[OpenAFS-devel] Call for consensus: configure defaulting behavior change

Jeffrey Hutzelman jhutz@cmu.edu
Wed, 29 Aug 2001 14:08:57 -0400 (EDT)


On Tue, 28 Aug 2001, Derrick J Brashear wrote:

> On Tue, 28 Aug 2001, Jeffrey Hutzelman wrote:
> 
> > > alternately continuing to create a dest area as now as
> > > "staging" and either springboarding to a final install or creating native
> > > packages from it...
> > 
> > I don't think we can avoid creating the 'dest' tree, because too many
> > people are already depending on that layout.  Similarly, it would be bad
> > to change the layout of what gets put into 'dest', since that would negate
> > the value in keeping it.
> 
> Who are these people and what are they doing? Anyone? (Not that keeping
> the complete tree is hard)

I think Derrick and I have both been less than clear here.

What I meant was, we can't avoid _being able_ to create ${SYS_NAME}/dest.
I didn't mean that running 'make' would always have to generate such a
tree; it should be perfectly reasonable to move to a model where 'make'
just compiles stuff, and people who need something in the form of
${SYS_NAME}/dest have to make some other target to get it.


What Derrick meant was, are there any people who depend on getting
${SYS_NAME}/dest as a side-effect of running 'make', and wouldn't be able
to deal with having to do something likke 'make dest' instead?

-- Jeff