[skoll] GME

[Don Hinton]

Hi Arvind:

> I understand what you are suggesting i.e have nightly build that will 
> build GME to spot compile errors etc when GME developers commit stuff. But 
> again, at this stage, I don't know if we can test *features* in a 
> scriptable format. 

Just because it hasn't been scripted successfully, doesn't mean it can't 
be.  But that's not the point.  The point is that we'd like to use skoll 
for more than just ACE+TAO, right?  In fact, as stated in the NSF 
proposal, we'd like to use if in skoll.  Well, *all* the other skoll tools 
at ISIS are windows only, so in order to do that, we need to have the 
skoll clients run on windows.

We also need them to be able to do MSVC++ builds too.  I don't think it 
will be hard for Cemal to set it up, but since it hasn't been done/tested 
yet, there may be some work involved.  I just gave him the details about 
how it could/should work.  The GME team will be responsible for writing 
tests/scripts for GME regression tests once we give them a framework that 
works.  If guitar will help, great.  But doubts about the GME testing 
phase shouldn't have anything to do with getting skoll clients running on 
windows, should they?

thanks...
don