Re: [MMUSIC] [Ice] Adam Roach's Discuss on draft-ietf-mmusic-trickle-ice-sip-14: (with DISCUSS and COMMENT)

[Christer Holmberg <[e-mail]>]

Hi,

>>>> Regarding core ICE, I still think we shall replace the RFC5245 
>>>> reference with a reference to draft-5245bis, for the following technical reasons:
>>>
>>> I would view this as a huge changes that would need some pretty 
>>> significant testing and discussion in the WG. Right now, as far as I can tell the browser have implemented ice not bis.
>> 
>> So, "the browser" supports e.g., both types of nomination?
>> 
>>> I would like WebRTC 1.0 to be what people have implemented - the 
>>> reason it is ice not bis is it is unclear if bis performs as wells as ice.
>> 
>> I assume our primary task is not to document what has been implemented, but to provide procedures on how to implement things.
>> 
>> As I said earlier, e.g., many of the trickle ICE procedures are based on the 5245bis procedures. If you want to implement 
>> trickle based on RFC 5245 there are details and procedures that you will have to figure out yourself, because they are not 
>> documented, which means that different implementations likely will do things in different ways.
>
> Although the trickle *technique* goes back to 2005 and thus predates the publication of RFC 5245 by ~5 years, formally speaking 
> you are correct about alignment of details between draft-ietf-ice-trickle and 5245bis.

Correct. The technique itself goes way back, but the standardized procedures (draft-ietf-ice-trickle) for implementing it are based on 5245bis.

Regards,

Christer