Re: addIceCandidate behavior

On 01/11/2013 1:37 AM, Harald Alvestrand wrote:
> Hm. Can we make a guarantee that an RTCIceCandidate always has a 
> non-malformed "candidate" attribute?
>
> What comes to mind is the case of RTCIceCandidate("Once upon a 
> time...") - that is, the string used to create it doesn't conform to 
> the candidate grammar.
>
> Perhaps the description (and the throw) is better applied to 
> RTCIceCandidate's constructor, so that an RTCIceCandidate is always 
> known to be syntactically valid?

     +1 That's a good idea.

Gili