On Tue, 13 Jul 2004, Johnny C. Lam wrote:

> I don't see any useful reason for any particular patch-naming scheme
> from a pkgsrc user's standpoint.  However, from a developer's standpoint,
> if a patch is linked via a canonical name to the file that it changes,
> then the CVS history for a patch will be much more likely to be
> relevant.

I fully agree with the cvs history argument. The particular scheme used
doesn't really bother me, as long as

  (a) it's documented,
  (b) the make target for regenerating patches supports it, and
  (c) cvs history becomes useful.

In the ideal world I'd also like

  (d) the filename to quickly identify the patch

but we don't currently have that, and as the vast majority of packages
have very few patches it's not proved to be a major problem (for me) so
far.

-- 
Gillette - the best a man can forget