Camelia, the Perl 6 bug

IRC log for #mojo, 2011-10-02

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

All times shown according to UTC.

Time Nick Message
00:13 metaperl is anyone benefitting from a nosql database when using mojolicious? I might choose one, but not sure how to integrate it with fluxflex
00:14 Akron metaperl: Sorry, don't know. But how's fluxflex doing currently?
00:14 metaperl Akron: financially? I dont know. their main competitor is dotcloud I think
00:15 Akron metaperl: I was more thinking of performance, development, features etc.
00:15 marty metaperl:  I use mongodb via the perl driver right now and like it. There is a native driver for Mojo in the works I think. But no estimated deliver on it.
00:16 marty The current perl driver requires anymoose which is a little heavy for me.
00:16 metaperl Akron: i see -- dotcloud has mongodb support .. tha's good to hear marty -- deployment to dotcloud is a little more work , but they seem more polished. plus you-know-who works there :)
00:16 metaperl i'm big on moose, so no probs here :)
00:17 marty Ahh ok.  Then might be a really good fit for you.
00:17 Akron MongoDB for Mojo is planned for 2.0 release, i heard ... ;)
00:17 marty \o/
00:18 Akron purl: true?
00:18 purl i guess true is http://search.cpan.org/perldoc?true or http://www.youtube.com/watch?v=ikkg4NobV_w
00:18 Akron Damn ...
00:27 Akron marty: http://irclog.perlgeek.de/​mojo/2011-08-21#i_4307373
00:29 tempire karma for tempire
00:29 purl tempire has karma of 39
00:33 marty Akron: Thanks.  enticing stuff.  Should be fun when it happens.  :)
00:35 Akron Definitively.
00:40 marty Mojolicous still blows be away daily.  It's amazing how feature rich it is with little (or no) dependancies.
00:42 marty Exciting times in Perl land.  :)
00:49 mattastrophe joined #mojo
00:50 marty wow.  It would be easy to become a glitch zombie.   fun game.
01:08 andrefs joined #mojo
01:30 metaperl sri: you're aware of the test failures on Windows in the CPAN mojolicious? http://pastebin.com/fF14TXdW
01:31 metaperl marty: what do you mean glitch zombie? what game?
01:48 MojoGuest11 joined #mojo
01:48 MojoGuest718 joined #mojo
02:10 mattastrophe joined #mojo
02:14 metaperl hmm, my clean urls are gone when using mojolicioius on fluxflex - http://karuna.fluxflex.com/dispatch.fcgi/hello
02:14 metaperl tempire: what issue do you have with fluxflex- http://irclog.perlgeek.de/​mojo/2011-08-17#i_4291309 ... its pretty nice for me so far
02:15 metaperl oh 404 error
02:15 purl well, 404 error is defined by its status, which is correctly set
02:26 metaperl it seems like the relation between local development and remote deployment is smoother with dotcloud, because you have to package your ::Lite app in a .pm to get things to work with fastcgi on fluxflex
02:27 metaperl ick - dotclouds CLI app doesnt play well with windows 64-bit
02:29 metaperl time to try out activestate's  - http://www.activestate.com/cloud#lc_top
02:36 metaperl actually, it looks like if I just put my start.pl ::Lite app in public_html it will run as a plain CGI
02:57 tempire have you tried redhat's cloud?
03:36 metaperl Mojolicious::Sessions uses ->load() and ->store()  ... http://search.cpan.org/~sri/Mojolici​ous-1.99/lib/Mojolicious/Sessions.pm ... but the ::Lite example simply places things in a hash - http://search.cpan.org/~sri/Mojolicious​-1.99/lib/Mojolicious/Lite.pm#Sessions
03:36 metaperl i dont think i want to try redhats cloud
03:43 metaperl I will be fine with fluxflex once I figure out why my cgi script is barfing - http://getsatisfaction.com/fluxfl​ex/topics/could_not_open_log_file
04:13 crab metaperl: what is your point? (wrt. load and store)
04:14 crab i ask because i suspect it's based on a misunderstanding of what load and store are for.
05:18 Eugene joined #mojo
05:34 kaare joined #mojo
06:01 mattastrophe joined #mojo
07:08 lammel2 joined #mojo
07:08 tabbi joined #mojo
07:14 abra joined #mojo
07:18 d4rkie joined #mojo
07:22 metaperl crab: I guess I dont see anywhere in Mojolicious::Sessions that usage that I see in the ::Lite docs
07:22 metaperl yet ::Lite is using M::Sesions underneath, no?
07:31 Vandal joined #mojo
08:13 crab metaperl: (a) users should not care about Mojolicious::Sessions at all, (b) the interface to the session mechanism is just the $c->session hash, whether in ::Lite or ::Licious
08:14 crab you should never be calling load and store.
08:14 metaperl "users" --- what do you mean "users" ... users of what?
08:15 crab mojolicious or mojolicious::lite
08:20 crab all a user needs to know is that things placed in $c->session will be persistent across requests (well, and that since this is implemented using a cookie, it's best not to put too much data in)
08:21 macroz joined #mojo
09:16 smpb joined #mojo
09:45 perlrocks Twitter: "Did in the mojolicious 2.0." (ja) --t_muto http://twitter.com/t_muto/​status/120434255154257920
10:18 MojoGuest974 joined #mojo
10:20 MojoGuest518 joined #mojo
11:08 * sri yawns
11:22 crab good morning, sri.
11:22 sri good morning mr crab
11:27 Akron joined #mojo
12:29 * sri wonders if marcus is still addicted to that game
13:14 Hinrik sri: have you considered using https://metacpan.org/module/Mo in Mojo?
13:15 sri Hinrik: why would i?
13:16 Hinrik I haven't looked at Mojo extensively but it looks like it rolls its own accessor generators, etc
13:16 sri for a very long time
13:18 sri back when i wrote Mojo::Base it was still uncool to use anything smaller than Moose ;)
13:18 Hinrik yeah, I was just thinking that it would be nice to have something that's pretty much Moose-compatible
13:19 sri so you have no real arguments?
13:19 Hinrik nope, just a thought
13:19 sri yea, i don't change design decisions unless there's a gain
13:20 sri especially if they're that popular for years
13:21 sri not to mention break every single app out there ;)
13:22 sri to my knowledge there are no mini-moose implementations yet that do chained accessors
13:24 sri Mojo::Base is really about embracing minimalism and getting the most syntactic sugar out of as little code as possible
14:40 andrefs joined #mojo
14:57 * marty waves
14:58 marty metaperl: By "glitch zombie" I mean glitch.com  (you've been warned, cute factor is addictive)
14:58 metaperl :)
14:59 metaperl sri - any comments on the windows test failures with 1.99?
15:10 metaperl I tried it (glitch) . moves a bit slow for me... but it might've fun if I kept going... thanks for linking me to it. I play Go more tha anything
15:11 metaperl Hinrik: you know Mojo is entirely self-contained?
15:12 metaperl I was going toask if Mojo should be builton POE - out of curiosity- I still dont even understand POE totally
15:15 metaperl somewhat off-topic - does anyone have a favorite (jQuery)? solution for building the nav menu for a website?
15:16 metaperl in a since "Mo" is in "Mojo" already; my $word = 'Mo' . 'jo';
15:29 Akron metaperl: navmenu - >CSS?
15:30 metaperl Akron: I dont undewrstand your shorthand, but I'll be using this - http://www.reddit.com/r/jquery/comments/khkai/j​query_easytabs_v23_released_ajax_tabs_and_more/
15:30 metaperl instead of the flowplayer jquery tools
15:31 D4RK-PH0ENiX joined #mojo
15:32 Akron metaperl: I was just wondering, what kind of navigation you want to create that needs javascript.
15:33 metaperl well it's just more automatic with javascript rewriting the dom for you
15:34 sri metaperl: i'm not aware of any open reports
15:34 Akron metaperl: Maybe, but navigation is so crucial, it has to work even with disabled js imho.
15:35 metaperl [21:30:20] <metaperl> sri: you're aware of the test failures on Windows in the CPAN mojolicious? http://pastebin.com/fF14TXdW
15:35 metaperl Akron: yes it depends on requirements. But if certain tabs need to load via AJAX, then you must have js enabled... what nav-tree builder do you use?
15:35 sri metaperl: that link is dead
15:36 metaperl sri - ok, this one works - com
15:37 metaperl this one - http://pastebin.com/3XN7a47s
15:37 Akron metaperl: As I don't do website dev very often, I stick to pure CSS navigation.
15:37 sri metaperl: always mention versions
15:38 metaperl 1.99 - the defauult one installed from CPAN
15:39 sri Perl version
15:42 sri we've recently established that ActivePerl 5.12 sucks big time and should be avoided at all costs
15:43 sri ActivePerl 5.14 seems ok
15:43 sri i generally ignore all reports that don't have at least that much information btw
15:45 sri my latest windows strategy is just to disable tests for specific versions we know are bugged :)
15:46 marcus just use strawberry perl
15:47 sri straberry has lost its edge
15:47 marcus rilly?
15:47 marcus just don't use windows
15:47 marcus :)
15:47 sri !
15:48 sri i get more reports for strawberry than activeperl these days
15:48 sri especially activeperl 5.14 works really well
15:48 Akron I made a little (or not so little) XML plugin for Mojolicious but I don't know if it's fine when I call it Mojolicious::Plugin::XML. http://goo.gl/QQBAl (Plugin) http://goo.gl/gJpSF (Document Base class). Would love to have some feedback!
15:50 sri ::XML does sound a bit unspecific
15:51 sri it looks more like a XMLBuilder or XMLRenderer or so
15:52 Akron That was my fear, however it's a pretty general extension for Mojo::DOM. XMLBuilder would look bad as a namespace for extension a
15:52 d4rkie joined #mojo
15:52 Akron extensions and base classes.
15:53 sri XMLBuilder imo :)
15:53 Akron As it inherits all methods from Mojo::DOM it can be used for generation as well as consumption, although there is no additional benefit for consumption.
15:54 sri ::XMLMonster
15:54 Akron Mojolicious::Plugin::XMLBuilder::Atom? Naaa ...
15:54 tabbi joined #mojo
15:55 sri my vote goes for Mojolicious::Plugin::XMLMonster
15:56 * sri goes to vote for more beer
15:56 estrai hi, how can I get mojo home so I can use it to tell local::lib where my local-lib is?
15:56 Akron I thought about SerialXML in first place, as it is especially useful for this kind of xml docs. Hm ... MoreXML is also fine.
15:56 Akron Okay - your vote is counted. :)
15:56 estrai something like use FindBin qw($Bin); use local::lib "$Bin/../local-lib"
15:57 Akron sri: What do you think about the plugin in general?
15:58 sri Akron: no opinion yet
15:58 Akron estrai: app->home?
15:58 sri in fact, i don't really understand what it does yet ;)
15:58 Akron sri: Diplomat!
15:58 purl diplomat is probably at http://www.lti.cs.cmu.edu/Research/Diplomat/ -- it's speech-to-speech translation on portable comupters
15:59 estrai Akron: that's not available on compile time I guess
15:59 Akron sri: Okay, that means the doc is bad ...
16:00 Akron estrai: $ENV{MOJO_HOME}?
16:00 sri judging by the base class it is something for XMLophiles
16:01 estrai Akron: I've tried that too, it's empty
16:02 sri Akron: are you aiming for something like http://builder.rubyforge.org ?
16:02 Akron sri: Yeah - it's something easily extensible without caring about namespaces etc. Useful if you have to create feeds with lots of namespaces, reusing general purpose XML modules.
16:03 sri estrai: use __FILE__ then to generate the path yourself, like the script in a normal mojolicious app does
16:03 Akron sri: Oh ...nice. Didn't know that.
16:06 Akron sri: The ruby builder is more general purpose. All things to powerful for my plugin should be done using Mojo::DOM and added as raw into the document tree. It's really for either text or element contet - not for mixed content (which is nice for meta and feed files).
16:07 estrai sri: cheers
16:07 Akron sri: But my namespace handling seems to be nicer. :)
16:11 vel joined #mojo
16:14 sri Akron: not sure the verbose api approach could work, i would try something with AUTOLOAD or maybe even serializing data structures
16:15 sri maybe a mix of all :)
16:15 sri don't know what your goal is though
16:17 sri personally i would not be uninterested in a cool AUTOLOAD api for building xml
16:20 sri but knowing you i bet the subclassing of Mojo::DOM wasn't unintentional, you have plans for filtering existing xml don't you? :)
16:23 Akron sri: What dou you mean with Autoload? In fact it uses AUTOLOAD for the extensions. And why is the API verbose?
16:23 Akron sri: Yes - Subclassing Mojo::DOM was absolutely intentional. ;)
16:24 Akron I hate the class variable approach for the Extensions, but it was the only solution I was able to get to work.
16:25 sri Akron: $xml->add(data => 'whatever') vs $xml->data('whatever') or similar
16:26 sri use classdata instead of variables then
16:27 sri use XMLMonster::Document; namespace 'whatever'; prefix 'fun';
16:27 mohan joined #mojo
16:29 sri i should mention though that i'm more a fan of template engines than xml generators though
16:31 Akron sri: I tried that - problem was, that I sometimes need the namespace in the API, e.g. in https://github.com/Akron/Sojolicious/blob/master/​lib/Mojolicious/Plugin/XML/ActivityStreams.pm#L11. When using Atom as the base class and ActivityStreams as an extension, a namespace function returned the function value of the baseclass. The variables were scoped.
16:32 Akron I like templates, but think they are ugly when dealing with lots of namespaces like OStatus does. And they are not easy to extend.
16:32 sri i see, your real goal is abstraction
16:33 * sri starts to get it
16:33 Akron Yeah, kind of. I want to have separated modules describing the api of their intention. And I want to use them as base classes and as extensions without any changes to the code.
16:35 Akron plugin('XML' => { new_feed => [qw/Atom ActivityStreams GeoRSS/]}); should work as well as => [qw/Base Atom/] etc.
16:35 sri i suppose with a good set of base classes that could result in very little actual code
16:35 Akron The actual code was the simplest solution I was able to get.
16:37 Akron And I was extremely happy that it's working! Although I don't like every part of the API - especially the variable thing. But it's semantically okay and easy to understand. In fact - it looks like there is less magic than there is. ;)
16:40 Akron sri: Regarding $xml->add(data => 'whatever'); - The "add" thing is, I think, pretty intuitive and nobody would think, you have an accessor as well. Of course, in your extension module, you can make, e.g. in an RSS module, something like "sub data {}" that's an accessor as well.
16:41 Akron Sorry - I meant somehing like "sub author".
16:43 Akron You expected to have something like Mojo::DOMs Autoloading with element creation? $dom->entry->[0]->title->text as well as $dom->enry->[0]->title('Hello!')?
16:44 sri Akron: ye
16:50 sri metaperl: you do realize that i won't be able to do anything unless you mention the Perl version right?
16:51 metaperl sri:  Strawberry Perl 5.10.1
16:52 sri oh, that's ancient :o
16:52 sri does it work in newer versions?
16:52 metaperl i dont know
16:52 sri ok, nothing i can do until i know
16:55 GitHub74 joined #mojo
16:55 GitHub74 [mojo] kraih pushed 1 new commit to master: http://git.io/gAuWTw
16:55 GitHub74 [mojo/master] fixed FreeBSD portability issues in tests - Sebastian Riedel
16:55 GitHub74 left #mojo
16:55 sri sucks that cpantesters is dead
16:56 sri but only a matter of time until enough people upgrade so i get better reports
17:13 noganex_ joined #mojo
17:47 sri interesting
17:48 sri tests don't pass with perl 5.12.3
17:48 sri even on os x
17:53 kaare joined #mojo
17:58 GitHub73 joined #mojo
17:58 GitHub73 [mojo] kraih pushed 1 new commit to master: http://git.io/Yc92Ig
17:58 GitHub73 [mojo/master] fixed more portability issues in tests - Sebastian Riedel
17:58 GitHub73 left #mojo
18:01 grim_fandango joined #mojo
18:05 perlrocks Twitter: "Starting to realize how much i'm depending on cpantesters, so many bad failing test reports for #mojolicious during the last few days. #perl" --kraih http://twitter.com/kraih/status/120559962102566912
18:20 sri hmm
18:20 sri looks like tests fail rather randomly with perl 5.12.3
18:28 sri hmm
18:31 sri i'm going to revert all test changes
18:31 sri these half assed reports need to be ignored
18:37 GitHub178 joined #mojo
18:38 GitHub178 [mojo] kraih pushed 1 new commit to master: http://git.io/--dGiQ
18:38 GitHub178 [mojo/master] enabled many tests with portability problems again - Sebastian Riedel
18:38 GitHub178 left #mojo
18:38 sri this needs to be done again, properly
18:39 sri and i'm going to keep it breaking until i get proper reports :)
19:01 sri http://groups.google.com/group/mojolicio​us/browse_thread/thread/f716bed0fb606000
19:02 sri lets gather information
19:03 GitHub99 joined #mojo
19:03 GitHub99 [mojo] kraih pushed 1 new commit to master: http://git.io/JNfhdw
19:03 GitHub99 [mojo/master] updated changes - Sebastian Riedel
19:03 GitHub99 left #mojo
19:12 sri whatever, i'm not gonna waste any more time on this
19:47 GitHub0 joined #mojo
19:47 GitHub0 [mojo] kraih pushed 1 new commit to master: http://git.io/fKVW9A
19:47 GitHub0 [mojo/master] removed a few too dangerous tests - Sebastian Riedel
19:47 GitHub0 left #mojo
20:14 perlite joined #mojo
20:20 MojoGuest44 joined #mojo
20:20 MojoGuest44 From: http://news.ycombinator.com/item?id=1277067 (1 hits)
20:22 metaperl i'm back with basically the same question. does anyone have a favorite library for setting up the navigation panel for a webapp? I suppose the code/logic for this belongs in the layout file?
20:25 metaperl a bit too hasty with my asking. I think I will just get the core tabs working - http://jqueryui.com/demos/tabs/
20:45 metaperl ah. now i've hit on my problem. Normally the layout wraps the content. But ajax tabs are planning to do the same thing
20:55 metaperl aha - basically manually specify the layout per template, not in the ::Lite app. And only have a layout in index.html.ep all the others are plain raw html that the ajax fills in
21:28 andrefs joined #mojo
21:29 lammel2 joined #mojo
21:30 metaperl hmm I've typed up the issue with getting a generic nav system to work with the layout system of mojolicious - https://github.com/metaperl/karuna/issues/1 ... it's not a simple chore
21:45 metaperl I'm starting to see the light -- akron was right -- css is really all you need :)
21:53 alnewkirk joined #mojo
21:55 D4RK-PH0ENiX joined #mojo
22:25 sri damn, i found the 5.12.3 bug
22:25 sri had a hunch it was a weaken thing, just like the malloc bug recently
22:35 sri funny, i only have to remove two lines of code to fix it
22:36 GitHub108 joined #mojo
22:36 GitHub108 [mojo] kraih pushed 1 new commit to master: http://git.io/y2k1VQ
22:36 GitHub108 [mojo/master] fixed Perl 5.12 compatibility - Sebastian Riedel
22:36 GitHub108 left #mojo
23:34 stephan48 sri: any infos how long Mojo::DOM would need for parsing a 4,4mb document?
23:36 sri stephan48: are you serious?
23:36 * sri bonks stephan48 on the head
23:37 stephan48 somewhat around 5 mins here after 5:34 mins the whole thing is finisht(with login req, generation req, logout req, data output)
23:39 sri try not to ask me questions that cannot be answered -.-
23:39 * sri prepares the tar and feathers
23:45 sri how long does it take to compile 5mb of C source code with gcc? :)
23:46 stephan48 good question 5mb of actual code or 5mb of comments?
23:47 sri that's the point! ;p
23:47 stephan48 first i cant test latter... would be testable!
23:47 sri no you can't, because there are too many variables unknown
23:50 stephan48 i could generate 5mb useless comments and compile it... the first would be a problem cause 5mb of code is a bit hairy to generate
23:53 sri i'm trying to tell you that the question you asked me was far too unspecific to be answerable, in my slightly drunk sense of humor ;p
23:54 sri i really should sleep… but i want to find this stupid last bug first :S
23:54 d4rkie joined #mojo

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