Re: SCons COM vs COMSTR tests
Bill Deegan <bill-cJFiu+DHMVC5azolltMz9laTQe2KTcn/@public.gmane.org>
| Newsgroups | gmane.comp.programming.tools.scons.devel |
|---|---|
| Message-ID | <CAEyG4CF1p719PCao-Wn3GcRmLM4o+8HBj2Jdrx__NW92nDjFmg@mail.gmail.com> |
William, I think I disagree. The ability to set one different from the other should work and should be tested. Though, this really comes from the Action Object.. I think ideally we'd be able to easily separate running unit tests from the system tests. (Though you can do that by specifying which parent directory now). -Bill On Wed, Sep 21, 2016 at 10:15 AM, William Blevins <[email protected]> wrote: > Team, > > I was looking at some tests and they seems to overlap in such a way that 1 > is subsumed by the other. > > For example, test/Java/RMICOM.py and test/Java/RMICOMSTR.py are identical > except that one sets only the RMICOM variable and the other sets both > RMICOM and RMICOMSTR (which of course means that one has extra stuff in the > stdout). It seems to me that RMICOMSTR subsumes the other, and only 1 test > has value. > > Having a lean test harness is a bonus. Tests take a long time to run > (especially if we are running redundant ones). > > V/R, > William > > _______________________________________________ > Scons-dev mailing list > [email protected] > https://pairlist2.pair.net/mailman/listinfo/scons-dev > > _______________________________________________ Scons-dev mailing list [email protected] https://pairlist2.pair.net/mailman/listinfo/scons-dev