On 2/3/16 14:43, Daniel Burnett wrote: > I'm probably reading this thread wrong, but it looks as if Martin says > the SDP error info is unnecessary, while Adam says Firefox does it > (and then some). This does not appear to me as a consensus either > way. Can either you (Martin) or Adam help me understand which is the > direction that would be best to go -- towards including line numbers > (or similar info) or away from it? Martin and I are saying the same thing. Headline: we don't need a special RTCSdpError object. The spec says to use the same error objects for SDP errors as we do for other errors, and that works fine. Unimportant but maybe interesting detail that should have no real impact on the specification one way or another: Firefox squirrels away diagnostic information in this error object to help in debugging. It does this for all errors, not just SDP errors. In the case of SDP errors, this diagnostic information is usefully clear about where the error occurred and what caused it. -- Adam Roach Principal Platform Engineer Office of the CTO
Received on Wednesday, 3 February 2016 20:55:16 UTC