This is a hammering of a few thoughts. I’ve been working with Ruby for about three years now and the Rails framework for about a full one now. I have to admit, the speed at which web applications can be built is really intriguing compared to statically typed languages like C++ but that’s probably a unfair comparison. Outside of Ruby, I’ve worked on hobby Web applications using Zend and Mojolicious with no real interest of taking it to market as a skill. Rails was the turning point.

Why Rails?

First and foremost, if you do a bit of back end Web development, the one thing that was always a concern was handling database migrations whenever entity relationship diagrams changed (though I feel they shouldn’t have). Defining a SQL schema by hand and having PHP execute it before it was ran in production (yes, bad practice, smell it) was less than routine but often enough to be a problem.

One day, while I was in California for an event at Stanford, another engineer suggested that we rewrite the entire platform using Rails (3.2.6, if that matters). I was really against initially out of a lack of knowledge of using the framework but I caved in. After a bit of shoddy pair programming, my whole opinion about the thing changed.

Today?

After using and working with a few Rails applications, I’m not so sure I’d want to continue using it. For starters, I’m a bit of a fan of being able to tweak and hammer things to a particular setup (take my laptop, for example). Rails is configurable, but in a very weird way. Again, good software has ‘sane defaults’, but I wouldn’t consider most of the defaults for Rails to be¹.

People are looking to the whole “let’s make an API!” stage with event based web stacks² powered by Node/JavaScript and decoupling the …. hogwash. For now, any serious disconnection from the current trends of application would only occur when HTML5 is 100% out the door. And we know when that’s happening.