On 02/10/2016 02:21 PM, Paul Kyzivat wrote:
Charitably, it may have been an attempt to apply Postel's Maxim, but
maybe in an inconsistent way at two different points in the code.
It seems to me to be a very wanting attempt to apply the Maxim, but this
is where my experience with the methodological guidance side of the
standards world falls short.
So, perhaps it's a bit pedestrian a point of view, but, as I see it:
- If one is to be strict in what one accepts, one must be consistently
so, not selectively.
- If critical validation on RR URI parameters is performed and a
parameter is found to be invalid, the response should be explicit, e.g.
a rejection with of 400 Bad Request.
UA B did not do this.
- But if validation is implicit, an RR header perceived to be malformed
should be silently discarded, not consumed and integrated into the route
set for the dialog.
UA B sent the Route header (constructed from the RR) in its BYE to the
calling party. If the RR header was malformed, why was it ingested and
utilised?
All this to say: if the stack is to be strict, be explicitly restrict,
but if not, then be consistently strict in every aspect of the same
fundamental issue, not merely some aspects, and if one is not going to
do be consistently strict, then do not behave in deleterious or
unwholesome ways.
-- Alex
--
Alex Balashov | Principal | Evariste Systems LLC
303 Perimeter Center North, Suite 300
Atlanta, GA 30346
United States
Tel: +1-800-250-5920 (toll-free) / +1-678-954-0671 (direct)
Web: http://www.evaristesys.com/, http://www.csrpswitch.com/
_______________________________________________
Sip-implementors mailing list
Sip-implementors@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/sip-implementors
