Re: [heycam/webidl] Add [InNamespace] to allow defining an interface in a namespace (#425)

> My thinking for advising @littledan to go with this instead of new syntax is that we already use extended attributes to control what gets exposed for a given interface (viz. [NamedConstructor], [NoInterfaceObject], and [Exposed]).

If we could end-up in a situation where those extended attributes are mostly there for legacy purposes, I think that would be better. We're adding new syntax for mixins soon, so [NoInterfaceObject] will be a legacy construct.

> Plus, it's always a bonus when we don't need to update every IDL toolchain to add a new feature.

Well, for namespaces to be at all useful, we'd be breaking the invariant that identifiers are unique, so we're in for a world of pain anyway (e.g. idlharness assumes interface identifiers are unique).

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/heycam/webidl/pull/425#issuecomment-324845150