- From: Harald Alvestrand <harald@alvestrand.no>
- Date: Thu, 09 Jul 2015 12:09:45 +0200
- To: Dominique Hazael-Massieux <dom@w3.org>, public-webrtc@w3.org
Den 09. juli 2015 11:25, skrev Dominique Hazael-Massieux: > On 09/07/2015 10:47, Harald Alvestrand wrote: >>> * there is no definition of what it means for an ICE candidate to be >>> malformed (the RFC doesn't define such a term) >> >> I think you're right. Syntax errors should be caught when constructing >> the RTCIceCandidate from the string, not when pushing the >> RTCIceCandidate into the PC. >> >> But the error should be SyntaxError, and it should happen whenever the >> candidate doesn't conform to the IceCandidate ABNF in the RFC. > > Note that the ABNF is a superset of what a meaninful candidate can be; > in other words, there are plenty of room to write ABNF-compliant > candidate parameters that can't be used in the end. > >> Where is it appropriate to say that this is a requirement on the >> RTCIceCandidate constructor? (and is this indeed appropriate?) > > I'd be OK with failing on construction, but I'm not sure what this would > gain us? > >>> * no browser seems to be doing this at the moment >> >> It definitely fails on adding to the PC, not on construction. (Chrome 43) > > It does fail, but not with a SyntaxError (which is what I'm suggesting > we remove). So you're suggesting we remove the SyntaxError, but continue failing ... what are you suggesting we use as a failure instead? I'm much happier with proposing "change error code X to Y in case W" than with "remove SytaxError".
Received on Thursday, 9 July 2015 10:10:18 UTC