Re: Format of candidate-attribute string

On Jan 16, 2013, at 10:35 PM, Eric Rescorla <ekr@rtfm.com> wrote:

> I don't see a lot of value in making a candidate object given that
> we have decided not to break out other parameters for (e.g. SDP)
> [yes, yes, I know comment 22].
> 
> WRT to the right string representation, given that 5245 defines the
> production without the a= and the \r\n and that they really are
> part of SDP rather than ICE, my preference would be to do
> without.
> 
> -Ekr
> 
> 
> On Wed, Jan 16, 2013 at 3:33 PM, Martin Thomson <martin.thomson@gmail.com> wrote:
> On 14 January 2013 07:40, Cullen Jennings (fluffy) <fluffy@cisco.com> wrote:
> > when the ICE spec changes we need to track that
> 
> That sounds like a feature to me.  There's something to be said for a
> container that holds extensions that conform to the ICE grammar, but
> if the IETF changes the rules, then it's probably not something
> trivial and backward compatible.
> 
> 

WFM - seem to be the bast argument so far