I have copied this to both the 802.3ch and 802.3cg reflector, since both projects make reference to single-pair autoneg.
Mehmet Tazebay noticed what appears to be a typographical error in the transmit state diagram for single-pair autoneg. We discussed this this morning on the 802.3ch ad hoc call, and Brett McClellan (the original contributor of the figure)
agrees that a typo seems to have been introduced in the editorial process.
On 4 transition branches, “*” (AND) operators, appear to have been replaced with “+” (OR) operators.
Researching the issue, I found that the original contribution that proposed the state diagram had these as “*”
http://www.ieee802.org/3/bp/public/nov14/mcclellan_3bp_03_1114_%20Autoneg_baseline_text_proposal_v0p4.pdf See page
25, and it was implemented as “*” in draft 1.1 (the first place this showed up): (see e.g., page 88 of D1.1), through 2.0. For example, they show:
complete_ack=false
∗
transmit_ability=true
∗
transmit_mv_start_done
However, in D2.1, the figure was redrawn (to change font sizes) according to comment 316 on draft 2.0.
When the state diagrams were redrawn it changes to the current form, with an OR (+) in Draft 2.1 – and it has remained so ever since.
I checked to see if there were any other comments on D2.0 changing this, and found none. (I can find only 2 comments on D2.0 relating to page 180 (one is comment 316, the other is unrelated), and I can’t find anything relating to “complete_ack”,
“transmit_ability”, or “transmit_mv_start_done”. )
I can only conclude this was a typo made on the implementation of comment 316 going from Draft 2.0 to Draft 2.1, which remained uncaught.
The proposal is to fix these in the revision draft through a late comment, as the most expedient way to fix a clear error. It is important that if anyone thinks this is OTHER than a clear typographical error, that they please speak up.
Below, I have copied Mehmet’s description of the typo and the reference figures so that you can see the branches which need to be changed.
(from Mehmet Tazebay):
-
Figure 98-8 has several typos. Several places an OR is indicated for state transitions when the condition should be an AND. Clause 98 is based on Clause 73. There are some important differences but figure 73-9 shows the expected behavior for the state transitions
that are common between them. (see figures below). This is definitely a problem with Figure 98-8.
Thanks,
Mehmet
-george
George Zimmerman, Ph.D.
President & Principal
CME Consulting, Inc.
Experts in Advanced PHYsical Communications
310-920-3860