> AFAICT, the encryption spec is a component of TCP-ENO. This wouldn't be
> "overloading", but a sub-field or variant.
> 
> IMO, TCP option kind numbers ought to be assigned only for independently
> useful options.

Ok, I stand corrected on that one ... please ignore that comment on 4.2.

Thanks, --David

> -----Original Message-----
> From: Joe Touch [mailto:[email protected]]
> Sent: Wednesday, June 29, 2016 12:54 PM
> To: Black, David; Derek Fawcus; [email protected]
> Subject: Re: [tcpinc] TCP-ENO: David Black's review
> 
> 
> 
> On 6/29/2016 5:45 AM, Black, David wrote:
> > Hi Derek,
> >
> >> > Having encryption params within the scheme being renegotiated seems
> likely, but not
> >> > changing the scheme,  as such I can't see the need to preclude reuse of 
> >> > the
> option
> >> > values for per scheme use once the scheme is selected.  The actual use I 
> >> > was
> expecting
> >> > was that the per scheme renegotiation would/could use those values for
> negotiating
> >> > new keys, or other per-scheme values.
> > Hmm - looks like I may have misread this text last night as releasing the 
> > option
> for arbitrary
> > usage.  I suppose what is there is ok, but I'd prefer to see another option 
> > kind
> allocated for
> > encryption spec usage (if we anticipate that being needed) rather than
> overloading the
> > TCP-ENO option kind.
> >
> > Thanks, --David
> 
> AFAICT, the encryption spec is a component of TCP-ENO. This wouldn't be
> "overloading", but a sub-field or variant.
> 
> IMO, TCP option kind numbers ought to be assigned only for independently
> useful options.
> 
> Joe

_______________________________________________
Tcpinc mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/tcpinc

Reply via email to