Hi Hugh,

Thanks for your reply.
Actually, Whil's book gave me the idea when he spoke about action points and that Geek-With-Clark-Kent-Glasses-And-A-Stopwatch.
I am now working out that idea.
Of course, any spec leads to a certain number of AP's (actionpoints).
If any of those AP's is within the framework one developer uses, as a standard, the client should still pay for those as well. After all there is some time and money spend on getting/developing that framework.
It would be the same as to say to Philips that you're not willing to pay for the tube that is in the TV because they did the development of that component a long time ago.
To direct this to software, if a form needs a toolbar with buttons for certain actions those are all AP's for me whether or not I have the toolbar developed already as a standard in my framework.

You are right if you say that it is hard to make an estimate on new work after collecting data on one app only. However, the more one develops the more data one has and the more acturate the outcome of an estimate will be.

Not only did I read whil's book but also the books by Steve McConnell, "The Software Project Survival Guide" and "Rapid Software Development" as well as "Code Complete". he gave me some ideas on the interpretation of the figures as well. These 4 books (from Whil and Steve) are "biblical-like" stuff for any developer, but hey, so are "about face" and the automation bible from Tamar and Della.

Boudewijn.