==April 8, 2009==
16:24 <@jhannah> t0m: can I whine about being perfectly happy with Catalyst-Plugin-FormValidator? You marked it deprecated 6 days ago. 16:25 <@castaway> jhannah: you could rewrite as not a plugin ;) 16:25 <@jhannah> So if I moved it to Catalyst::View::Role::FillInForm and Catalyst::Controller::Role::FormValidation 16:25 <@jhannah> everyone would be happy? 16:25 <@castaway> role? 16:26 <@castaway> I dunno, whats current $trend ? ;) 16:26 <@rafl> eww.. '::Role::' in role names :-) 16:26 <@jhannah> that was mst's suggesions. I don't know anything about roles 16:27 <@jhannah> who says "Data::FormValidator itself is not recommended for use" ? 16:27 <@jhannah> that package is not marked deprecated 16:28 <@castaway> good Q that 16:28 <@castaway> dice tossing? 16:28 <@jhannah> t0m r9643 16:29 <@jhannah> seen t0m? 16:29 < purl> t0m was last seen on #catalyst 4 hours, 43 minutes and 48 seconds ago, saying: dhoss: kd, I think. 17:14 <@t0m> jhannah: gnip, looking 17:15 <@t0m> jhannah: and yes, DEPRECATION 17:16 <@jhannah> t0m: who says "Data::FormValidator itself is not recommended for use" ? 17:17 <@t0m> me 17:17 <@t0m> formvalidator validates content 17:17 <@t0m> not structure 17:18 <@t0m> FAIL 17:18 <@jhannah> ok, so mst's Catalyst::Controller::Role::FormValidation suggestion is DOA too? 17:19 <@jhannah> trying to plan my production transition to SomethingElse 17:19 < confound> are FormValidation and FormValidator the same thing? 17:19 <@jhannah> ya. sorry 17:19 <@jhannah> err... I don't know if that was my typo or mst's suggestion 17:20 <@jhannah> since my happy production MyApp's rug has been yanked :) 17:20 <@t0m> jhannah: Well, my big issue with DFV is that if you have it @stuff, it could say yes, when @stuff wasn't what you expected.. 17:20 <@t0m> especially in the case where you had ?param=value1¶m=value2 17:20 <@t0m> and you get { param => [qw/ value1 value2 /] } 17:21 <@t0m> if your solution fixes (or doesn't fall for) that evil. 17:21 <@t0m> Then all is well. 17:21 <@t0m> Please make marcus ship code with a better deprecation notice :) 17:21 <@t0m> I've just been attacking things uninformed people accidentally fell over, which have been untouched for 3 years. 17:22 <@t0m> And then complained about on the list. 17:22 <@jhannah> -ponder- so if I can write a test which shows this hypothetical I don't yet understand, and then patch it, then it might be undeprecated? 17:22 <@jhannah> and/or I'd be happy to be the official finger of blame until I $transition 17:22 <@t0m> jhannah: No, as its a plugin :) But the underlying module itslef, I'm sure can be used less evily :) 17:23 <@t0m> the issue is where you end up with some_function(%params);, or more specifically, some_function( param1 => $c->req->param('param1') ); 17:24 <@jhannah> so plugins are uncool and are to be killed? 17:24 <@t0m> as if ?param1=one¶m1=two 17:24 <@t0m> then bad things happen. 17:24 <@t0m> Unless you _need_ to fiddle with dispatch, its not a plugin 17:24 <@t0m> adding random shit to $c is evil 17:24 <@t0m> I want multiple form processors in one app. 17:25 <@t0m> This conversation validates how there is no one right solution ;) 17:25 <@rafl> (although probably not as evil as the hacks we pull to keep it working in 5.8 ;-) 17:25 <@t0m> And if the only solution is $c->form 17:25 <@jhannah> ok. I'm happy to do The Right Thing. this is production, so they'll pay me to transition to $something_else so... just trying to figure out what that is 17:25 <@t0m> Then you're totally screwed for changing over, ever :) 17:25 <@t0m> $self->form - totally fine 17:26 <@t0m> that can be on a controller by controller basis :) 17:26 <@t0m> which is a small enough chunk to convert at once.. 17:26 <@t0m> rafl: Are you calling my code a hack? 17:26 <@t0m> bah 17:27 <@t0m> you're still sore as B::Hooks::EndOfScope didn't do what I wanted 17:27 <@t0m>17:27 <@rafl> yeah, that too. but that's fine now you use one of my other B::Hooks to get the dev server restarting right ;-) 17:28 <@jhannah> $c->form evil, $self->form ok? but *::Plugin::* namespace is dead regardless? ... What Catalyst namespace would be acceptable if I make Data::FormValidator use less evil? 17:29 <@rafl> a controller role or at least baseclass 17:29 <@jhannah> or do I need to punt all my stuff to Catalyst::Controller::FormBuilder? 17:29 <@t0m> jhannah: Catalyst::Controller::FormValidator is fine :) It's a controller base class (or perhaps a role?) 17:29 <@jhannah> oh, ok. want me to write that on github.com? or start writing it in Cat SVN somewhere? 17:30 <@t0m> the point being anything you can compose onto a subsection of your app, as opposed to applying to your whole app 17:30 <@jhannah> roger that 17:30 < confound> jhannah: *::Plugin::* isn't dead, just very, very overused 17:30 <@t0m> jhannah: either/or, cat svn _slightly_ prefered from my end right now 17:30 <@t0m> Only because I haven't taught the smoke tests about git 17:31 <@jhannah> is it cool to Module::Starter an "svn mkdir" in the Cat repo then? 17:31 <@rafl> jhannah: either is fine. in my experience you get a lot more contributions from others by using github tho :-) 17:32 <@t0m> jhannah: CX::Starter is worthwhile IMO 17:35 <@jhannah> t0m: So it's cool if I "svn mkdir Catalyst-Controller-FormValidator" in http://dev.catalyst.perl.org/repos/Catalyst/trunk/ and grab another C::C::* as a template to get started? 17:35 <@t0m> jhannah: yes. You may not be able to as you don't have those permissions. 17:36 <@t0m> jhannah: if so yell 17:36 <@t0m> it shall be sorted 17:38 <@jhannah> woot! r9668 17:38 <@jhannah> I'm unstoppable! mwoohahahhahahaha 17:39 <@t0m> Good man 17:40 <@t0m> well volunteered 17:40 * jhannah wonders if anyone will notice an svn cp of Catalyst::Plugin::FormValidator 17:40 * jhannah ducks, hides back in his little SVN hole 17:42 <@jhannah> t0m: you volunteered me by pushing my $production off the DEPRECATED cliff ;) well played, sir 17:44 <@t0m> awesome. 17:44 <@hobbs> t0m is THE DEPRECATOR 17:44 <@t0m> I'll do that more 17:44 <@t0m> purl: t0m is also THE DEPRECATOR 17:44 < purl> okay, t0m.