- From: Philip Jägenstedt <philipj@opera.com>
- Date: Wed, 25 Jun 2008 16:51:27 +0700
- To: HTML WG <public-html@w3.org>
Hi! I'm assuming that the progress events in the HTML 5 spec is intended to eventually refer to http://dev.w3.org/2006/webapi/progress/Progress.html This progress events draft states that exactly one of the events "error", "abort" and "load" must be dispatched. This is easy in most cases, but what about infinite media resources (streaming) or media resources so large that the browser can never cache the entire resource at once (certainly not in memory, and not on disk on devices either). http://www.whatwg.org/specs/web-apps/current-work/multipage/video.html#networkstate LOADED: The entire media resource has been obtained and is available to the user agent locally. Network connectivity could be lost without affecting the media playback. The "load" event is dispatched when networkState equals LOADED If the LOADED networkState is reached only when the entire resource is cached locally and the load even dispatched only then, none of the "error", "abort" or "load" events will be dispatched. This doesn't seem like a big practical problem, but to be compliant with both specs one would have to cache the entire resource. The simple solution would be to request that the requirement in the progress events spec be removed, does anyone have any other ideas? Unless the definition of the LOADED state is changed, a note to point out that it may not be reachable is in order. Somewhat related, is the intention to have the "stalled" event added to the progress events spec? -- Philip Jägenstedt Opera Software
Received on Wednesday, 25 June 2008 09:52:16 UTC