Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [HSSG] Backwards compatibility

Title: Backwards compatibility
This is what I was getting at by backwards compatibility.  If we were to do a serial high speed - 40, 80, 100, whatever - it would be nice to smoothly upgrade one end at a time, as has been done in copper.  But I guess Brad answered the question - if there is nothing in the 10 Gig optical PHY's to negotiate with that door is already shut.  The only question is whether to start putting those kind of hooks in for the future.

From: Brad Booth [mailto:bbooth@xxxxxxxxxxxxx]
Sent: Tuesday, August 22, 2006 2:38 PM
To: STDS-802-3-HSSG@xxxxxxxxxxxxxxxxx
Subject: Re: [HSSG] Backwards compatibility

Good question.  Unlike copper PHYs that use Clause 28 auto-negotiation to determine the capabilities of each PHY and to use the highest common denominator, nothing to-date exists in 802.3 for the optical PHYs to do the same.  Depending on the approach taken by the task force (once formed), this may be a viable option to consider.  For example, an n x 10G solution could use a base channel to communicate the capabilities of each end of the link.

From: Kevern, James D [mailto:james.kevern@xxxxxxxxxxxxxxxxxxx]
Sent: Tuesday, August 22, 2006 11:25 AM
To: STDS-802-3-HSSG@xxxxxxxxxxxxxxxxx
Subject: [HSSG] Backwards compatibility

Excuse me for not knowing 1 & 10 Gig well enough, but is it reasonable to consider backwards compatibility with the corresponding implementations (i.e. short and long wavelength)?  Or is backwards compatibility a given?