This post originated from an RSS feed registered with Java Buzz
by dion.
Original Post: Oh no. Dr. Java thinks AOP is B.S. :)
Feed Title: techno.blog(Dion)
Feed URL: http://feeds.feedburner.com/dion
Feed Description: blogging about life the universe and everything tech
Dr. Java has gone on a rant about AOP (watch out Hani!).
His main contention is that AOP doesn't save the developer time or increase productivity.
How are you measuring that?
Does it make sense to start a small web application project and use AOP all over? Probably not (at least yet).
Can it increase productivity for large projects? I would argue that it can under certain scenarios.
For example, when you modularise the cross-cutting concerns there is LESS for a developer to worry about in their business logic. So, you are much more productive.
Take persistence with Hibernate for example. In aTrack and with Spring, you setup a persistence policy, and then the framework will automatically setup sessions, flush correctly, commit, rollback, you name it. Much more productive, and it is guaranteed to be correct. A new developer can't come onto a project and miss step 4, and suddenly screw things up.
AOP isn't for everyone. It is still new. But the core ideas make a lot of sense!
Previous Topic
