| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
I disagree that we need to add "exposed interface" as a qualifier to the pacing mechanism objective. One reason is that the word "exposed" within the context of 802.3 implies a connector which is not going to exist for the 10GMII. The other reason is that the pacing mechanism has not been selected, so there is no reason to limit its location within our objectives. I agree that in an implementation of an integrated MAC and PHY the need for a pacing mechanism may be moot, but that is implementation specific.
Thanks,
Brad
-----Original Message-----
From: Devendra Tripathi [SMTP:devendra.tripathi@xxxxxxxxx]
Sent: Thursday, September 09, 1999 2:59 PM
To: stds-802-3-hssg@xxxxxxxxxxxxxxxxxx
Subject: Re: Why not have both?
Paul,
>Would you accept a modification to item 4 which stating:
> 4) Agree that a pacing mechanism be employed to throttle the MAC's transmit
> data rate down to a rate which is compatible with the payload of
> OC-192c/SDH-64.
>If so, I think I could accept item 4.
In cases where phy and MAC are integerated, the pacing may become
irrelevant, as,
variable IPG or some other scheme could be used to that effect. Thus if
we should have the above statement, it should be qualified with the case where
MAC/PLS to PHY interface is exposed.
Thanks,
Tripathi.