Re: Instantaneous metadata cues

I'm pretty sure that one should see an exit event for every enter event.

I am less convinced one should see enter/exit pairs for times that did not get 'sampled'.  For a start, you can drive a system crazy by putting 'too many cues too close'. I think that the model that the UA skips forward as needed to keep up makes sense.  You also use uniform logic when seeking, skipping ahead to keep up, and so on(as, in fact, time is only sampled at discrete instances anyway).


On Mar 2, 2012, at 8:12 , Glenn Maynard wrote:

> On Fri, Mar 2, 2012 at 9:46 AM, Victor Carbune <victor.carbune@gmail.com> wrote:
> This situation is currently specified in the spec through the concept
> of "missed cues", so enter & exit events are dispatched as soon as
> such a cue is detected through normal playback (even if the playback
> position never lied within the exact range of the cue).
> 
> This is mentioned in section 4.8.10.8 Playing the media resource [1]:
> specifically in the rules starting with "When the current playback
> position of a media element changes (e.g. due to playback or seeking)
> [...]":
> 
> The phrase "missed cue" doesn't appear in the spec.
> 
> But after searching, I see it's in http://www.whatwg.org/specs/web-apps/current-work/multipage/the-video-element.html.  It's irritating to waste time because we have two copies of a spec that aren't being kept in sync, never mind the consequences if people implement out of date specs.  Why are they out of sync?  It's bad enough to have TR's confusing things...
> 
> -- 
> Glenn Maynard
> 
> 

David Singer
Multimedia and Software Standards, Apple Inc.