similar to: a year of rails magic

I just noticed I frequently encounter patches that need to be updated because of applied changes. Maybe this can be automated? Not automating the "updating of the patch" part, but the part where a comment is posted to the ticket so the submitter is informed. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups

I was working with acts_as_taggable earlier and thought I could make it a bit friendlier. I submitted a ticket and patch #8279. Basically, there are three differences. 1. Fixed the deprecated :dependent => true 2. Added migration generator (it has really specific requirements on table layout, so might as well generate it). 3. Expanded the README to explain using the migration and to include

I''m having a wee bit of trouble building documentation for Edge Rails (rev 7187). When I attempt to run the doc:rails taask (or any doc:* task), I get the following error: -- BEGIN -- rake aborted! uninitialized constant RubyToken::AlreadyDefinedToken /Users/jherdman/Projects/crosslisting/vendor/rails/activerecord/ lib/../../activesupport/lib/active_support/dependencies.rb:263:in

Filter chaining appears to behave in differently than Rails 1.2.x. It seems that I''m getting a lot of errors along the lines of: ActionController::ActionControllerError: filter #<ActionController::Filters::ClassMethods::BeforeFilterProxy:0x322f468 @filter=#<ActionController::Filters::ClassMethods::SymbolFilter: 0x322f4b8 @filter=:login_required>> was in the wrong place!

I''ve been having a look through the code responsible for printing how much time was spent in Active Record at the end of an action. My reading of the code + experiments suggests that any database time after the render is not counted. It seems to me that in controller_runtime.rb in Active Record, append_info_to_payload should in fact read def append_info_to_payload(payload) super

I''ve submitted a small patch to make Rails behave properly with the Erubis <%== %> construct. For some reason the current behaviour of that tag in Rails 3 is to escape the contents _twice_ which is probably a bug. I offer three suggestions why this is a good idea: - The syntax is cleaner. It can avoid a lot of .html_safe and raw in your views. I especially like the conciseness of

Yesterday I answered a question regarding polymorphic URL helpers on Core ML and noticed that the module has no documentation. I''ve documented it and rewritten unit tests using Mocha. I also optimized some of the code slightly (nothing major, though). The patch also includes two fixes by Geoff Buesing, who has done awesome work in this area in the past.

Looking over the discussions and Trac tickets am I correct in assuming that the simply_helpful''s form_for not respecting the default_url_options (specifically only_path) is a side-effect of some other default behavior choices? Personally, I set only_path to true as my default as I think except were you are dealing with https it makes life much easier (especially in testing) and is just

Upgraded from rails 3.0.7.rc1 to 3.0.8 today and ran into a problem when rendering haml: TypeError in Pages#home Cannot modify SafeBuffer in place Full Trace: activesupport (3.0.8) lib/active_support/core_ext/string/output_safety.rb:122:in `gsub!'' haml (3.1.1) lib/haml/compiler.rb:382:in `block in build_attributes'' haml (3.1.1) lib/haml/compiler.rb:358:in `each'' haml

Hi all, I have a need for making ActiveResource have the ability to specify the format much like you can specify the site. So, right now (unless I am not understanding the code, which could be, since I am a Rails and Ruby newbie), all calls end up at http://something/call.xml. I want to be able to call http://something/call.rss or .vcf, etc. This is stemming from the Highrise api in 37

I get an error using Rails 3 beta3 on Ruby 1.9.2-preview1 undefined method `current_period'' for nil:NilClass (NoMethodError) This commit should fix it http://github.com/rails/rails/commit/a0cc94b32e5b4e4dce50787eed1208406f99ea82 But I forgot how to ensure I am running on edge with Rails 3. Is it a special bundler command I need to use? Thanks! -- You received this message because

I came across this when switching from rails 1.2.6 to 2.0.2, there is a change in the way routes are generated between these to version due to the routing optimization in rails 2.0.2. In version 1.2.6 it was possible to overwrite the default_url_options method in ActionController::Base to define parameters attached to each generated url. Due to the routing optimization in rails 2.0.2 this is not

303 (See Other) is the status code a resource returns to tell the client about the new resource it just created (typically after a POST). 301/302 is the status code some servers return to tell the browser where to go next. It works because browsers ignore the distinction and treat all three status codes the same way, and few people understand the difference. But when developing an application

I recently submitted a patch to work around an issue I was having & added a new optional behaviour to the to_xml method but on further reflection I wonder if the default behaviour should be changed. The gory detail, documentation and test case is all here... http://dev.rubyonrails.org/ticket/7004 Here''s a brief example... class Person < ActiveRecord::Base has_many: children

Hi, I''m getting a little confused how Rails handles loading models. I''m on development mode, using Rails 3.1.0.rc1 on ruby 1.8.7 My application models are: Curso Legislacao Usuario And I have another model that is loaded from a gem Im trying to build: Cidade My gem is actually a simple engine: require ''active_record'' require

Hello guys, I went into milestones page (https://rails.lighthouseapp.com/projects/8994-ruby-on-rails/milestones) and found out that Rails 2.3.6 was due Jan 15, 2010. Does it mean all tickets in that milestone should be cleared first? If so, then I would pay attention to clear tickets in that bucket first. Thank you, Prem Sichanugrist (sikachu) -- You received this message because you are

I''m using render_to_string in an after_filter, which renders a partial that is sent to a web service. Here''s the code: class MyController < ApplicationController after_filter :post_to_webservice def show # show logic end def post_to_webservice stuff = render_to_string :partial => ''stuff'' webservice_post(stuff) end I get a DoubleRenderError in the

Hi, For some reason, this harmless block of code in my application controller def index @wide = true end is causing this hideous error. I have read the error means that the error is raised when attribute has a name reserved by ActiveRecord. Does that mean that none of my models can have columns named "toggle"? Here is the error and framework trace:

I think I''ve found a bug with Edge. I''m trying out Ryan Bates'' multi-object form technique shown in one of his Rails-casts (railscasts.com/episodes/75). If you use a fields_for similar to that shown (here: http://pastie.caboo.se/110480), you get a Server Error 500: ------ Status: 500 Internal Server Error Conflicting types for parameter containers. Expected an

rescue_from stores exception class names in a hash table, and associates them with handlers. When an exception is raised there''s a name lookup, and if an entry is found its handler is invoked. In particular rescue_from does not emulate Ruby''s rescue semantics with regard to inheritance. Which is the rationale? Don''t you think taking is_a? and declaration order