|Thread Links||Date Links|
|Thread Prev||Thread Next||Thread Index||Date Prev||Date Next||Date Index|
I’ve been slow to catch up with this thread and want to thank Brad for summarizing the discussions that happened later in the week after John’s ICAID meeting. I think Brad has captured well the essence and general consensus that we came to.
We had a lot of experienced 802.3 leadership perspectives that were concerned about the feasibility of one large project potentially handling 3 MAC rates simultaneously. We also heard the recommendation that we make sure we don’t initiate a process that may, due to IEEE rules, prematurely terminate any path of study we want to do.
It is clear we have interest in ensuring progress on both single lane 50Gb/s Ethernet and multiples (100Gb/s and 200Gb/s) Ethernet. We’d be defining new MACs and PCS’s where we don’t have them as well as new PMDs for new and existing MAC rates.
I’m putting together the CFI request to capture this. I’ll be setting up a series of pre-CFI review meetings for those who are interested to attend. I’ve gathered an email distribution list already from those who asked to be included after the May and July Flash mobs we had on the topic. I’ll send meeting details to that group to not flood the dialog reflector - please send me an email if you want to be included and aren’t on the list (or not sure you are on the list).
On 9/24/15, 12:12 PM, "Brad Booth" <bbooth@xxxxxxxx> wrote:
Unfortunately you decided to head home early and missed some of the discussion that occurred in Bonita Springs.
One aspect that I liked in the discussion was the formation of two projects: 50G and n*50G. They could have one consensus meeting, two CFIs, two study groups, two PARs and two task forces that are coordinated and work in parallel. The nice aspect of this is that it permits each project to operate on a timeline that suits their requirements, rather than creating a dependency that could impact one or the other.
Splitting a PAR after the project has started is something we should try to avoid. Amending and modifying PARs is also something worth avoiding. Both add delays to the work being done.
So, while there are some that may feel there is opposition to 200G or n*50G as being a project in 802.3, I believe that is a false interpretation. To me, it is not a question of do we need to do 200G/n*50G, but rather what is the most efficient and effective means to manage and accomplish the work required. IMHO and with past experience as a chair and chief editor, I believe that two projects would be the optimal path forward.
On Wed, Sep 23, 2015 at 10:44 AM, Chris Cole <chris.cole@xxxxxxxxxxx> wrote: