Wednesday, February 2, 2005

Natural selection for frameworks in Ruby vs Java

Posted by admin

Dion Almer is pondering the difference in how the Ruby and Java communities have dealt with natural selection in web-application frameworks. About Rails, he writes:

There were various web frameworks available to a Ruby developer, and then Ruby on Rails came on the scene… How did natural selection kick in with Rails? Developers started to use it. They liked it, and the community grew and grew. As the community has grown, so has the quality of the software itself (as well as documentation).

So a case of natural selection where a popular choice emerges from the opinions and usage of many. Dion contrasts this with the “unnatural” selection in Java where the de facto champion Struts is being replaced by a committee-driven JavaServer Faces:

So, the progression has had little to do with natural selection. JSF wasn’t suddenly the best framework out there that the Java community jumped on. It was made by committee. It was funded by large corporations. And, as such, it doesn’t have the quality of a natural winner.

It’s fascinating what drives the forces of natural selection. And how it takes a long time, or doesn’t occur at all, in some communities. Or whether its even good or bad that it happens (I, surely biased, strongly believe in the good of it).