Re: Why not have both
Howard:
We are in agreement on these points, provided the generic MAC or MAC
interface provides a mechanism for use by WAN PHYs to pace the effective
MAC data rate at the OC-192c/VC-4-64c payload rate. I believe the PAR needs
to include a statement on a pacing mechanism, but does not need to select
between the many options discussed on the reflector.
Paul
At 10:14 AM 9/10/99 -0700, Howard Frazier wrote:
>
>
>
>As others have pointed out, the model I presented is just that, a
>model. It is not intended to constrain implementations, but it is
>intended to define the scope of work for the 10 Gigabit Ethernet
>standards project.
>
>Implementers can, and probably will, build a variety of devices.
>That's the wonderful thing about implementers. They are creative, and
>they respond to the needs of their customers. As standards weenies, our
>job is to write a document that describes the essential requirements
>for interoperability, and let creativity, inovation, and the demands of
>the market define the product specs.
>
>If we can agree that we will
>
> Have two PHYs, one for the LAN, and
> one for the WAN
>
>and we can agree that
>
> Within the context of the 802 standards
> architecture, the two PHYs operate below
> identical individual 802.3 MACs
>
>and we can agree that
>
> The PHYs may require the definition
> of new management attributes
>
>Then we will have made a huge amount of progress, and we will be able to
>draft and gain approval of a PAR with the supporting 5 Criteria. The
>standard that we eventually produce will fit nicely into the family of
>802 standards, and implementers will go out, work their magic, and deliver
>interoperable products that customers will just love.
>
>Howard Frazier
>Cisco Systems, Inc.
>
Paul A. Bottorff, Director Switching Architecture
Enterprise Solutions Technology Center
Nortel Networks, Inc.
4401 Great America Parkway
Santa Clara, CA 95052-8185
Tel: 408 495 3365 Fax: 408 495 1299 ESN: 265 3365
email: pbottorf@xxxxxxxxxxxxxxxxxx