[AFS3-std] Re: rxgk: Version numbers in XDR structures

[Matt W. Benjamin]

Hi,

Right, I don't see how we are altering XDR, rather rx.  I would have said that by creating the rxUnion we have increased the potential utility of rx for some hypothetical applications other than afs, but the utility remains potential until such an application (using the rxUnion, in fact) exists? 

Also, it wouldn't seem that extension of rx with new primitives changes things for rx standardization--who would do it, when it 'must' be done, etc.  Is there a group other than the afs community with an expressed interest in standardizing rx?

Thanks,

Matt

----- "Jeffrey Altman" <jaltman@secure-endpoints.com> wrote:

> On 3/10/2011 11:22 AM, Andrew Deason wrote:
> > On Thu, 10 Mar 2011 11:08:07 -0500
> > Jeffrey Altman <jaltman@secure-endpoints.com> wrote:

> afsUUID is not a fork of XDR.  It is a definition of an XDR construct
> that is assumed to be present as a building block for all AFS3
> protocol
> definitions.  Its scope is to the AFS3 protocol which does not
> include
> RX.  The RX protocol is independent of AFS3 and so its security
> classes
> must be implemented independently of AFS3 XDR extensions.
> 
> RX could define its own rxUnion that parallels the afsUnion. 
> However,
> as soon as we have done that we have demonstrated that there is a
> need
> for the type that extends beyond just AFS3 and it should be
> standardized.
> 
> Jeffrey Altman

-- 

Matt Benjamin

The Linux Box
206 South Fifth Ave. Suite 150
Ann Arbor, MI  48104

http://linuxbox.com

tel. 734-761-4689
fax. 734-769-8938
cel. 734-216-5309