archieving test results

[Stefan Seefeld]

hi there,

knowing that Nathaniel is working on this topic, I'd like
to understand what use cases you have in mind as to how
to archivate results.

I'm thinking of a way to annotate a test run with an
identifier describing the version of the code being tested
(a timestamp, a revision number, or similar) as well as
target information (platform, a list of all tools being used,
etc.), and then store the test results in a database.

Of course, I would have to make sure that the code being
tested is under control of a versioning tool, and I need
full control over the environment that could affect the test
runs.

I may still want to run tests manually with locally changed
code, in which case the test results probably should not
go into the database.

Will this or a similar use case be supported ? I guess
I could just write a 'result stream' myself that does
just what I describe. So the question is: will this kind
of functionality be supported by qmtest itself, or is that
something for an extension ?

How is this done for gcc, with a very high number of tests and
targets ?

Thanks,
		Stefan