UVR wrote:

<snip>

> Can't say I disagree with your position, having advocated it myself in
> the past.  However, thus far there has been a lack of interest amongst
> CTI-TET users to use checkenv (or to RFE that CTI should use checkenv).
> Please feel free to submit one: stc/cti/tet.

RFE 6818022 filed.

> At present, most CTI-TET test suites just do the checks they need sans
> checkenv.  It works mainly because the tests currently being performed
> are simple and trivial ones like
>         [[ -x some_binary ]] && do_something
> If the checks you need are also of this type, you may want to choose a
> similar approach.  This is not to say you shouldn't file the RFE though.

That's what I am doing but find it sub-optimal even for the minimalistic 
nc test suite. I guess most of the authors of CTI-TET based test suites 
realized that checkenv cannot easily used and reverted to their own 
solutions when developing. IMHO, a test suite framework should offer the 
capability of checking the environment before the configure and execute 
phases (in robust way) to offer a standard which can be adhered. For 
CTI-TET it should be sufficient just to stick a checkenv_def file to the 
top level directory of given test suite and the framework will do the rest.


v.

Reply via email to