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

[802.3AZ] Comment #110


In draft D2.1 we had a comment that identified an error in the nomenclature used by the Layer 2 (DLL) section of the draft. Specifically, that the nomenclature had not been updated to what we agreed to use in the wake-shrinkage ad-hoc and that the initialization values used should point to the minimum values specified for each PHY type. The BRC agreed with the comment and accepted the remedy. Unfortunately, the adopted remedy introduced some unintended consequences:

(a) The nomenclature needed to be updated in all other associated L2 sub-clauses in the draft.

(b) Changing the initialization states in the two machines to point to LOCAL INITAL TX VALUE and LOCAL INITIAL RX VALUE respectively leaves the init values unconstrained as these constants, as currently defined, have values picked by each system.

(c) Deleting PHY WAKE VALUE completely would have clashed with Comment 111 as the later asks to round up those values and the former deletes the item.

Rather than attempt to fix the unintended errors introduced by the adopted remedy, which would have been outside the scope of our process, the editor implemented the remedy agreed upon, reviewed the issue with the commenter and several L2 experts and added an editor's note to highlight the issue. Hopefully this email can serve to start a discussion on how best to remedy the error in the next comment cycle.



Michael J. Bennett
Chair, IEEE P802.3az, Energy Efficient Ethernet Task Force
Tel. +1 510 486 7913