I decided to move to GitHub Pages and Octopress, thinking it would be easier to blog by pushing my posts to a git repository.

Although the new site is up and running, I have plans to ‘spruce it up’. One: I still need to learn what extras are available (i.e. Octopress plugins, Jekyll plugins). I will not be replicating the entire site here to over there. Two: I have a particular theme in mind and will be working to convert it into Octopress’s theme structure. Three: I’ll be updating my Feedburner configuration so, hopefully, my RSS followers will be able to automagically pickup the new site’s posts.

Alright – Time to go!

Cheers

Since RVM come onto the scene, I’ve heard great things about it. But everything I heard was in the context of Snow Leopard (Mac OSX 10.6). For reasons better left unsaid, I’m still on Leopard and was doubtful I would be able to take advantage of RVM.

So I did what most people do: I turned to Google. At first I didn’t find any helpful posts. However, one post stood out and I diligently followed its directions. I made sure I had the right version of XCode. I updated SQLite3. I then installed RVM. Here’s where the fun begins.

RVM plays hard to get

I followed the installation directions, updating my .bash_profile as instructed. I then restarted my Terminal session and patiently waited…WTF! My window disappeared! I finally was able to edit my .bash_profile and commented out the line that loaded ~/.rvm/scripts/rvm. Now what?

I then decided to manually source the rvm script in the hopes of seeing an error or something. No dice. Great! No output and my window keeps disappearing. I search the available help and troubleshooting guides: Nothing. I go back to Google: Nothing. I find nothing that is similar to my problem. *sad trombone*

So I head over to the IRC channel: #rvm on freenode.net. And a couple of folks start helping me out. First I tell them about some warnings I received during installation. I then reinstall (and rvm reload) upon their suggestion. No change in behavior. I tell them, “There’s no output and then after a second or two my Terminal session disappears.” They check my shell: echo $SHELL => /bin/bash. Nope that’s not it. They ask for rvm info. Well I have to run ~/.rvm/bin/rvm because I can’t load RVM as a function. The output looks ok. They ask about my .bash_profile; I also include my custom prompts, thinking I might have a problem there. Nope, all of that looks good.

In between the IRC exchanges I learn about set +e and set -x. Using those two, I’m finally able to see the output from the script and, as I had suspected, the script is stuck in an endless loop! Great! How do I debug that? BTW set +e means: Don’t exit on error and set -x means: debugging output

The folks helping me are out of ideas; one even nukes and reinstalls RVM; it works for him. It must be something unique to my machine. I feel I’m in store for a lengthly and, quite possibly, fruitless search. Thankfully I have some output captured, early output before the session window disappears. I load up MacVim with the scripts and start looking at each line, comparing it to the output.

Foiled by ‘:’

As I was going through the output I noticed a ‘cd ..’ command. I couldn’t find it in the scripts. Thinking, “somewhere they have to be ‘cd’ing because it’s in the output.” And then it dawned on me: wait a minute, I have ‘:’ aliased. Just at that moment I was pinged from IRC. I excitedly replied that I might hove found the problem. On line 40 of ~/.rvm/scripts/initialize is: : \. Substituting my alias for ‘:’ the output matches; it’s trying to ‘cd’ into several directories. I remove the alias and reload my Terminal session.

Drum roll please: rvm is a function. SUCCESS! Now I can take advantage of RVM.

Funny thing is: no one thought to ask about my aliases. Assuming I would know that ‘:’ is null, a do nothing command. I do now. And hopefully this will help someone else with similar problems in loading RVM.

Cheers!

1. Clone your repository into a new working directory.
2. Create new private repository in GitHub; keep the resulting page up because you’re going to need the new repository url.
3. Update project’s origin remote.
4. Push changes up to GitHub.
5. View the new repository in GitHub.

  $ git clone [repository url] [new name]
  $ cd [new name]
  $ git remote add xorigin [new repository url]
  $ git remote rm origin
  $ git remote rename xorigin origin
  $ git push -u origin master

Cheers!

wget http://www.kernel.org/pub/software/scm/git/git-1.7.3.2.tar.gz
tar xzvf git-1.7.3.2.tar.gz
cd git-1.7.3.2
./configure
sudo make && sudo make install

Verify by running git –version.

Manpages

wget http://www.kernel.org/pub/software/scm/git/git-manpages-1.7.3.2.tar.gz
cd /usr/local/share/man
sudo tar xzvf [download_dir/]git-manpages-1.7.3.2.tar.gz

Verify by running man git.

Cheers!

I will not steal from you the opportunity to read Jeremy’s post and Jim’s post. Go read them, for they provide the context behind this entry; I’ll wait for you.

I may ramble; I may not. But the journey should be fun. Enjoy!

What are they talking about?

I believe both approaches have merit and may even be complimentary. However I feel both approaches have a different set of motivations and goals which it seeks to achieve.

In my short, illustrious career of some 17 years I’ve been involved in several different types of employments: full-time employee, contractor, but all in an organization I did create or have influence over. Over this time I developed likes and dislikes; we all do. And when I seriously started thinking of working for myself, my old prejudices came to light, shaping the kind of organization I wanted to grow.

I believe both posts come from a position of the ideal organization they would like to have, based on the experiences both gentlemen have accumulated over their careers, and, dare I say, their lives. Like was mentioned in a comment: “top-down” vs. “bottom-up”. But I think there’s more to it. I will not pretend to speak for Jeremy or Jim.

What does my Co-Working future look like?

I recognized early on that I would not have employees, that Juicy Parts Software would be a one-man show. As with everything, there are pluses and minuses to this approach. This approach allows me to focus on the relationship aspect of the project. But it does limit the kind of services I can provide, on the surface. Because I value the relationship above the technology I value a large and diverse network of professionals. There is where my approach, I believe, more closely aligns with Jeremy’s post. Working with people with whom I’ve built a relationship is like an ongoing interview process wherein I don’t have to have employees.

However I also recognize the allure of Jim’s post. As with the aspect of employees, I knew I didn’t really want to be involved with the business part of my business (I quickly outgrew that position, though), wanting to focus on what I’m good at: building resilient software solutions. Being able to offload all the icky bits of business is attractive to me. But, I don’t believe I would partake in Jim’s approach for personal reasons: I need to know AND trust the people I task with running my business. The one thing I’ve learned over the years: everyone has different motivations. Even if everyone really likes a CPA or a Lawyer, there’s no guarantee I’m going to like them. In that respects I find Jim’s approach no different from the typical corporation. And though it would seem ownership is the differentiator, I posit: that won’t scale. Maybe it doesn’t have to, but if it doesn’t you lose economies of scale when it comes to negotiating rates and what not.

I readily admit that I’m not a business guru; that’s why I try to listen to everyone around me that have successful businesses. But for me co-working is a chance to work with people whom I enjoy being around without any pretense of trying to build a larger organization above the community of which we’re a part.

Cheers!

I’ve heard nothing but good things about RubyNation and, this year, I’ll be there. They’re offering a great conference: Speakers and Agenda.

It’s nearly sold-out so you should stop reading and register: http://www.rubynation.org/registration

Cheers!

I just knew there had to be a way to get to the keys in gmaps_api_key.yml (this file is used by the YM4R plugin) because I didn’t want to copy/paste the values into environemnt.rb and I didn’t want multiple definitions in each evnrionment’s file, and this is Ruby!

So I’m going to try this method: one line in environment.rb

GeoKit::Geocoders::google = Ym4r::GmPlugin::ApiKey.get

Cheers!

[GEM_DIR]/rubyscript2exe-0.5.3/bin/rubyscript2exe

Before:

  gemdir	= File.expand_path("..", File.dirname(__FILE__))
  realstuff	= File.expand_path("realstuff.rb", gemdir)
  isapplication	= File.basename(File.dirname(__FILE__)) == "bin"

  $0.replace(realstuff)	if isapplication

  load(realstuff)

After:

  gemdir	= File.expand_path("..", File.dirname(__FILE__))
  realstuff	= File.expand_path("realstuff.rb", gemdir)
  isapplication	= File.basename(File.dirname(__FILE__)) == "bin"

  $_0 = realstuff
  alias $__0 $0
  alias $0 $_0
  $0.replace(realstuff)	if isapplication

  load(realstuff)

[GEM_DIR]/rubyscript2exe-0.5.3/realstuff.rb

For me this starts on line 620:
Before:

    if __FILE__ == $0
      $0.replace(File.expand_path("./init.rb"))

      TAR2RUBYSCRIPT	= true	unless defined?(TAR2RUBYSCRIPT)

After:

    if __FILE__ == $0
      $_0 = File.expand_path("./init.rb")
      alias $__0 $0
      alias $0 $_0

      TAR2RUBYSCRIPT	= true	unless defined?(TAR2RUBYSCRIPT)

Again: This works for me. YMMV!

Cheers!

I think the main reason my blogging has almost become nil is that I’m not sure what I want to use this blog for. I’m on Twitter. I run the CVREG. I just don’t know what to do this blog. I had ideas. But I don’t have the enthusiasm and energy right now – I’ve been focusing those on the projects I’m either running or helping with; maybe I’ll talk about my projects (at a high-level since I’m under NDA for some).

Maybe.

Until then…Cheers

Cheers!