I agree. My understanding is that the ordering of things within a header field value is only important if the ABNF and/or RFC explicitly indicate that they are.
A similar question was asked within the following thread. https://lists.cs.columbia.edu/pipermail/sip-implementors/2015-August/03028 0.html RFC 3455 provides an example where ordering is important. RFC 3455 defined P-Charging-Function-Addresses so that the order of ccf parameters and ecf parameters is important. Interoperability issues concerning ordering (or maybe duplicate parameter names) may have been the reason why RFC 7315 made ABNF changes which are not backward compatible with RFC 3455. > -----Original Message----- > From: sip-implementors-boun...@lists.cs.columbia.edu [mailto:sip- > implementors-boun...@lists.cs.columbia.edu] On Behalf Of Paul Kyzivat > Sent: Tuesday, October 06, 2015 12:16 AM > To: sip-implementors@lists.cs.columbia.edu > Subject: Re: [Sip-implementors] Ordering of Via field value parameters > > On 10/5/15 7:23 PM, Maxim Sobolev wrote: > > David, IMHO that "same ordering" clause refers to the "header values" (i.e. > > individual "via" lines), not to the order of parameters within ONE > > header value. Order of values is important, because it defines your return > path. > > Which is why the clause is there, I believe. Order of parameters on > > the other hand has no particular meaning. > > This is also my interpretation. > > I know of implementations that parse the parameters of a sip header field and > put them into a hash map. Then they can be serialized again, but the order in > which they were parsed will not be preserved. (This would be what would > happen when creating a response.) In my understanding this is a valid > implementation of the spec. > > Thanks, > Paul _______________________________________________ Sip-implementors mailing list Sip-implementors@lists.cs.columbia.edu https://lists.cs.columbia.edu/mailman/listinfo/sip-implementors
