Subject: Re: Version Naming/Numbering
To: George Michaelson <firstname.lastname@example.org>
From: Bill Studenmund <email@example.com>
Date: 09/23/2004 10:59:43
Content-Type: text/plain; charset=us-ascii
On Thu, Sep 23, 2004 at 04:23:16PM +1000, George Michaelson wrote:
> Confusing people is one thing.=20
> Breaking #if logic is another.
> Dealing with code is about a different kind of POLA than with people.
> When I find code won't make because of cpp problems, its very
> depressing. The track record for re-syncing changes into larger systems
> (X11?) isn't brilliant here. GCC for instance, could be expected to take=
> a while to bed in a significant alteration of how >=3D logic works for=20
> tuning the outcome of that code-make process.
> I very much hope that you can come up with a fix which works for both
> code and people, but if not, I'd suggest that making it work for code
> matters more.
We get to choose how we map text NetBSD version numbers into numbers for=20
comparison, so we can always choose something that works.
Also, note that checking version numbers is still questionable. We have=20
multiple branches of development, and some features get pulled from one to=
another. I remember we had one tool recently that was added, then pulled=20
to 2.0, then pulled to 1.6. So 1.6Z didn't have this tool, nor did 2.0A,=20
but 1.6.3 will.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (NetBSD)
-----END PGP SIGNATURE-----