开云体育

I like Hill & Charlie's answers...

I figure when I have a pre-conceived idea for the design, it's because at
least some situations have pressures that make the "known" designs a good
idea.

When I have a theory about a pressure existing, I'll choose a test in that
area and see if it reveals the stress point I think exists (and would be
solved by my design idea). (Sometimes I'll find the design idea isn't worth
it, other times it is.)

It pays to be sensitive to the refactoring step, using it to create the
needed concepts/objects/functions/whatevers.

What I think all this does is let the generality I really need emerge, and
leave the rest in reserve should it be needed in the future.

It's not to say I've never just added a class I thought would be needed,
but I think the emergence is interesting to watch & see what drives it. One
of the things I really like about Kent Beck's TDD book is his use of the
basic "currency" example, but he ended up with a different solution than
what he'd usually created.

--Bill Wake

--
Bill Wake
Industrial Logic, Inc. @IndustrialLogic
Coaching | Training | Assessment | eLearning