Re: [MOBILE-IP] New GRE Draft Extensions

[Rene Tio <tor@redback.com>]

On Mon, Mar 13, 2000 at 09:50:03AM -0800, Gopal Dommety wrote:
> At 06:43 AM 13/03/00 -0800, David Meyer wrote:
> >
> >> This is an addendum to the draft-meyers. Using a different version
> >> number will definetly make it easy. If the version number is 
> >> to be changed then we have to change it in  draft-meyers. 
> >
> >	We've been through this and decided not to change the
> >	version number.
> 
> I am not suggesting we change the version number. I was clarifying that we
> cannot change the version no in draft-dommety. If at all it is done
> it has to be done in draft-meyers.
> 

I don't see why this has to be so.  draft-meyer is pretty much completely
backward compatible to 1701, whereas draft-dommety isn't.  Thus draft-meyer
doesn't require a version number bump but draft-dommety in its original form
did.

However, if the Sequence Number field were changed to 32 bits, as Gopal
recently suggested, and the only change was to better define the behaviour
of the Sequence Number field, then the need for bumping GRE versions I feel
would be unnecessary.

If, however, you are going to add an Ack sequence number, *and* use the Key
field as some form of tunnel demux, then it's beginning to look a whole lot
like PPTP and I'd probably suggest splitting out the PPTP GRE portions into
a separate draft and making draft-dommety look like it (and bump the version
numbers, eck).

This discussion is sounding very circular at the moment.

Cheers,
Rene