Re: Should <video> buffer control be tri-state?

On Jan 8, 2010, at 3:09 AM, Kornel wrote:

>> Would it not be better to abandon autobuffer, and support a buffer
>> attribute? With values of buffer="yes" buffer="no" and buffer="auto",
>> which would be equivalent to how browsers handle no autobuffer now?
>
> I suggest calling it preload (preload="no", preload="metadata",  
> preload="all"?).
>
> "Buffer" may be confused with playback buffer, but this attribute is  
> only partially related to it (it doesn't affect buffering one once  
> playback has started).
>
> The term "preload" has been used for images for long time now, and  
> might give authors better idea what this attribute is for.
>

Since this discussion has settled down, I went ahead and filed a bug  
summarizing the conclusions. I thought Kornel's proposal was the best  
and got the most positive response, so I primarily highlighted this  
approach (with the additional suggested "playthrough" value and a few  
potential alternate names). But I also mentioned the other ideas.

Here is the bug: http://www.w3.org/Bugs/Public/show_bug.cgi?id=8731

I encourage Working Group members with an interest to Cc themselves,  
and to comment further in the bug if they feel it is appropriate.

Regards,
Maciej

Received on Wednesday, 13 January 2010 07:37:58 UTC