Re: Fwd: [Bug 23367] Move exceptions into IDL

On 10/03/2014 04:44 PM, Jan-Ivar Bruaroey wrote:
> On 10/3/14, 5:13 AM, Harald Alvestrand wrote:
>> Based on this stuff finally happening, I propose that we change our
>> definition:
>>
>> [NoInterfaceObject]
>> interface MediaStreamError {
>>     readonly    attribute DOMString  name <http://w3c.github.io/mediacapture-main/getusermedia.html#widl-MediaStreamError-name>;
>>     readonly    attribute DOMString? message <http://w3c.github.io/mediacapture-main/getusermedia.html#widl-MediaStreamError-message>;
>>     readonly    attribute DOMString? constraintName <http://w3c.github.io/mediacapture-main/getusermedia.html#widl-MediaStreamError-constraintName>;
>> };
>>
>> to read
>>
>> [NoInterfaceObject]
>> interface MediaStreamError : Error {
>>     readonly    attribute DOMString? constraintName <http://w3c.github.io/mediacapture-main/getusermedia.html#widl-MediaStreamError-constraintName>;
>> };
>
> Looks great!
>
> One question, we define three of our errors (NotSupportedError,
> NotFoundError and AbortError) as "same as DOM4" [1], but they're just
> namesakes. i.e. they're of type MediaStreamError, not DOMException
> they way the prose is written. Just doublechecking that that's what we
> want, or was the intent to return errors more identical to DOM (i.e.
> create errors of type DOMException for those)? - FWIW I think I prefer
> MediaStreamError, just unsure curious what the redirect is supposed to
> buy us?

The only difference with using DOMException (that I can see) is that we
get a Code field along for the ride. I don't see any value in that;
others might.

But as Anne's just proved again, we're caught in a twisty little maze of
rules, none alike, so there might be subtleties that I have not understood.


>
> [1]
> http://w3c.github.io/mediacapture-main/archives/20140909/getusermedia.html#error-names
>
> .: Jan-Ivar :.
>


-- 
Surveillance is pervasive. Go Dark.