Re: Some observations regardig the progress towards Debian 3.1
On Sat, Nov 29, 2003 at 02:02:24PM -0600, Steve Langasek wrote:
> On Sat, Nov 29, 2003 at 07:57:47PM +0100, Adrian Bunk wrote:
> > On Thu, Nov 27, 2003 at 07:53:47PM +0100, Goswin von Brederlow wrote:
> > > Adrian Bunk <email@example.com> writes:
> > >...
> > > > I haven't found it explicitely mentioned, but the logial version number
> > > > for a binary NMU of version 1.0 would be 1.0-0.0.1 .
> > >
> > > A binary NMU implies you haven't changed the source. If you change the
> > > version number you have changed the source and must upload it too.
> > > Thus binary NMU must have the same version number.
> > >...
> > That's wrong.
> Nevertheless, 1.0-0.0.1 is an impossible version number for a
> recompile-only NMU unless the previous MU was numbered 1.0-0 (which
> would be quite atypical).
> You could have 1.0-0.1.1 (recompile-only NMU of a source NMU of a
> non-native package) or 1.0-1.0.1 (recompile-only NMU of a MU of a
> non-native package), but 1.0-0.0.1 would be quite strange indeed.
> Incidentally, I don't see that the developer's reference specifies how
> to handle a recompilation-only upload for a native package. Would that
> be 1.0.1? 1.0-0.0.1?
And I also have a problem with using "1.0-0.1" as meaning "NMU of
package whose previous revision was 1.0", since in a non-native
package it would mean "initial version of upstream 1.0, as NMU".
There is a big difference in between, since the 1st one refers to a
(usually) simple fix over 1.0, and the second one implies that 1.0 is
a brand new codebase !
That (and the problem you mention) would not happen if native Debian
packages systematically had "-1" appended. But that looks like an
Yann Dirson <firstname.lastname@example.org> | Why make M$-Bill richer & richer ?
Debian-related: <email@example.com> | Support Debian GNU/Linux:
Pro: <firstname.lastname@example.org> | Freedom, Power, Stability, Gratuity
http://ydirson.free.fr/ | Check <http://www.debian.org/>