Lan Barnes wrote:
Fair enough, although IMHO the "write the test first" thingie in agile-dom
is YA rewrapping of RAD and other "we doan need no stinkin requirements"
thinking. Show 'em, do it, repeat, rinse ... developers love that. Only
problem is, it (almost) always ends up a mess. And then they have to find
a new rubric under which to do it all again.
Agreed. I do not believe in unit tests as a substitute for design.
I'm also not convinced about test-first on initial code writes where you
don't have knowledge about "correct", yet. You wind up writing a *lot*
of tests that get thrown out because the code is still too malleable.
I tend to write one or two core behavioral tests while the code is
highly malleable.
However, the *moment* I have to track a bug, unit tests get written.
The unit tests serve as breadcrumbs through the code that guide me
through the code I have checked thoroughly, and the code that may not be
quite so well checked.
The unit tests also serve to monitor how many edge cases were forgotten.
If too many unit tests start piling up in debugging, that area needs a
design rethink.
-a
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
