W3C home > Mailing lists > Public > public-html@w3.org > June 2009

Re: Firing media events early for throttled downloads

From: Boris Zbarsky <bzbarsky@MIT.EDU>
Date: Sat, 06 Jun 2009 14:36:44 -0400
Message-ID: <4A2AB73C.7050200@mit.edu>
To: John Foliot <jfoliot@stanford.edu>
CC: "'public-html'" <public-html@w3.org>
John Foliot wrote:
> * "author proposes, user disposes" (W4A 2006 Keynote Address)
> 	www.w4a.info/2006/presentations/09-Rhys-Lewis/09-Rhys-Lewis.html

If you want to do this, and remove the autoplay attribute, you also need 
to remove or limit scripted control over playing the video; otherwise 
the removal of autoplay is just a symbolic gesture.  I assume we agree 
on that, right?  At the same time, we do want to allow page-provided 
video controls, which means allowing the script _some_ way to play the 
video when the play button in said controls is clicked.

How would people feel about this proposal:

1)  There is an autoplay attribute.
2)  The page can call play() on the video.
3)  Whether play() works (and whether autoplay does anything at all) can
     be subjected to the same sort of controls as window.open() is
     now.

This would allow the UA to present the user with an adaptive experience, 
including whitelisting of sites where autoplay should be allowed, 
prompting when the page wants to play, etc.  As a personal example, I 
would want to whitelist a site like youtube for autoplay (if I'm there, 
I want to see the video).  I'd want to blacklist a site like 
money.yahoo.com for autoplay (because they have gratuitous videos that 
randomly start playing when you load the page, and are often below the 
fold).

The main drawbacks I see is that the window.open restrictions are 
heuristics that differ in different browsers.  But I think requiring 
that play() in an onclick handler work would not be unreasonable.  I 
don't believe any current browsers block window.open in onclick handlers.

Are there other drawbacks I'm missing?

-Boris
Received on Saturday, 6 June 2009 18:37:26 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 September 2014 09:39:04 UTC