The IP over ATM Mailing List Archive by date

[taddy@cc.gatech.edu (Rajesh Talpade)]

Hi Folks



> Andrew wrote:
> >>1a. *change* the interface definition in ipmc-12 section 6.2 to allow a MARS
> >> to use a point-to-point VC for the ServerControlVC if it knew
> >> it would never be adding more than one leaf - as written, a MCS is allowed
> >> to reject the ServerControlVC SETUP if it did not have the pt-mpt bit set.
> 
> Grenville wrote:
> > But since the text does not _require_ an MCS to reject pt-pt calls, ipmc-12
> > has the flexibility you're after already.
>
Andrew wrote:
> OK, then at least one reader of ipmc-12 got confused into thinking that
> section 6.2 was a part of the specification. Can you please put a big 
> disclaimer there somewhere noting that this is not a part of the 
> proposed/draft standard right now? If you do want it to be a part of
> the specification then it needs some more work.



We plan to enable an MCS to accept both pt-to-pt and pt-to-mpt calls
from the MARS as the ServerControlVC in the next version of marsmcs-01. The
MARS can then decide, as Andrew indicated, whether it wants to use a pt-to-pt
or a pt-to-mpt VC as the ServerControlVC.

As a matter of fact, it might be a good idea to resolve any issues that
one may find about MCS behavior in marsmcs-01, rather than
having to do another rewrite of ipmc-12.

This is one more reason for an MCS document to exist, so that MCS issues that
are left open in ipmc-12 can be resolved here. The marsmcs-01 was partly
targetted to satisfying this need.

However I am getting the feeling that some folks may not be keen on the
prospect of having to define one version of multiple MCS for MARSv1, and
another for MARSv2. Nevertheless, a document describing the MCS
architecture in detail is needed, to enable even a single _MCS_ to be used
_unambiguously_ with ipmc-12.

If there is general consensus on this observation, we can reduce the scope
of marsmcs-01 such that it describes the MCS architecture in detail, and leave 
the allocation mechanism out of it. The allocation mechanism in marsmcs-01 
could be made part of a separate document, an experimental RFC perhaps. 


Regards,
Rajesh.




-- 
Rajesh R. Talpade		(404)-315-1227 	     taddy@cc.gatech.edu
http://www.cc.gatech.edu/computing/Telecomm/people/Phd/taddy