Re: Public feedback on HTML5 video

Tab Atkins Jr. wrote:
> The autobuffer attribute should never be anything more than a
> suggestion from the author.  The presence or lack of it should be
> ignorable by the browsers whenever they find it appropriate.

The problem as I see it is that because autobuffer is only a boolean 
attribute, it's currently only possible for the author to suggest the 
video should be autobuffered, but no clear way for the author to suggest 
that it shouldn't.  i.e. When the attribute is present, the author 
thinks autobuffering would be the optimal behaviour for users that can 
handle it.  But when it is absent, the author has not provided any 
indication either way and left it entirely up to the browser, perhaps 
based on the user's own preferences or connection speed, or whatever.

e.g. A browser being used on a dial up connection or slow public wifi 
connection could detect that and opt not to autobuffer, whereas a user 
on a high speed connection may opt to autobuffer whenever possible, 
regardless.

But the spec does not address the issue raised in John's post, whereby 
the author may wish to help conserve server bandwidth by suggesting that 
the video only be downloaded by those users who choose to watch it.

John gave an example of a blog home page containing many articles, one 
of which may contain a video.  But the reader may only be interested in 
one of the articles which doesn't, so it doesn't make sense to waste the 
bandwidth of either the server or user, and providing a way for an 
author to indicate this would be useful.

One way to address this would be to allow autobuffer to accept the 
values "on" and "off", just like the autocomplete attribute.  Of course, 
the browser should be free to ignore the attribute either way.

-- 
Lachlan Hunt - Opera Software
http://lachy.id.au/
http://www.opera.com/

Received on Thursday, 24 December 2009 11:48:10 UTC