Re: [tcpm] I-D Action: draft-ietf-tcpm-1323bis-07.txt

["Scheffenegger, Richard" <[e-mail]>]

Hi,

> -----Original Message-----
> From: Pasi Sarolahti
> 
> 
> So, I'd like to second David's request: additional input would be much
> appreciated, based on the options below.
> 
>> So it'd be really good to get some other comments from a larger pool
>> of people on where this should go so we can resolve this:
>>
>> 1) No late enablement of TSopt: once it's negotiated, it's on for
>>   the rest of the connection.


Note, that text in 1323bis-08 now aligns with option 1) - explicitly disallow late TSopt enablement.


>> 2) Allow for late enablement of TSopt, i.e., Richard's text.
>>
>> 3) Allow late enablement of TSopt, and we hash out all the explicit
>>   details of every corner case we can think of.

As Joe pointed out, late enablement with a 10 (12) -byte option in the SYN is kind of pointless. Also, such a mechanism warrants a full draft in it's own right (and there are three authors working on that, see 
http://tools.ietf.org/html/draft-trammell-tcpm-timestamp-interval-00
and 
http://tools.ietf.org/html/draft-scheffenegger-tcpm-timestamp-negotiation-05 

which really is mostly a problem statement / requirements draft, but leaving out the late negotiation part in the published draft.



I hope that going with option 1 meets more the consensus of the WG.

Regards,
  Richard