>Hi, Denis.
>
>>"We fought with it for several months and finally wrote it out of the project. I'm sure the reason is because we had a database that wasn't designed the way MM wanted and we had a miserable time trying to adapt MM."
>>
>>Could you give details on that? I mean what's so special about MM databases?
>>
>>I understand that it's better to build a framework. I'd prefer to buy one (short-term) so I can make my application faster.
>
>Are you sure? If you're business is developing frameworks and you know the plattform pretty well, it might be. Otherwise, you will end up with somethig you know well but has very little users testing it and proving the design (maybe just you), and with a poor roadmap to say the least.
>
>Yes, if you have the time, the skills and a business opportunity to build your own framework, it can be a good idea. For many independent developers, I would rather think this is not the case. I don't know William, but I guess he probably make his analysis and he probably came up with a good business case.
>
>Regards,


I agree with Martin's opinion. It will take a long time to develop and test a good and reliable framework. And everytime a new version of the framework will be released, you may have to dig into your framework to use (and retest) latest features that makes sens.

Are you aware of the free CSLA? http://www.lhotka.net/cslanet/