similar to: Re: Soft failing deleting-type migrations

TDD, SVN and migrations seem to be a natural combination for model development: create a test, add the column with a migration, write the code, and check it all in when you''re done. The trouble is that this ends up creating one migration per field. I guess it''s not horrible in and of itself, but it does have a code smell to it. Has anyone come up with a different way of doing

Using 1.0.8, I just realized that I''ve been scrolling through pages of backtraces when rspec is supposed to clean them up for me. This is a new development machine, so I''m not sure if older versions worked or not. I''m running Ruby 1.8.6, Rails edge, Mac OS X 10.4.10. I''ve verified that the QuietBacktraceTweaker is the one that''s running.

I'm using exim4 to deliver into dovecot 1.1, and using plus-addressing to deliver mailing lists directly to appropriate folders in my "lists" mailbox. I created a file called /folder-map.lists that looks like ## folder-map.lists dovecot: Dovecot lpedge: Linux PowerEdge ... and so on. So mail to lists-lpedge at jay.fm goes to the "Linux PowerEdge" folder. The one

I''ve got a spec that loads a fixture in the "before" block. This works fine running scripts/spec, but when I run rake spec instead, I get: ActiveRecord::StatementInvalid in ''User in fixture :quentin with an IM service but no IM name should be invalid'' Mysql::Error: Cannot delete or update a parent row: a foreign key constraint fails

I started using the new Model.should_have(1).records expectation in rspec_on_rails, and quickly realized that my fixtures were remaining loaded, even in contexts that didn''t use them. Bug or feature? Jay Levitt

I know that a given search index doesn't get built the first time until you do an actual search on it from an IMAP client. I don't much care about that; I'm the only user, and I don't create new folders too often. At worst, I could create a script to go search every folder once. But what I'm not clear about: Does it then get updated on every delivery by LDA? Or does the

This is such a vague request it''s not even worthy of a ticket, but: When rspec does support integration testing, it''d be really nice to be able to do things like this plugin does for standard Rails tests: http://www.jasongarber.com/articles/2006/10/24/easier-testing-of-forms-form_test_helper In a nutshell, it finally closes the loop between "the empty form sent by a

The 0.7.0 vendor/selenium/README.txt references a "start_browser_once.patch" file. This doesn''t seem to be included in either the rspec or selenium-rc distributions, and Google finds nothing... where might I find that patch? Jay Levitt

Setting up a few hosts in a small domain, I thought I'd save keystrokes by defining ssh_config aliases for their non-canonicalized names. For example: Host myhost HostName myhost.example.com Host *.example.com Port 23 ... I would have expected that "ssh myhost" would then start a session on port 23 (instead of 22). And I've seen blog/list posts that suggest the same

I installed OpenSSL 0.9.5 in the default location (/usr/local/ssl) on my Solaris 2.7 box, and then tried to install OpenSSH 1.2.2p1. Configure (with no arguments) complains that it can't find working OpenSSL libraries. The last few lines of config.log show a whole bunch of attempts to compile the test code, all failing of course. My /usr/local/ssl DOES have an include/openssl directory

Now that 0.7.0 is forcing me to write properly compartmented specs, I thought I''d give the "dead easy Selenium support" a try. Is it supposed to all the things that rspec normally does for me, or is it just for using rspec-like grammar to surround Selenium tests? Specifically, should it roll back the database between specifications? I just did a quick test, and the result

When developing applications, Webrick will automatically reload controllers, models, etc. if the source changes. For obvious reasons, it doesn''t do this for plugins, which makes it a pain to develop said plugins. I was hoping Mongrel might have some facility to say "watch this directory too and reload any changes". Obviously, re-init''ing all plugins would be

Every once in a while, I pipe up and whine that "belongs_to" should really be ''refers_to" [http://dev.rubyonrails.org/ticket/2130]. It''s time again. I''ve got a bunch of stuff going onto eBay, so, like any good engineer, I don''t just post it manually: I design a Rails-based inventory database that creates a semantically-correct XHTML/CSS auction

[OK, since I stupidly posted this over the July 4th week[end] and got no responses, let''s try it again...] As a perpetual Rails dabbler, I make an awful lot of typical beginner mistakes. And the logs are always either too verbose or too sparse to help me. And it started me thinking. I came from one of the largest production OLTP systems in the world, and the one thing that helped

I have a plugin that uses RSpec, and it''s being used by a Rails app that does not otherwise use RSpec (and does not want to). I can install the RSpec gem, but of course there is no RSpec-on-Rails gem. And I wouldn''t really want to include the whole rspec-on-rails plugin inside my plugin, even if such a thing would work. Yet it would be nice if the app''s continuous

My spec contains: response.should_have_tag "<legend>", :content => "Add new Contact" The output is: Spec::Expectations::ExpectationNotMetError in ''The ResourceController should do something when I try to edit contacts'' .. <legend>Add new Contact</legend> .. should include [[{:content=>"Add new Contact",

I''ve been successfully using rails_spec_server for some time. Somehow, in the process of deleting and recreating my rspec directory a few times for SVN (or, less believably, in the process of building SVN 1.4.0 and its dependencies), something got hosed. Now, script/rails_spec takes 5-10 seconds to get past DRb.start_server. This is on a fast Windows machine under cygwin. Anyone

As usual, list price ($400)... Jay Levitt

The docs in the SVN version of file_column state that url_for_file_column will return paths relative to the application root, and that if I pass it to image_tag, I''ll get an absolute URL. This doesn''t appear to be the case, at least with Rails 1.1. image_tag returns the same app-root-based path that it''s given. Doc bug? Rails change? file_column bug? Jay Levitt

I''m developing a library called Hark. The structure is: lib/ hark/ session.rb mixins.rb hark.rb hark.rb: ==== module Hark include Hark::Mixins ... end application.rb: ==== class ApplicationController < ActionController::Base include Hark end my_controller.rb: ==== def login hs = session[:hark_session] = Hark::Session.new ... end So application.rb