Subject: Re: Version Naming/Numbering
To: Matt Thomas <matt@3am-software.com>
From: Christos Zoulas <christos@zoulas.com>
List: current-users
Date: 09/17/2004 19:52:15
On Sep 17,  4:42pm, matt@3am-software.com (Matt Thomas) wrote:
-- Subject: Re: Version Naming/Numbering

| At 04:38 PM 9/17/2004, Christos Zoulas wrote:
| >In article <20040917210706.GE1281@netbsd.org>,
| >Bill Studenmund <wrstuden@netbsd.org> wrote:
| > >
| > >I thought after the 2.0 release we were going to move current's version so
| > >that we didn't have this problem. I've forgotten what the exact fix was,
| > >but I know there are a few ideas floating around. I don't remember if it
| > >was we were going to 2.9 or 2.99 or we were going to 3F & friends and
| > >arranging it so that NV(3.0) > NV(3F).
| > >
| >
| >That's what I thought too, and 2.99 was luke's proposal IIRC. Can't we just
| >agree on that and move forward?
| 
| Does that mean 2.99A, 2.99B, 2.99C, .... 3.0?
| Or 2.99.1, 2.99.2, ... 3.0?
| Or 2.99, 2.991, 2.992, ... 3.0?

I am not particular to any of the first 2 [the third might not fit well,
because the minor version is 2 hex digits in __NetBSD_Version__].

The first has the advantage that it makes it clear that this is not a release
because the scheme is different, and the second has the advantage that it is
not different from the release scheme :-)

I think I prefer the second.

christos