>>>>> On Thu, 30 Oct 2003 09:18:07 -0500, Alain Javier Guarnieri del Gesu <[EMAIL
>>>>> PROTECTED]> said:
> Sound argument. Are you suggesting then that a distinction be made
> between unit, integeration, and acceptence tests in the project
> directory structure?
Yes, this is something I've wanted for quite sometime, just never
really figured out a good way to do it.
Thinking about concentric circles, three rings, our code in the center...
*Programmer tests* ( unit tests ) are closest to the code in the first
ring. These are run with every compile of the code ( Some could even
say there are four rings and the compile is the first ring of our
tests. If it doesn't compile, its almost like a test failure )
Third is our *programmer integration tests* and they may run after all
projects have been built and their programmer tests pass. These are
really slow programmer tests or tests that have interactions with code
from different products or external resources ( files ). Sometimes
these are just poorly written programmer tests ;)
On our outer ring is the *customer tests*, these run nightly and may
have full end-to-end interaction with deployed systems.
The number of rings could vary. I like my tests as close to the code
as possible.
--
=====================================================================
Jeffrey D. Brekke [EMAIL PROTECTED]
Wisconsin, USA [EMAIL PROTECTED]
[EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
