
kungfuftr theorbtwo: check the auto-smoke url above 20:37
not a huge amount of difference tbh
theorbtwo I like it. 20:38
kungfuftr main difference is to use fonts designed for screen readin 20:40
anyways... pub 20:41
stevan rgs: pugs has limited string substitution actually
jabbot pugs - 1351 - Add @ARGS to script read from STDIN
theorbtwo I wish I knew what to do about that firefox bug. I haven't the foggiest idea where to start looking if it's a known bug, or who to point it out to who would know. 20:43
kungfuftr theorbtwo: which one? 20:44
rgs stevan: wow
stevan rgs: s:perl5/// is supported 20:45
rgs ah, with pcre
Corion theorbtwo: What Firefox bug?
stevan and s:perl5:g/// is supported, but limited
yes with pcre
theorbtwo Look at that smoke with firefox. 20:46
Note that some cells have the text show in them.
stevan revdiablo: blosxom,.. we need to port FileHandle, File::Find, File::stat, Time::localtime, and more of CGI first :)
theorbtwo Now view source, and compare that with the source you see from wget.
stevan s/(re)vdiablo/$1/ 20:47
kungfuftr theorbtwo: refresh (it's got to do with how quickly the css can be applied to said HTML) 20:48
theorbtwo Oooh, interesting.
Khisanth hrm 20:49
theorbtwo I've been dealing with file:/// URLs on that mostly, and thus didn't see changes from reload to reload.
Khisanth stevan: wouldn't FileHandle require objects?
stevan Khisanth: yes, so it would all have to be procedural at this point
Corion To those working on the -Cbackend compilers - should the compile tests for the examples/*.p6 go into a separate test or should they stay with my generic tests for the -C switch?
kungfuftr PUB 20:51
Limbic_Region theorbtwo - is Jess still awake? 21:00
theorbtwo Aye. 21:01
castaway more or less 21:02
Limbic_Region looking for a specification of ANSI SQL
castaway any particular one?
Limbic_Region toy project - can't seem to find anything other than SQL-92
yes, 99 preferably 21:03
castaway sec
desert-island.dynodns.net/SQL/ansi-...5-1999.zip 21:05
(but you never heard it from me..)
the BNF is in part 2 21:06
Limbic_Region why - is it not supposed to be publicly available?
castaway costs money @ISO
$18 or so if you feel like paying ;)
Limbic_Region oh - well, then as soon as I am finished please make the link go away since this channel is logged and publicly available 21:07
castaway so tell me when you're done
eep
theorbtwo The ISO and ANSI put out non-free standards.
castaway well, I gathered it from elsewhere ont' net too
Limbic_Region I found the 92 specification publicly in several places 21:08
castaway yup. that one was
drafts of this are about too, but not easy to find
Limbic_Region done - thanks! 21:09
castaway welcome 21:10
g'night
Limbic_Region is there no part 3? 21:11
castaway I only have whats there, tisnt all
Limbic_Region gotcha - 2 is probably all I need anyway - thanks again and sweet dreams 21:12
castaway thanks.
jhorwitz leave #perl6 21:15
sorry. :)
jabbot pugs - 1353 - have to commit to make other changes 21:41
pugs - 1352 - ArgParse.hs - minor changes
Corion lightstep: Thanks for fixing ArgParse.hs :) I still haven't grasped the finer details of Haskell (and forgot to clean up my global exported subroutines :) ) 21:48
Oooo - concatMap() - good to know about it :)
lightstep i'm trying to abstract unpackOption.? now 21:49
jabbot pugs - 1355 - changed CGI::Pugs to CGI 21:51
pugs - 1354 - deleteing the old CGI
Darren_Duncan New question: why was the CGI.pm port put in /ext rather than /modules? Is it required by Pugs to run?
New question: why was the CGI.pm port put in /ext rather than /modules? Is it required by Pugs to run? 21:54
repeated for the newcomer
stevan Darren_Duncan: the ext/ directory is for modules which work now 21:55
the modules directory is for one which will work later
Darren_Duncan that's the only distinction?
stevan Darren_Duncan: yes, pretty much 21:56
Darren_Duncan so when /modules items work, they will move over?
stevan Darren_Duncan: it is all pretty ad-hoc for now, hopefully by that point we will have FreePAN or something like it
neither ext/ or modules/ is ideal
Darren_Duncan meanwhile, we just need to be internally consistent
stevan Darren_Duncan: yes, that will make it easier to split off when the time is right 21:57
Darren_Duncan also, does Pugs have object support now? CGI.pm was object oriented
stevan Darren_Duncan: nope
CGI.pm is just a proof of concept right now, and more like the CGI.pm procedural interface
Darren_Duncan I understand ... it just sounded like a complete port 21:58
stevan (although I am not going to do all that HTML stuff, people should use templates)
Darren_Duncan is what it was advertised as
PerlJam stevan: as long as it doesn't inherit all of the cruft from perl5's CGI.pm I'll be happy
stevan Darren_Duncan: nothing is really complete yet :)
PerlJam: my intent right now is proof-of-concept/curiosity
PerlJam: but future plans are more akin to CGI::Simple or CGI::Lite 21:59
Darren_Duncan As I recall, the documentation in CGI.pm itself said it should be deprecated in favor of the CGI::* modules, and has said so for a long time; therefore, why port it?
PerlJam stevan++ sanity!
stevan Darren_Duncan: its not really a port, when it comes down to it
its a brand new CGI module for a brave new world ;)
PerlJam It's CGI.pm the way it should have been from the beginning only in perl6?
stevan PerlJam: yup :) 22:00
Darren_Duncan that's a nice thought
stevan Darren_Duncan: honestly it should probably go into examples/ but examples/ implies it will never be useful outside that context 22:01
Darren_Duncan I agree
stevan and while it is currently very limited by Pugs's own limitations, it can still be useful 22:02
Darren_Duncan examples/ implies not serious
stevan for those who like to live on the bleeding edge :)
Darren_Duncan I haven't checked yet, but is there a way to write Perl 6 code that talks to C libraries?
under Pugs
PerlJam Darren_Duncan: not that I know of. 22:03
stevan Darren_Duncan: Totally OT, it looks like I am going to need L::KT for my next project
Darren_Duncan stevan, do you have any further questions on it?
stevan Darren_Duncan: actually look at the SHA stuff, that is C -> Haskell, although not yet exposed to perl6 though (i think)
Darren_Duncan hopefully the P5 version corresponding to whats in /modules will be up today
stevan Darren_Duncan: not yet, but when I start development I surely will 22:04
Corion reads the Parsec documentation and likes it. Is Parsec used in Pugs somewhere?
Darren_Duncan I asked because ideally once object support is in place, I will want to talk to actual databases, and some of their interfaces are in C
such as SQLite 3
Corion Oh. Grep tells me it's used everywhere :) 22:05
stevan Corion: I think so,.. check src/Internals.hs it does most of the importing
lightstep Corion, the whole Rule system is a ripping of Parsec. it isn't imported anywhere, just copied
Corion lightstep: Ah :)
stevan Darren_Duncan: look at svn.perl.org/perl6/pugs/trunk/modules/SHA1/ 22:06
Darren_Duncan looking ...
Corion I'll port the command line parser to Parsec in the next round I guess. It's currently a mess :(
But not tonight :)
Corion goes to sleep.
PerlJam Corion: you know ... pugs really could use a perl6 rules parser ... ;-)
stevan I think that is C binding to Haskell, then the start for inline Haskell support in Pugs
Corion PerlJam: Sure - later :)
. o O (zzzzZZZZ)
Khisanth PerlJam: isn't that what PGE is for? 22:07
stevan Darren_Duncan: ingy is the guy to talk to regarding all this
Darren_Duncan okay
PerlJam Khisanth: There's nothing that says we can't have two parallel efforts to achieve the same goal is there?
PGE is written in C and can be bound into pugs to get perl6 rules in pugs. Or, using Parsec, a perl6 rules parser could be written directly in haskell and pugs can get them that way. 22:09
I thikn both are healthy efforts.
Darren_Duncan so it looks like, if I want to talk to C the way pugs is now, I write the interface in Haskell, and then the Perl 6 module just does an import
or so SHA1 seems to work
still, I'll wait for working objects before I try anything of the sort 22:10
PerlJam Darren_Duncan: why wait when you can create? ;) 22:11
Darren_Duncan priorities
the Perl 5 versions aren't even alpha yet
I mean the stuff I haven't even started porting to perl 6 yet ... (SQL::Routine IS alpha, LKT is Released) 22:13
everything else is prealpha
22:47 PerlJam is now known as SexyThang, SexyThang is now known as PerlJam
jabbot pugs - 1356 - factoring ArgParse.hs a bit 23:01
jdv79 stevan, does IOC prereq anything that has xs involved? 23:20
kungfuftr bah... heartburn sucks
stevan jdv79: nope, pure perl 23:22
jdv79 certain?
that would be perfect
stevan jdv79: actually it uses Scalar::Util, which uses C
jdv79: but IIRC that is pretty portable 23:23
kungfuftr Scalar::Util is a box of ultimate voodoo 23:29
well, sections of it are anyways
stevan kungfuftr: I am pretty sure I only use blessed(), is that portable? 23:30
kungfuftr well, how do you mean 'portable'? 23:35
erk, huge amount of parse failure -esque things happening from the last automated smoke here 23:38
lightstep my bad 23:39
(probably, i touched ArgParse)
kungfuftr checks logs
dunno, lot of haskell code changes from the previous build 23:40
jabbot pugs - 1357 - CGI work: added limited url_encode/decod 23:41
crysflame Scalar::Util is core, isn't it? 23:56
Limbic_Region yes, since 5.007003 23:57
Limbic_Region should really install Module::CoreList instead of looking at the source on CPAN everytime

Powered by Cro and the Raku® Programming Language.

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

Thank you!