Heroku wants you to use a CNAME to point a subdomain of your site to their proxy address, and I got that up and running right away. www.imatightwad.com was set up and running with little trouble, but without www things came apart. Now, I could pay $5 a month to Heroku for a wildcard match on the domain, and that’d be fine. But what if I want to set up blog.imatightwad.com in the future? What if I want to point help.imatightwad.com somewhere? I like using Dreamhost to control my domains.

While the amount of control that they allow over DNS and other details of the setup is admirable compared to other places I’ve tried, there are a few DNS records that they won’t let you edit. After trying several approaches and looking for any way to do a redirect of the primary domain that would be supported by Dreamhost, I finally caved in and contacted support. Maybe the solution was right under my nose. In less than 30 minutes I had a response, telling me to reenable hosting for the domain, using their radio button to redirect no subdomain to www, and they would remove the uneditable A record that were causing me trouble. And now everything is happy at TightWad.

So kudos to Dreamhost for being so responsive, and to Heroku for offering such a great hosting service. I’ve never been able to deploy a Rails app so easily. Oh, and finally, thank you to Brian Hogan and Kevin Gisi for their work on Lazy Developer, which made moving our data between servers painless.

I’ve been contributing to Mentors, and it’s been a really great experience. We’re maintaining 100% test coverage in RSpec, and we have a ton of Cucumber stories. I’m learning to test a lot of things that I hadn’t considered how to test before. Thorough tests are not something I’ll ever skimp on again.

I’d also like to take a moment to say that it’s been great working with Brian again. It’s been years since we did any sort of project together, and I’m enjoying collaborating with him.

This was the first plugin I ever wrote (the only plugin I’ve written so far, actually), and it had been over a year since I first released it. Going back over it this past week, there were chunks of code that I didn’t understand anymore, plus functions trying to do way too much without being clear about what was happening. I also added unit tests galore (more lines of test code than lib code!) and a series of rake tasks for getting the JS libraries and cleaning them back up for testing.

Cleaning The Code

I’ve just started reading Clean Code, and I’ve been talking with Brian Hogan about it quite a bit as he’s been reading it, so I’ve tried to take the nuggets I’ve learned from those places to heart as I did this refactoring. The new version doesn’t have the mess of class variables at the beginning, though there are still a few.

Most importantly, this new version is understandable. I have method names that actually describe what’s happening, and methods that are much simpler to understand at a glance. No more ridiculous chains of functions, causing untold pain as they slowly unravel.

New Features

One of the nicest things about doing this rewrite, besides the challenge and the learning experience, was that it’s so easy to add in new features now! As I worked through the code, I got the idea to handle library dependencies, so if you include Scriptaculous through the javascript_include_tag, it’ll make sure that Prototype is ready and loaded first. I’m doing the same with jQuery and jQuery UI. This is a very minor feature, but one that I couldn’t imagine trying to add in the old version of the code.

Try IncludeGoogleJs

So why don’t you grab IncludeGoogleJs and give it a shot? Put it through the paces, break it, and let me know what I missed. Or maybe it’ll help lighten the load on your server, and your users, a bit. I’d love to hear about that too.

%w[@cdwarren @bphogan].each do |keyword|
  FakeWeb.register_uri(:get, "http://search.twitter.com:80/search.json?q=#{keyword}",
  :string => JSON.parse(File.new("#{RAILS_ROOT}/test/files/search/#{keyword}.json").read))
end

I tried loading the file in several different ways, using :response and :file instead of :string, but they weren’t happy with what I was doing either.

Oh, and for the record, this is how I’m loading the XML files. Many thanks to Brian Hogan for getting me going on this part.

%w[cdwarren bphogan].each do |screen_name|
  FakeWeb.register_uri(:get, "http://twitter.com:80/users/show/#{screen_name}.xml",
  :string => REXML::Document.new(File.new("#{RAILS_ROOT}/test/files/profile/#{screen_name}.xml")))
end

Oh, and don’t forget to require 'json' if you’re going to use JSON.parse!

In users_controller.rb just add one line to the successful_creation method.

self.current_user=@user unless @user.not_using_openid?

That’s it! Your users will be on their way even faster now. Take advantage of the glory of OpenID. Many thanks to Brian Hogan for pointing me in the right direction when I started looking in to this.

Include_Google_JS is a plugin for Ruby on Rails that will use Google’s hosted copies of the major Javascript libraries.

All you have to do is add :include_google_js => true to javascript_include_tag in your layout and you’re in business.

Upcoming features will include:

• The ability use more than the default JS libraries (Prototype and Scriptaculous) when using :all or when files are declared.
• Version checking to include the right version of the library as what is contained in /public/javascripts.

I’m also hoping to submit this is as a patch to core soon.

After you enter the code in iTunes that links the AppleTV to the computer, the first option you can select in the AppleTV menu is your iTunes library. While this isn’t a huge issue, what happened for my dad was that he clicked in to that library, thinking that’d finish setting up the connection. What actually happened was that he was now browsing his library remotely.

Because he was browsing his computer’s iTunes library, the AppleTV never began synching. It showed up in iTunes, but most of the options, like Auto and Custom sync, were not there. He could play all the music, see all the videos, but everything was extremely limited. There was no obvious indication of what was happening.

We ended up unlinking the AppleTV, restarting it, and recreating the connection. Since I’d set up several others before, I was familiar with the process and backed out after the link was made instead of moving in to the remote library. Everything started synching and we were good to go.

I’ve also put up a link to my Tumblr blog, which I have neglected for quite some time. I always wanted to use it to share the random videos, photos, and wacky things that I came across, but it slipped from my mind with great consistency. I’d like to develop a ‘web presence’, as the kids say, and this is part of my efforts to do that.

Beyond that, there’s not a lot to report from me. Going through things one day at a time, getting ready for spring and summer and making the most of my great new place.

I definitely need to purchase a bike soon - the weather is so great outside, and I want to make the most of it. Walking around yesterday confirmed for me that I definitely want to make the most of my chances to be outside.

Oh, and further taking advantage of spring, I walked home from work on Friday. It wasn’t too bad, only took an hour to get out the door to work to walking in to my apartment. It was a nice chance to reflect and get a further appreciation for where I’m living.

I headed to the Mall of America after work, with my laptop and Ziphone loaded on it. I got in to the Apple Store, and 5 minutes later I had a phone working with T-Mobile and I was online.

A few days in to using it I’m pretty pleased. Edge is definitely slow, but I have no point of reference for what 3G would be like. I have to sell my iPod Touch now, but I don’t expect that to be too difficult.