Re: [Gnu-arch-users] NEW POLICIES (draft)

[Aaron Bentley]

Andrew Suffield wrote:
> On Mon, Oct 04, 2004 at 02:59:47PM -0400, Aaron Bentley wrote:
>
> > Andrew Suffield wrote:
> >
> > > On Sat, Oct 02, 2004 at 03:11:47PM -0700, Thomas Lord wrote:
> > >
> > >
> > > > ** The Solution: Cascade Branches
> > > >
> > > > There is a simple and elegant solution to the problem with 
> > > > long term branches and that is to use "cascade branches",
> > > > as illustrated in ./cascade.ps.
> > >
> > >
> > > On reflection, this is just a prism structure under a different name,
> > > plus a rule that there must be a single line of ancestry through all
> > > of them (which is actually useless and should be ignored - this isn't
> > > used for anything, so regular freeform prism structure will work just
> > > the same; the freeform structure is just a set of 'cascades' anyway).
> >
> > The single-line-of-ancestry rule means it's not a prism.
>
>
> Not one, but several. It's a sequence of prisms tacked on top of each
> other. 

Yes, but they're prisms with an input of one, and that has a qualitative 
difference; when the same upstream merges arrive by more than one input, 
conflicts will be produced.  Conflicts due to duplicate upstream merges 
cannot happen with an input of one.

Aaron
--
Aaron Bentley
Director of Technology
Panometrics, Inc.