"I can refactor a class diagram in Rose 10 times faster than the fastest
Smalltalk programmer with the slickest development environment can rewrite
and retest the code. So I can get dozens (hundreds on a larger project) of
refactoring iterations in before code while working through the scenarios."
You're making a bet that you are right when you do those refactorings.
You're comfortable making that bet, and more power to you.
I can't do that. I tried and it just didn't work for me, so I started going
in the other direction. The more situated my decisions, the faster I went
and the less stressed I was. So I turned the knob all the way to 10 and it
still worked for me.
When I refactor, I am absolutely certain the system is getting better
because the code is getting better (closer to the simple design ideals). If
the code is not getting better, I stop.
There's a lot we agree on- the need to learn as we go, the need to
transform artifacts instead of inventing new things all the time, the
importance of careful analysis and design thought. I can't stand the risk
of not having the feedback of code and tests. It sounds like you can't
stand the extra friction of having to transform the code instead of just
the diagrams, at least early in development.
The tortoise and the hare, but this time it's not clear which way to bet.
Do NOT sent signoff request to email@example.com, please!!!
To unsubscribe from this list, please send email
Body: unsubscribe otug your_full_subscription_address
This archive was generated by hypermail 2b29 : Thu Jul 13 2000 - 22:37:57 EDT