Steve James <s.james@bigfoot.com> writes:
> On Sunday 22 Jun 2003 10:25 pm, Robert Pluim wrote:
> > Steve James writes:
> > > On Sunday 22 Jun 2003 9:24 pm, cmpilato@collab.net wrote:
> > > > Thanks for the find-n-fix, Philip. I'm still confused as to what made
> > > > this bug appear, but I'm content to know it's been killed.
> > >
> > > Are the tests always run for both local repositories *and* DAV?
> >
> > Nope. You have to explicitly request DAV testing (because you need to
> > have compiled mod_dav_svn and set up apache, which not everyone does).
> >
> Hi Robert. More precisely, I meant are the tests run for both access
> modes by the project 'process'? I'm presuming there's a collection
> of platforms doing smoke testing somewhere "out there" -- I've seen
> the corresponding mailing list. For this system, it would be
> appropriate to set-up for DAV testing. If this is true, I wonder if
> there a hole in the system? Regards, Steve.
To run the test suite over ra_local, ra_dav, and ra_svn is, I'm sorry
to say, just plain prohibitive to forward motion (generally
speaking). I mean, we're talking about a four-hour process for some
folks. So yeah, I admit that I don't run the tests over all three RA
layers with every commit.
My policy is just to try to use some common sense when deciding what
to test. For example, a change to mod_dav_svn doesn't need any
testing except over ra_dav -- no other RA layers should be touching
that code.
In the case above that Philip solved, the problem, by all indications,
should *not* have been RA-layer-specific. (In fact, I *still* have
trouble believing that they really were RA-layer-specific.) So I ran
only the ra_local tests.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Jun 23 00:35:16 2003