Re: pkg/45284: native Kerberos on Solaris (again)

[Jörn Clausen <joern.clausen%uni-bielefeld.de@localhost>]

The following reply was made to PR pkg/45284; it has been noted by GNATS.

From: =?ISO-8859-1?Q?J=F6rn_Clausen?= <joern.clausen%uni-bielefeld.de@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: Thomas Klausner <wiz%NetBSD.org@localhost>, 
pkg-manager%NetBSD.org@localhost,
 gnats-admin%NetBSD.org@localhost, pkgsrc-bugs%NetBSD.org@localhost
Subject: Re: pkg/45284: native Kerberos on Solaris (again)
Date: Fri, 26 Aug 2011 11:01:24 +0200

 >   On Fri, Aug 26, 2011 at 10:33:35AM +0200, Jörn Clausen wrote:
 >   >  Again: My question is: Why is this defined at all in this package?
 >
 >   Because sometimes there are changes which break the API for all
 >   customers.
 
 This may be true for a package that can only be provided by pkgsrc. But 
 for a package that can be replaced by a native one, this does not make 
 sense. Because the change in the pkgsrc package is not felt by the 
 customers.
 
 >   And a minimum requirement also makes it easier to package software,
 >   since you don't have to specify a minimum version in every package (as
 >   long as it can live with the minimum version and doesn't need a newer
 >   one, then you have to add an API depends there).
 
 But what if the majority of the dependencies would be happy with an 
 older version? And why does not every other package define its version 
 this way?
 
 BTW, if this is not clear: I *want* to keep the native version, even if 
 it is older than the pkgsrc version. Because I assume that the binaries 
 in the native version are able to e.g. make use of cryptographic 
 hardware. And as long as I have no dependencies that need v1.8 of 
 Kerberos, I don't see why I am forced to use the pkgsrc version of this 
 package.
 
 -- 
   Jörn Clausen                             
joern.clausen%uni-bielefeld.de@localhost
   Hochschulrechenzentrum                 http://www.uni-bielefeld.de/hrz/
   Universität Bielefeld