On 2/3/16 11:23 AM, Sheldon Patry wrote:
hello,
By reviewing RFCs I came with that question:
RFC 3840, section 9 says:
The following production updates the one in RFC 3261
<https://tools.ietf.org/html/rfc3261> [1
<https://tools.ietf.org/html/rfc3840#ref-1>] for
contact-params:
contact-params = c-p-q / c-p-expires / feature-param
/ contact-extension
However, on
https://tools.ietf.org/html/rfc3261
I don't see 3840 in the list "Updated by"
Is there any reason?
Cases when "updated by" is and isn't used are subtle and sometimes
controversial. But in general, when a new document makes an extension
via an explicit extension point in the sip syntax, then it typically
isn't marked as an extension.
Also,
with only RFC 3261,
Contact: <sip:[local_ip]>;mobility=fixed
was valid.
However, with RFC 3840,
Contact: <sip:[local_ip]>;mobility=fixed
is invalid, we MUST have
Contact: <sip:[local_ip]>;mobility="fixed"
How is that possible that 3840 "broke" 3261?
With just 3261,
Contact: <sip:[local_ip]>;mobility=fixed
was *syntactically* valid, but the mobility parameter was undefined. So
you could put it in, but it would be ignored by anything that supports
3261 but not 3840. And note that:
Contact: <sip:[local_ip]>;mobility=****
was also syntactically valid.
RFC3840 provides a semantic definition for
Contact: <sip:[local_ip]>;mobility="fixed"
but not for the others. It narrows the allowed syntax for those that are
semantically defined.
This doesn't *break* 3261. You can parse it according to 3261 and ignore
it, or you can parse it according to 3840 and process it.
Thanks,
Paul
_______________________________________________
Sip-implementors mailing list
Sip-implementors@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/sip-implementors
