Four causes of proprietary Java frameworks
I just read On proprietary Java frameworks … and from my experience there tends to be four common threads to why these homegrown frameworks exist, all witnessed first hand.
First: A big company gets sold on another big company’s grand idea. This happened to the US Dept. of Agriculture that got sold on IBM’s pandora framework. A total disaster and completely unproductive environment. Ok, they didn’t create it, they bought it.
Second: The lead developer wrote the framework and won’t give it up. Ah yes, the not-invented-here syndrome. Probably the most common. I have yet to meet a good developer that wasn’t egotistical in some way, or perhaps evangelical is more appropriate.
Third: Sometimes related to the second. The framework didn’t solve some particular need. This doesn’t happen so much with web frameworks because the variety is so high that they cover just about every conceivable option. This really happens at the data layer. Tools like Hibernate, CocoBase, Toplink, Entity beans tend to be too rigid for a lot of applications.
Fourth: Some wacko in management or the ‘architecture group’ doesn’t want to use open source because they are afraid of viruses, security holes, etc. Never mind the hundres and thousands of developers that use this stuff on a daily basis with no reported issues.
So there you are, four reasons why companies can’t ditch that homegrown framework.
Don’t miss anything, subscribe!
Did you enjoy this post? Why not leave a comment below and continue the conversation, or subscribe to my feed and get articles like this delivered automatically to your feed reader.
Comments
Perhaps there is another valid reason you are missing: open source sucks ass in java. Look at the crap out there!!!
There are rails-like frameworks some lucky developers use, and have used for years. We have had one, in java, since 2001. Its elegant, fast, and has been a major commercial advantage over the last 5 years. Do you honestly think that hibernate is the best java can do for data? Far from it.
I enjoy giving up my own frameworks. A few years back I made an XML and reflection-based GUI library for Swing. After learning Swing a bit more, I realised that my Swing code could be written fairly well in Java itself, and a month ago finally deleted all traces of that framework from my source code (but not my source control!). I found that a better solution than extending the framework to support more and more ’special’ cases.
@carsten
I’ll buy the argument that it had been ahead of its time. I also see that once a framework has been widely used within a company it can be difficult to change.
@ben
I’m not a rails fan by any means, but I certainly don’t endorse or like hibernate either.
@ricky
It can sometimes be hard to give up your own stuff, especially when there is nothing really ‘wrong’ with it. I’ve done it a few time myself, but it was always because there was something better out there. Like you, I also never deleted the source 
Five or 4b: When we are doing it ourselves, we can give timely support and develop it in the direction the market needs.
Or 2b: The idea of the framework had been ahead of the time.
Of course no company gives enough resource to develop a framework properly and maintain it, nor do you have the resources to adapt to the new trends