Proposal: mechanism for local patches

[Jim Trigg]

On Tue, Dec 2, 2008 at 9:08 PM, RW <fbsd06 at mlists.homeunix.com> wrote:
> I wonder if portsnap actually needs to behave the way it does.
>
> Portsnap stores its compressed snapshot as one .gz file for each
> port plus one for each additional file (files in Mk/ etc). When you
> do an "update" any modified snapshot files are extracted over
> the appropriate location in the ports tree.
>
> The reason that "portsnap extract" deletes patch-files is that before
> each .gz file is extracted, the corresponding file or port directory is
> deleted. I wonder why, if an "update" can decompress over the top of a
> port, an "extract" need to delete it first. I can't think of any good
> reason offhand.
>
> Modifying portsnap not to delete extra files is just a matter of
> deleting one line. The behaviour of portsnap extract would then be
> virtually identical to csup. Alternately, it wouldn't be much harder to
> create a new portsnap command.

I would presume that it does that to get rid of "standard" patch files
that are no longer part of the port...

Jim Trigg