Re: TDD and Refactoring
From: Phlip (phlip_cpp_at_yahoo.com)
Date: 12/26/03
- Previous message: JXStern: "Re: TDD and Refactoring"
- In reply to: Harry Erwin: "Re: TDD and Refactoring"
- Next in thread: Uncle Bob (Robert C. Martin): "Re: TDD and Refactoring"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Fri, 26 Dec 2003 17:05:05 GMT
Harry Erwin wrote:
> The issue for me is that baby steps do not lead to major design
> improvements that are based on rethinking a larger scale pattern.
The goal is not baby steps - that's just the steps towards the goal.
My "Deprecation Re*" used baby steps in the new module, but not from the
deprecated module. So they go with the full benefit of learning from its
design & its mistakes.
The goal is if you are busy deep in a refactor party, and someone busts thru
the lab door and claims the biggest customer's orneriest representative
showed up unannounced and requested to see the latest features, you have the
option to not demo iteration's release. You can hit the test button, demo
the application, unscripted, and let the customer's representative play with
it.
For each command, control flow sometimes goes thru the old version, and
sometimes thru the new version, and the tests don't care.
-- Phlip
- Previous message: JXStern: "Re: TDD and Refactoring"
- In reply to: Harry Erwin: "Re: TDD and Refactoring"
- Next in thread: Uncle Bob (Robert C. Martin): "Re: TDD and Refactoring"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]