yade-dev team mailing list archive

[Bruno Chareyre <bruno.chareyre@xxxxxxxxxxx>]

>     > Ok, I think it is good to have such tests anyway. We can keep them
>     > separately from regression ones.
>     > Let's call "check tests" or...
>     >
>     > And they should not be mandatory for packaging. "Just to check". If
>     > somebody commits something and it sufficiently changes result
>     value - it
>     > can be serious.
>

Ok, I have a few "check-test" designs in mind, but now I'm wondering
exactly how they should be integrated.
Clearly, there should be some automatic check-testing and reporting,
else it will be just adding more scripts in trunk and we can't expect
that they will be tested after each commit.
 
Anton, suggested that they should not be mandatory for packaging. I'm
not sure what it implies technicaly.
We could perhaps also disable failure/success mails for this special
group, being able to browse per-commit reports and find _when_ something
started failing is enough. 

Should we add a script similar to __init__.py, that would call check
tests and generate reports? Or is it better to include them in
__init__.py, setting them not-mandatory somehow?

Cheers.

Bruno