Hi,I've been implementing an ICE library that will be used for HIP NAT traversal and I found the NICE draft useful. I think it's good to have a document that explicitly tells you which parts of ICE are safe (and which not) to ignore if you are using ICE in other than SIP environment.
Jonathan Rosenberg wrote:
4. connectivity checks
[...]
What might differ is pacing;
IMO, this would be a really important part of the NICE document. The pacing formulas etc. in the ICE draft seem to be well thought out for RTP but I'm not that convinced of the non-RTP case. For example, minimum RTO of 500ms seems quite high to me. Some discussion and design guidelines or maybe even a few "timer profiles" that implementors could reference to could be useful.
Does this seem like: a. its useful to folks producing protocols that make use of ICE,
Yes.
b. its a good direction
Yes. Cheers, Ari _______________________________________________ mmusic mailing list mmusic at ietf.org https://www.ietf.org/mailman/listinfo/mmusic