At 9:18 AM +0100 7/4/01, Vincent Massol wrote:
>Regarding chaining, I absolutely do not wish to preserve any state as in my
>opinion any tests _must_ be independent of another. These are _unit_ test. I
>agree that for functional and acceptance tests it is useful but not for unit
>tests where it is very dangerous (side effects) and breaks the model.

The state is maintained in an object which is instantiated for each test, and is 
therefore not maintained across tests. The intent is to verify that you can set up a 
particular state in one call and use it in another, but that is part of a single unit 
test.  As I say, it is not clear to me that this is necessary for in-container tests.

-- 
------------------------------------------------------------------------
Russell Gold                     | "... society is tradition and order
[EMAIL PROTECTED]                 | and reverence, not a series of cheap
                                 | bargains between selfish interests."
http://www.httpunit.org          |   - Poul Anderson, "Iron"

Reply via email to