> One suggestion...
> 
> On 8/3/2016 5:31 PM, Black, David wrote:
> >>>    To avoid unexpected connection resets, TEPs SHOULD support a normal
> >>> > >    mode of operation that does not make use of data in SYN-only
> >>> > >    segments.  Moreover, implementations SHOULD enable data in SYN-
> only
> >>> > >    segments only if explicitly requested by the application
> > Usual IETF practice is to require implementation of mechanisms when there
> > are circumstances that require them (to need but not have is bad).  On that
> > basis, that first "SHOULD" ought to be a "MUST".
> >
> > The second sentence would be better phrased as a "MUST" for the
> > default configuration - that would allow additional means of turning
> > on SYN-only data for an admin who knows what she's doing.  Separate
> > requirements about when SYN-only data "SHOULD" or "SHOULD NOT"
> > be used would be fine.
> 
> FWIW, my usual expectation is that SHOULD/SHOULD NOT always describes
> the requirements of an exception. I.e., if there's a reason to give
> wiggle room, it should be stated.

+1 - that's a good general principle, and I think some of the ADs are now
reviewing for that ...

Thanks, --David



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

Reply via email to