“The race is long and in the end, its only with yourself”

Progress on Domain Pigeon (DP from now on) has going well. Several weeks ago I estimated the site was about 90% complete, which turned out to be very optimistic. Deploying a Rails app, especially for someone with no experience, takes time to do properly. I also made the mistake of not writing unit tests until recently. The “Aha!” moment came the first time I saw DP running on Dreamhost and realized there wasn’t an easy way to make sure everything worked properly.

Lesson Learned: When the roads are covered with fog, plan for extra travel time.

Lesson Learned: Prepare thoroughly. When it comes time act, you’ll be glad you did.

The plan was to launch several days ago, but some last minute preparation and a ski trip this weekend has postponed it.  I’d like to be around to respond to the initial feedback and to address any problems that arise.

Also, updated the blog layout, as the old one was, well, bleh.

And finally, a good FastCompany article titled “What Should I Do with My Life“, courtesy of HackerNews. Here’s a quote:

But the article itself flipped that connotation inside-out. It argued that with the economy in a tailspin, it was unsound economic theory to have millions of drone workers shuffling to work every day doing jobs at quarter-speed they didn’t care about, so they weren’t very productive at, and certainly didn’t add value at. The economy would never get kick-started if our workforce was uninspired and didn’t innovate. So the article — really a manifesto — suggested that the way to get business going again was for its basic building blocks — the workers — to do something they were really good at, or were inspired by, or cared about, where they would work extra hard, and innovate their way out of this black hole. Now it was not a permission slip to quit your job, nor a doctor’s note to take a year off (I’ve never taken more than two weeks off in 23 years), but it did suggest the economy might be better off, long-term, if the square pegs found their square holes and the round pegs found their round holes, rather than everyone just wondering where the next big thing would be and gravitate to it like moths.

Right now my main priority is getting an alpha version of Domain Pigeon launched. It turns out deploying it isn’t as easy as uploading a few files like it was with PHP.

The last three days can neatly be summed up by this post’s title and for a bit of elaboration I…

– Learned how to use Github to host a repository

– Finally had a need to learn and use SSH

– Re-learned how to use Git, as I haven’t been using it nearly as much as I should have

– Learned how to use Capistrano to deploy a Rails application and how to deal with a mess load of error messages that pop up when you do things incorrectly…

Here are some good resources for anyone else looking for information on these topics:

Github’s Git Cheatsheet

Capistrano’s Getting Started Article

Deploying Rails on Dreamhost with Passenger (there are several Dreamhost/Passenger/Capistrano articles out there but this was the best and most recent)

Github’s Deploying with Capistrano

Suso.org’s SSH Tutorial

Unix File Permissions

Philly.rb Google Group (thanks guys for the help)

Hack for using deploy:web:disable with Passenger

Peepcode’s Capistrano Video

About sixteen hours of the command line, trouble shooting, and dozens of tutorials later, a web page miraculously appeared! (It seems so easy in hindsight)

And now comes the fun part… prepping for actual people.

Domain Pigeon is probably around 90% ready to go.

There are a few finishing touches I need to make before the first iteration is launched sometime in early January. The plan is to have a few friends test it, then ask the Philly.rb folks for feedback, then maybe HackerNews, depending on how things go.

One thing lesson learned from ALL IN Expert is that you can’t wait for a perfect product in order to launch. A better approach is to get something decent, launch, get feedback, then iterate over and over until you have a great product. If I had done that sooner with ALL IN Expert, I’d have saved myself several weeks of work because I would have realized that there wasn’t a strong market for the product I was building. Therefore, with Domain Pigeon, I’m drawing a line and not adding any new features until I get feedback on the ones I have.

There was a thread recently on HackerNews about this. Brian Lash sums it up pretty well:

Throw the launch mindset to the wind.

I think our preoccupation with a huge launch owes more to its sexiness than to its practical importance for a startup. You may catch lightning in a bottle with a launch that’s met with a ton of fanfare. But you can also build an enduring success by focusing your time and energy on sale #1, then #2, and so on.

Your odds are long one way or the next (and granted things change when you need to observe network effects to succeed) but it seems it’s almost always the better shot to build fast, iterate faster, and lean into whatever success you find along the way.

If anyone is interested in helping test out the initial version in about three weeks, please shoot me an email and we’ll go from there.

Happy holidays —