[whatwg] RWD Heaven: if browsers reported device capabilities in a request header

[James Graham]

On Mon 06 Feb 2012 05:00:55 PM CET, Boris Zbarsky wrote:
> On 2/6/12 10:52 AM, Matthew Wilcox wrote:
>> 1) client asks for spdy://website.com
>> 2) server responds with content and adds a "request bandwidth& device
>> screen size header"
>
> Again, the "screen size" is not invariant during the lifetime of a 
> page. We should not be encouraging people to think that it is....

No, but there is a different *typical* screen size/resolution for 
mobile/tablet/desktop/tv and it is common to deliver different content 
in each of these scenarios. Although people could load the same site on 
desktop and mobile set up to have the same viewport dimensions, it is 
not that probable and, only one of the two is likely to be resized.

A typical thing that people want to do is to deliver and display *less* 
content in small (measured in arcseconds) screen scenarios. If you are 
only going to show a subset of the full content it would be nice to 
only do a subset of the backend work (database queries + etc.) and 
transfer a subset of the full data. At the moment this is possible, but 
you pay for it with an extra RTT (at least as far as I can tell). I am 
sympathetic to the view that it would be desirable to be able to 
minimise the cost of generating a reduced-functionality page without 
burning the savings on extra round trips.