the real Nate Clark

I. Create a new Partition

Your Macbook comes with a pretty big hard drive (mine is 200GB), but it’s probably configured as just one big volume … that’s how most laptops are. You can use Disk Utility to resize your main “Macintosh HD” partition to allow some space for your new encrypted one.

WARNING: Disk utility is not supposed to destroy any data in this process, and it worked fine for me. However, I still recommend you back stuff up anyway, in case of a problem or failure. Disk partitions gone wrong can be disastrous to your data.

1. Insert your Mac OS X install CD
2. Choose the Install Mac OS X app from the CD, and click the restart button
3. Your Mac will boot from the CD. After you select your language, don’t continue with the installation. Instead, look up to the menu bar and choose Disk Utility from the Utilities menu.
4. Highlight your hard drive and then click over to the Partition panel
5. Now you can resize your “Macintosh HD” to make room for a new partition. Click the little resize handle in the bottom right of the graphic, and drag your HD smaller. I shrank my Macintosh HD to leave about 80GB for my encrypted partition … should be enough for all my stuff.
6. Now that you have some space, click the plus icon to add a new partition in the freed space. You can tweak the sizes of the two partitions to your liking, and create a name for your new one (I called mine just “Data”). Be careful not to make any other changes to Macintosh HD … we don’t want to risk data loss.
7. Leave the Format setting at the default, Truecrypt will reformat the drive later anyway, and click Apply.
8. Hold your breath … and you’ve just repartitioned your hard drive!
9. Now exit Disk Utility, then go back to Utilities -> Startup Disk and change the boot disk back to “Macintosh HD” and reboot. If everything goes as planned, you should be back at your desktop, and a new hard drive called “Data” shows up right there next to “Macintosh HD”.

II. Install Truecrypt and Encrypt the Partition

Now we have to set up the encrypted volume. There are a few options here, and I recommend that you read and understand the TrueCrypt documentation first.

1. Eject the empty “Data” drive by right-clicking on it on the desktop
2. Download Truecrypt and install it. I wrote this guide using TrueCrypt 6.0a.
3. Launch TrueCrypt from Applications, and click the Create Volume button
4. Choose the second option, “Create a Volume within a Partition/Device”
5. Select your new partition (you should be able to identify it by the size and the device, mine was on /dev/rdisk0s3)
6. Now there are several options to choose. I did not need the crazy security of a hidden volume, so I went with the standard encrypted volume. I also stuck with the AES encryption default. Go through the dialogues to create your encryption keys and passphrase or keyfile.
7. Next, you need to reformat the drive. This will take a while, mine took about an hour. I may have been able to get by with a “Quick Format” but I wasn’t sure, so I did the full format. Click through all the scary warning messages and get it started …

IV. Mount your Encrypted Drive

Once the formatting is done, you’re ready to start saving your encrypted files. You’ll be brought back to the main Truecrypt screen where you can now mount your new encrypted drive.

You’ll have to do this after every time you boot your computer to re-mount your encrypted partition.

1. Click on the first “Slot” and click Select Device.
2. Choose your encrypted partition (i.e. /dev/rdisk0s3) and OK and Click Mount
3. In the Mount dialogue, click the Options button and you can change the Mount Location. To be more Mac-like, I called it ”/Volumes/Data”
4. Type in your passphrase and click OK

After a couple of seconds, your new encrypted drive should pop up on your desktop and in Finder and you’re ready to go!

class NinjaTurtle < ActiveRecord::Base
end

class Leonardo < NinjaTurtle
  has_many :katana
end

class Donatello < NinjaTurtle
  has_one :bo
end

class Michaelangelo < NinjaTurtle
  has_many :nunchaku
end

class Raphael < NinjaTurtle
  has_many :sai
end

@donatello = Donatello.new

The straightforward way would be to just write an action called new in each of the subclasses controller, but that’s a lot of repetition of basically the same thing.

The solution? Make sure your subclass controller inherits its parent class controller, then you can write a generic new action in the parent controller.

class NinjaTurtleController < ApplicationController
  def new
    @ninja_turtle = params[:controller].camelcase.constantize.new
  end
end

class LeonardoController < NinjaTurtleController ; end
class DonatelloController < NinjaTurtleController ; end
class MichaelangeloController < NinjaTurtleController ; end
class RaphealController < NinjaTurtleController ; end

Now, any request to /leonardo/new would create a new Leonardo object and assign it to @ninja_turtle. Similarly, a request to /donatello/new would create a new Donatello object. The beauty is that there is really only one new method that is flexible enough to create the right type of object depending on the controller that handled the request.

How it works

I discovered yesterday that Ruby’s class names are constants. That means that there is a constant named Leonardo that refers to the Leonardo class.

If we’re using Rails convention, which we should be, the controller name that corresponds to the Leonardo class is leonardo. We can grab the controller name from request parameters.

  params[:controller]                        #=> "leonardo" 
  params[:controller].camelcase              #=> "Leonardo" 

and then we use the Rails built-in method constantize to convert our string into a constant, and invoke the new method on the object that we’re trying to create. It essentially evalutates to Leonardo.new, which is exactly what we wanted!

References:
Programming Ruby
Infovore – Getting a class object in Ruby from a string containing that class’s name

Log In Using SSH

I use PuTTY for all of my SSH needs. Fire it up and connect to your domain using SSH. After you enter your password, you’re at your home directory (replace nateclark here with your username). Its worth mentioning that you should log in with the same username that your web application runs as (whatever you set in the DreamHost control panel when creating your domain).

[/home/nateclark]$ 

Create a Directory for Compiled Packages and Gems

To keep things neat, I created a directory called .packages under my home directory for any compiled packages. You could download and compile stuff right in your home directory, but that could quickly get cluttered. The . in front of the directory name makes it a hidden directory, so it won’t be listed in a regular ls command. You have to use ls -a to list all directories including hidden ones.

In my ~/.packages directory, I currently have installed my own versions of Ruby 1.8.6, Trac 0.10, Python 2.4.3, ImageMagick 6.3.0, Subversion 1.3.2 and a few other smaller things.

I also have a ~/.gems directory to store all of my own gems.

Set Up Your Paths

Ok, this is the important part—your path variables tell the shell where to look for executables and gems. We’ll set these up in your ~/.bashrc file, which is executed by bash for non-login shells. For regular login shells, you want to use the same path variables, and ~/.bash_profile sets this up. I’ve chosen to source ~/.bashrc at the end of ~/.bash_profile. For Linux newbies, ~ is a shortcut for your home directory.

My ~/.bash_profile looks like this:

# ~/.bash_profile: executed by bash(1) for login shells.

umask 002
PS1='[\h:$PWD]$ '
alias ll="ls -l" 
EDITOR="/usr/bin/vim" 
. .bashrc

And my .bashrc looks like this:

# ~/.bashrc: executed by bash(1) for non-login shells.

export TZ=EST5EDT # Sets my timezone to Eastern U.S. time
export LD_LIBRARY_PATH="$HOME/.packages/lib" 
export PATH="$HOME/.packages/bin:$HOME/.gems/bin:${PATH}" 
export GEM_HOME=$HOME/.gems
export GEM_PATH="$GEM_HOME:/usr/lib/ruby/gems/1.8" 

NOTE: Some people (including me) have had problems with Dreamhost’s shared gems conflicting with gems that you install locally. To force your environment to use ONLY your local gems and not the Dreamhost managed gems at all, change the last line above to:

export GEM_PATH="$GEM_HOME:/usr/lib/ruby/gems/1.8" 

Now that these paths are set, simply log out and log in again to get them to work. Or, you can just source the file at the prompt:

. ~/.bash_profile

To test if it worked, just echo your path. You should see something like this:

$ echo $PATH
  > /home/nateclark/.packages/bin:/home/nateclark/.gems/bin:/usr/local/bin:/usr/bin:/bin:/usr/bin/X11:/usr/games
$ echo $GEM_PATH
  > /home/nateclark/.gems:/usr/lib/ruby/gems/1.8

The paths that point to your directories are listed first, and then the DreamHost shared location is next. Sweet.

Configure Your Rails Environment

In your Rails applications, you’ll also have to tell Rails where to look for your gems. Add this line to the top of your config/environment.rb, file:

ENV['GEM_PATH'] = '/home/nateclark/.gems:/usr/lib/ruby/gems/1.8'

Install Your Packages

Ok, now you’re ready to install your packages into your ~/.packages directory. You can do this just like you would manually compile a package normally, with the only difference that you need to use the --prefix=$HOME/.packages option when you configure. For example, here’s how I installed Ruby 1.8.6:

First, install readline. This is required if you ever want to use script/console on your Dreamhosted rails app.

$ cd ~/.packages
$ wget ftp://ftp.cwru.edu/pub/bash/readline-5.2.tar.gz
$ tar zxvf readline-5.2.tar.gz
$ cd readline-5.2
$ ./configure --prefix=$HOME/.packages
$ make
$ make install

Now, download and compile the latest version of Ruby:

$ cd ~/packages
$ wget ftp://ftp.ruby-lang.org/pub/ruby/ruby-1.8.6.tar.gz
$ tar zxvf ruby-1.8.6.tar.gz
$ cd ruby-1.8.6
$ ./configure --prefix=$HOME/.packages --with-readline-dir=$HOME/.packages
$ make
$ make install

Occasionally, Dreamhost will kill a process that is using a lot of CPU or memory and would be bogging down the server. A few times, they have killed my make command. If this happens, just run it again until it completes successfully.

Then, make sure you’re actually using the new version:

$ which ruby
  > /home/nateclark/packages/bin/ruby
$ ruby -v 
  > ruby 1.8.6 (2007-03-13 patchlevel 0) [i686-linux]

Update: Some of you had problems with gem. That’s cause I left out the part about installing rubygems. Oops.

$ cd ~/.packages
$ wget http://rubyforge.org/frs/download.php/20989/rubygems-0.9.4.tgz
$ tar zxvf rubygems-0.9.4.tgz
$ cd rubygems-0.9.4
$ ruby setup.rb config --prefix=$HOME/.packages
$ ruby setup.rb setup
$ ruby setup.rb install

Install Your Gems

Installing gems is just as simple as always. Since your $GEM_HOME is set, all your gems will go into the directory that you specified. For example, install your own version of Rails:

$ gem install rails --include-dependencies

And make sure that you’re using the right one:

$ which rails
  > /home/nateclark/.gems/bin/rails

Thats it! Now you can manage your own versions of pretty much any package, library or gem. Of course with that comes the responsibility of keeping everything patched and up to date.

Let me know if I’ve missed anything. Good luck.