Quite honestly, I never understood why this title was ever considered. After all, would you buy a book on how to build a watch, or would you buy a watch. Maybe that is a bad analogy.

If you have a problem to solve, what are you going to do? Buy a framework or build one? Chances are the smart move is to buy one. Besides, so much is in the public domain on frameworks that you are can get plenty of advice for free. i.e., there is no need to pay $40 or $50 for the privledge.

Before you tackle the job of building a framework, you should decide if you should in fact build one. Before doing that, you should see what is on the market. Measure those features up to what you need. Chances are, the gap will be small or non-existent...

I often ask the question of when did it become a requirement that programmers let go of their common sense. Seriously, it becomes a simple economic decision.

If you decide to build a framework, chances are, you will spend 200-300 hours at least. Even at a $50/hour rate, you are still talking about a cost of $10-$15 Dollars. Or, you can buy a framework - like FoxExpress which is the only one I recommend for a few hundred dollars. Chances are, that framework will do everything you need. If it does not, it is easily customizable.

A treatise on how to build frameworks is an interesting academic excercise. From a business standpoint, it made no sense whatsoever.

And remember this, when you build a framework, it is YOUR spin and take on things that counts. Usually, if it is the correct course of action to build a framework, then you have encountered a very unique business problem or you have very unique business requirements. When you think about it, nobody could ever tell you how to do that anyway.

A good book on analysis and design - or several - would be a far better investment.

< JVP >