RE: Dropping the Local Overbooking Multiplier (LOM) method from DS-TE specs?

["Francois Le Faucheur (flefauch)" <flefauch@cisco.com>]

Jerry,

>> Also, I argued against having LOM progressed separately, I 
>> think that would just complicate matters, and leave the BC 
>> models, Proto extensions, etc. in a state of limbo (possible 
>> awaiting a LOM extension later).  I had argued to either 
>> keep LOM, or drop LOM, but not leave it in the 'in between' state.

I'd like to come back to your point above because:
	- (i) it is a concern that Sanjaya also expressed
	- (ii) I think we can find a way which does not leave things
more in a limbo than keeping it in as an option
	- (iii) I have the impression that putting LOM in an
Experimental draft is a solution that everybody else could probably
leave with (even if not their first choice) so this may be a way to get
out of the "impasse". Ie it is the second choice of people who favor
keeping LOM in the spec and of those who favor dropping LOM altogether.

We could basically take the exact same text that is currently spread
over -proto, -mam, -russian (and perhaps -mar if you've put some in
there) and group it into this new -LOM draft. So frankly, it terms of
how things are specified to work it would just be perfectly equivalent.
If your concern is that MAR may get left out, we could work on that,
perhaps try to work together on a section for the MAR-specific LOM
aspects (just like there would be a russian-specific and mam-specific
section cut-and-pasted from -russian and -mam).

The only real difference in making the same text appears in Exeprimental
draft is really to say that the TEWG would like to get more live
experience on DS-TE before deciding whether this should make Standards
Track or not. This seems to perfectly reflect the situation.
We've made the same decision for RDM, MAM and MAR: we've made them
Experimental to indicate that we needed more experience to decide which
one should become Standards.
I can't see that LOM is less Experimental than some of the BC models ^)

If we go that way, could you not live with LOM as Experimental?

Thanks

Francois