Activity Graph

Page 1 of 1


[19:14:43] angelixd: has joined #ruby
[19:50:49] angelixd: Remote host closed the connection


[13:51:14] angelixd: Ping timeout: 258 seconds


[08:43:12] angelixd: Ping timeout: 276 seconds
[08:44:04] angelixd: has joined #ruby


[21:01:42] angelixd: Ping timeout: 244 seconds
[21:02:33] angelixd: has joined #ruby


[19:58:19] angelixd: Ping timeout: 260 seconds
[19:59:54] angelixd: has joined #ruby


[01:23:04] angelixd: Ping timeout: 240 seconds
[01:23:56] angelixd: has joined #ruby


[22:32:57] angelixd: Ping timeout: 260 seconds
[22:39:50] angelixd: has joined #ruby


[22:34:30] angelixd: has joined #ruby


[17:20:46] angelixd: *.net *.split


[16:09:37] angelixd: has joined #ruby


[03:01:13] angelixd: Remote host closed the connection


[00:18:02] angelixd: has joined #ruby
[00:18:27] angelixd: has joined #RubyOnRails


[21:04:51] angelixd: Ping timeout: 246 seconds
[21:05:38] angelixd: has joined #RubyOnRails
[21:05:38] angelixd: has joined #ruby


[15:36:51] angelixd: *.net *.split
[15:37:00] angelixd: has joined #RubyOnRails
[15:37:00] angelixd: has joined #ruby


[14:41:47] angelixd: has joined #RubyOnRails
[14:41:47] angelixd: has joined #ruby


[20:37:18] angelixd: Hrm, I need to clone an AR (Rails 3.0.x) object, update some arbitrary attributes, then save it. I can't use update_attributes because the object doesn't exist in the DB yet, and assign_attributes wasn't introduced til a later version. anyone have any suggestions?
[20:38:21] angelixd: I mean, short of just using send over the hash of attributes :-P


[21:43:05] angelixd: Does anyone have any clues as to why in my factory_girl factory, I would have *both* a FactoryGirl::Declaration::Association and a FactoryGirl::Declaration::Implicit in the declarations?
[21:43:25] angelixd: specifically, for the same attribute


[19:46:49] angelixd: Hey, which runs first: config/application.rb, or config/initializers/*.rb ?
[19:48:56] angelixd: GeekOnCoffee: I'd get my answer, but if I miss a gotcha that's not part of the regular documentation I'll end up chasing my tail for hours
[19:53:10] angelixd: parndt: thanks :-)


[17:38:22] angelixd: hey guys, I'm writing an adapter to get my persistence class working with rails. I'll need to overwrite the 'persisted?' method from ActiveModel::Model, of course. Is that method meant to be a 'dirty' flag, or is it meant to be used if the object is persisted at all?


[21:34:18] angelixd: hey, does anyone have experience with using views for non-model classes?
[21:34:54] angelixd: radar: thanks
[21:37:43] angelixd: I have an application that needs to send some moderately complicated XML and/or JS to RabbitMQ. I've been using RABL so far, and I'd like to re-use it for the generation of messages, how would I go about rendering those templates from the model layer in action hooks?
[21:41:12] angelixd: well, the architecture isn't set yet, but what I would like to do is have this in my model class: after_create { produce_message }
[21:50:02] angelixd: so quick question, because the rails magic is contagious: if I suffix all my message classes with 'Message' (e.g., GizmoCreatedMessage) and throw those all in say, app/messages in my project root, can I just have a app/view/gizmo_created_message.rabl and will that automagically render?
[21:50:41] angelixd: Spaceghost|work: any suggestions or leads, or is it time to go fishin'?
[21:53:19] angelixd: Spaceghost|work: hrmmm, okay. so my message-passing code lives in lib/. So I guess I have to set up any template associations there myself
[21:53:26] angelixd: which isn't really all that bad
[21:59:14] angelixd: spaceghost|work: you're probably right. it's just that the MVC model is almost as applicable when it comes to messaging as it does to HTTP requests
[22:03:13] angelixd: hahaha :-P well, maybe I should start open endedly: I am trying to find a good way to structure messaging in my application. I have an existing set of messages I need to send, but I'd like to model them in an OO fashion. are there any good examples of apps that do this?
[22:09:42] angelixd: spaceghost|work: I am sending RabbitMQ messages to a builder queue when some model objects get created or updated.
[22:11:50] angelixd: Spaceghost|work: yes, it is, and no, this is work.
[22:14:30] angelixd: Spaceghost|work: you're probably right. I'm biting of a lot at once. I think I'm going to step away from the computer for a bit for some perspective. Try and do little steps at once
[22:15:38] angelixd: Spaceghost|work: that's very kind of you. I appreciate the help :-)


[16:26:39] angelixd: Hey guys, I have a situation where I want to change up the way that my models store data, and I'm not sure of the best way to do it
[16:29:24] angelixd: I need to version my objects, but I need to be able to have multiple objects be live at the same time. I also need to be able to represent asssociations as being between the specific versions of the object, and not the whole object through time
[16:30:45] angelixd: rushed: I know :-(
[16:32:14] angelixd: I am thinking the simplest way to do something like this is to have a Gizmo class, and a GizmoVersion class, with the former representing the business object, and the latter being a plain ActiveRecord model
[16:32:57] angelixd: Has anyone tried to do something like this before? does anyone have any pointers or any gotchas that I should be aware of?
[16:43:29] angelixd: rushed: It's kind of a hard dependency of the data model, unfortunately. Conceptually, I have a good model for how it should work. The problem is I'm relatively green when it comes to Rails metaprogramming. I need to apply this "versioning" scheme to multiple object types, so I need code this up as at least a module, base class, or something of that ilk.
[16:44:01] angelixd: Are there any resources for doing rails model meta-programming, or should I just be cracking open lots of libraries?
[16:51:58] angelixd: rhizmoe: fair enough :-P


[17:28:15] angelixd: rompelstilchen: rabl is another option for doing data views. I really like it since you can re-use templates for relations


[18:46:25] angelixd: hey guys, I had sort of a general architecture question. I'm considering using can(can|tango) for permissions on my application, and I'm in a situation where tracking the user_id won't be enough. I'd like to add roles to the objects themselves, but it seems that can(can|tango) handles permissions on the class level, not the object level. Am I going to have to roll my own on this, or am I thinking about the solution wrong?
[18:53:20] angelixd: danneu: I'd like to have a system where each object has a set of roles whose permissions it would accept. So I could say, have two groups "company-a" and "company-b" that could tag all their objects with their
[18:53:40] angelixd: role, and that everyone who had the company role could perform those actions on them
[18:55:12] angelixd: spaceghostc2c: I totally agree managing the object roles is pretty easy. What I'm worried about is the scope of the permissions engine
[18:57:32] angelixd: spaceghostc2c: I'd like to use something off-the-shelf to save time, but cancan and cantango appear to focus on permissions on a model class level
[19:03:35] angelixd: danneu: that's about 80% of the way there
[19:05:23] angelixd: the case I'm looking at is let's say I have Employee with roles a,b,c and Clipboard with b,c,d ; and each role specifies abilities.
[19:23:54] angelixd: danneu: well, what I would *like* would be that the permissions engine would notice that the user and object share roles b & c, and then merge those permissions lists
[19:25:57] angelixd: danneu: so yeah, in this case, a,b,c,d are all role objects