Re: pkg/46308: pkg_add should not display warnings for known compatible platform mismatches

[Joerg Sonnenberger <joerg%britannica.bec.de@localhost>]

The following reply was made to PR pkg/46308; it has been noted by GNATS.

From: Joerg Sonnenberger <joerg%britannica.bec.de@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: 
Subject: Re: pkg/46308: pkg_add should not display warnings for known
 compatible platform mismatches
Date: Sat, 7 Apr 2012 21:21:58 +0200

 On Sat, Apr 07, 2012 at 07:10:06PM +0000, Jan Schaumann wrote:
 > The following reply was made to PR pkg/46308; it has been noted by GNATS.
 > 
 > From: Jan Schaumann <jschauma%netmeister.org@localhost>
 > To: gnats-bugs%NetBSD.org@localhost
 > Cc: 
 > Subject: Re: pkg/46308: pkg_add should not display warnings for known
 >      compatible platform mismatches
 > Date: Sat, 7 Apr 2012 15:07:18 -0400
 > 
 >  --S6vg04ofUPzW4qJg
 >  Content-Type: text/plain; charset=us-ascii
 >  Content-Disposition: inline
 >  Content-Transfer-Encoding: quoted-printable
 >  
 >  Joerg Sonnenberger <joerg%britannica.bec.de@localhost> wrote:
 >  =20
 >  >  On Sat, Apr 07, 2012 at 06:10:01PM +0000, 
 > jschauma%netmeister.org@localhost wrote:
 >  >  > pkg_add displays a warning whenever a package is installed for a platf=
 >  orm that
 >  >  > does not match the current platform 100%.  Under NetBSD, however, we k=
 >  now that
 >  >  > certain platforms will be compatible.  In particular, any package buil=
 >  t under
 >  >  > major version N should work on other hosts running major N, even if th=
 >  e build
 >  >  > and deploy hosts have different minor versions.
 >  > =20
 >  >  That's not a correct assumption to make, IMO.
 >  
 >  We have incomaptible minor versions?  That seems like a problem to me.
 >  
 >  How about compatibility across tiny versions?
 
 Well, 5.1 is the state of the netbsd-5 branch at some point between
 5.0.0 and 5.1.0 release, right? So it might have the new X libraries or
 might not have them. Impossible to tell. x.y.z and x.y.z+1 are most
 likely compatible, not sure if we ever pulled up a binary incompatible
 ABI change. I'd still say it is backwards compatible only, e.g. new
 functions may exist in z+1.
 
 Joerg