A recent blog entry about the failure of a Rails based project caused a storm of controversy. Predictably, the side of the debate with stakes in competing technologies saw this as sign of the demise of Rails - particularly as the post talks at length about switching back to PHP. However, some took a closer look at what was actually said in the post. Austin Ziegler points out a few non-technological related problems that broke the project's back:
Another problem Austin brings up is that the project was a rewrite of existing software.
The hype cycle is also brought up in this debate. The hype cycle is an idea for describing the process with which products or technologies make a splash in the market. The speedy rise of Rails indicates that hype was one factor it achieved it's current popularity and mindshare so quickly. This is an approach it shares with many technologies, such as Java, XML or venerable AJAX.
One stage of the hype cycle comes after expectations in the technology peak - which is followed by the Trough of Disillusionment. This is the point when real use of the technologies has uncovered some problems or dispelled some myths. It also comes after the period Peak of Inflated Expectations, in which some adopters might believe the technology to be a silver bullet, fit to solve all problems. Raised expectations might explain the fact that a PHP-only shop considered rewriting part of their software with Rails.
While some suggest this to be bad for Ruby, history shows that other technologies went through these same stages. Java, for instance, experienced this stage long ago when big public projects such as Corel Office for Java project failed in the late 90's or the cancellation of Netscape's Javagator, a rewrite of Netscape Navigator in Java.
Some interpret this recent debate as the backlash for Ruby, although that seems a bit late. Earlier this year, a bigger, more public Ruby based project seemed close to failure. Twitter, a messaging system, built on Ruby, experienced serious performance troubles. A remark that the performance of the Ruby interpreter might be at fault for this, caused a big stir and raised concerns about Ruby.
Now, 6 months later, things look different however. Case in point: Twitter is still up and running, has shed its performance issues, and is still implemented in Ruby. The solutions to fix the problems are now documented to everyone around the net, for instance in slide sets of talks "Scaling Twitter" or "A Small Talk On Getting Big. On the whole, the problems turned out to be architectural.
He ignored his existing experts for the new technology. Neither he nor his employees knew Ruby aside, perhaps, from playing around with it. This wasn’t a technology that was deemed to be appropriate from experience; this was a technology deemed appropriate by management (sorry Derek, you might still be getting your hands dirty with code, but you’re still management).This points to the fact that the company's software was fully PHP based, with all developers doing PHP development. A software company completely invested in PHP rewriting a big piece of software in Ruby on Rails does raise questions. Particularly, as the original post clearly states that there was only one full time Rails developer plus the writer of the blog entry, who states he's much more comfortable with the PHP environment, while the rest of the developers and the software of the company stayed with PHP.
Another problem Austin brings up is that the project was a rewrite of existing software.
Derek approached the project as a whole-environment ground-up rewrite with a One Big Day deployment, without considering ways to phase it in over time. It’s almost always possible to find interface points where you can replace one broken piece at a time. Ultimately, this is what the Rails folks wouldshould tell you anyway: replace one area at a time, each with a different codebase. Interface them as REST-ful services. Don’t make them depend on a single database schema.This has been brought up by many. David Heinemeier Hansson points to an article series "The Big Rewrite" by Chad Fowler detailing problems with project rewrites.
The hype cycle is also brought up in this debate. The hype cycle is an idea for describing the process with which products or technologies make a splash in the market. The speedy rise of Rails indicates that hype was one factor it achieved it's current popularity and mindshare so quickly. This is an approach it shares with many technologies, such as Java, XML or venerable AJAX.
One stage of the hype cycle comes after expectations in the technology peak - which is followed by the Trough of Disillusionment. This is the point when real use of the technologies has uncovered some problems or dispelled some myths. It also comes after the period Peak of Inflated Expectations, in which some adopters might believe the technology to be a silver bullet, fit to solve all problems. Raised expectations might explain the fact that a PHP-only shop considered rewriting part of their software with Rails.
While some suggest this to be bad for Ruby, history shows that other technologies went through these same stages. Java, for instance, experienced this stage long ago when big public projects such as Corel Office for Java project failed in the late 90's or the cancellation of Netscape's Javagator, a rewrite of Netscape Navigator in Java.
Some interpret this recent debate as the backlash for Ruby, although that seems a bit late. Earlier this year, a bigger, more public Ruby based project seemed close to failure. Twitter, a messaging system, built on Ruby, experienced serious performance troubles. A remark that the performance of the Ruby interpreter might be at fault for this, caused a big stir and raised concerns about Ruby.
Now, 6 months later, things look different however. Case in point: Twitter is still up and running, has shed its performance issues, and is still implemented in Ruby. The solutions to fix the problems are now documented to everyone around the net, for instance in slide sets of talks "Scaling Twitter" or "A Small Talk On Getting Big. On the whole, the problems turned out to be architectural.