Excerpts from Philip Brown's message of Fri Jul 31 14:41:08 -0400 2009: > I think applying the patch should be part of "get source", and there should > not be a mandated separate one.
Ok. For a 'broad strokes' approach, that would be acceptable, imo. (This implies that the 'fetch' step would also encompass 'extract' which is fine.) > > 3. I'd then say both 'build' and 'package' should be individually > > callable steps.... > > 4. A 'clean' target is always a good thing to support. > > 5. A set of variables should be honoured by any Makefile such that a > > build system knows where to find outputs. > I think this is making it too complicated. > It needs to be *simple*,for most widespread use. Ok, as long as there is some sort of convention where automated files could expect to find their outputs. The Variable approach lets the tool set that as it desires, which makes changing it somewhat easier, but as long as good choices were made up front... Are you suggesting that a build step separate from package is also too much though? I think that is an important distinction, but arguments to the contrary are welcome. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 GPG Key Id: 8E89F6D2; Key Server: pgp.mit.edu Contact me to arrange for a CAcert assurance meeting. _______________________________________________ maintainers mailing list [email protected] https://lists.opencsw.org/mailman/listinfo/maintainers
