This weekend I finished migrating this blog over to WordPress, thus ending a six-month experiment with Typo, the WordPress-like blog tool built on top of the web application framework Ruby on Rails. I originally went to Typo for two reasons: To evaluate RoR performance on my ISP (The much-maligned Dreamhost) and to keep up with the numerous blog platforms available floating about in the ether. In the end, I gained a better understanding of some determining success factors for open source software projects.

Rails was extracted from 37signals‘ BaseCamp in July of 2004. The framework’s two defining principles — “Convention over Configuration” and DRY — resonated with developers looking for a way to rapidly develop RIAs. Two years later, Rails hit its tipping point: Social applications using Rails as a framework moved toward the mainstream; Bitter Java author Bruce Tate left Java.net and joined the Ruby community; Apple decided to ship Ruby with Leopard.

Unfortunately, most affordable hosting providers don’t yet support Rails (it requires a larger dedication to support and infrastructure than say, PHP), and those that do provide only a minimal level of performance. While the level of service is improving with the support of virtual private machines, it still isn’t cost effective for hosting small applications that will gain any kind of traffic. Because you can’t easily host a Rails application, the network of available developers is much smaller and more disparate than for an app built on a more widely supported platform. The result is a cascade effect for open source failure: Less developers write more code with less documentation — leading to more bugs, more scope creep, and longer release schedules.

While this may not always be the case, it certainly appeared to be for Typo.

Piers and Frédéric are only two folks with a limited amount of support, and their project is suffering because of it. When there are more affordable options available for Rails applications, Typo will shine. I therefore submit some suggestions for improving their likelihood of success:

Thoroughly document the application for your developers. This is the biggest barrier to Typo’s success. Get some documentation out there, from installation to upgrading to theming. Documenting the application will help determine the road map for the future.

Add a section to the Typo website about installing on hosting providers. Bloggers (like me) don’t want to waste time installing patches on their web server — we want it up and running quickly and painlessly.

Let people know when users should or shouldn’t use Typo. If you’re running a Rails environment already, then you should use Typo. If you’ve got a $10 monthly hosting fee, then you should consider other options.

I wish the Typo team good luck. Rails support is growing fast (check out the increase in job postings for Rails developers), so Typo will only have more opportunity to succeed. In the meantime, if you need me I’ll be blogging.