On Thu, 1 Sep 2005 13:40:31 +0200
Lubomir Sedlacik <salo@Xtrmntr.org> wrote:
> On Thu, Sep 01, 2005 at 07:32:56AM -0400, D'Arcy J.M. Cain wrote:
> > However, I can see how this is different in that the package(s) that
> > it fixes are other than the one being patched.  But the Python24
> > package itself wasn't broken.  If we bump the PKGREVISION wouldn't we
> > also want to change all the extension dependencies to require that new
> > revision as well?
> 
> why?  if the python modules built, there is no reason for them to bump
> their PKGREVISIONs.  if they didn't, well there is a new version of
> python24 for a reason.

No, I didn't mean to bump the PKGREVISION in the modules.  I meant that
the modules needed to recognize that the PKGREVISION in python24 needed
to be noticed.  Otherwise the modules will continue to fail to build
with no indication why.

In addition, the modules don't need to be built with the fixed python24
package if the python24 package already built precedes the original
fix.  They will still build with that older version.

It gets messier.  The previous fix didn't bump the PKGREVISION so we
can't point to a particular version and say that it needs to be updated
if you are at that level.  Mind you, I bumped the PKGREVISION 4 days
later for an unrelated fix so the window is only 4 days long.

-- 
D'Arcy J.M. Cain <darcy@NetBSD.org>
http://www.NetBSD.org/