Are IDL names like `Observer` too ambiguous?

[keithamus]

Looking through the spec here, I see Observer is an IDL type but that might be a little ambiguous, as there are lots of "Observer" like things that exist across various specs, for example ResizeObserver, MutationObserver and so on.

This has practical ramifications, as Webkit already has an Observer class, and Firefox also has an Observer class that are for different things. It would be useful to disambiguate against those, and it might be easier to simply rename this specifications Observer type.

Might I suggest a few alternative names:

• SubscriptionObserver
• SubscriberLike (my preference)
• ObservableSubscriber
• SubscribeCallbacks
• ObserverCallbacks

[domfarolino]

Sure, I'm not opposed to renaming this in the spec. Ideally (and likely) other engines' Web IDL implementations have some sort of custom syntax to allow generic-named Web IDL interfaces/dictionaries to take on unique-named implementation name, as to prevent a real conflict. Chromium has this, but yeah there's an argument even outside of that to rename this to be more specific.

Personally I'd like to stay away from using Subscriber in the name, since while there is a conceptual straight line you can draw between Observer and Subscriber, they are really not the same. Sticking with the precedent of ResizeObserver, MutationObserver, and so on, I'd prefer SubscriptionObserver. Do you want to make a PR?