W3C home > Mailing lists > Public > www-style@w3.org > August 2009

Re: Projection media type in fullscreen browser mode

From: Anton Prowse <prowse@moonhenge.net>
Date: Tue, 18 Aug 2009 20:03:57 +0200
Message-ID: <4A8AED0D.5040908@moonhenge.net>
CC: www-style@w3.org
Řyvind Stenhaug wrote:
> On Tue, 18 Aug 2009 16:11:28 +0200, Aryeh Gregor 
> <Simetrical+w3c@gmail.com> wrote:
> 
>> On Tue, Aug 18, 2009 at 5:17 AM, Řyvind Stenhaug<oyvinds@opera.com> 
>> wrote:
>>> To clarify, if the projection media type is used anywhere it will 
>>> render to that media type and ignore "screen". If there is nothing 
>>> projection-specific it will render to "screen" media.
>>
>> Then Opera is deciding its media type partly based on the markup
>> provided by the author?  That's not how I expect media types to work
>> -- they're meant to query what sort of device is being used, not
>> return information about the current page's markup.
> 
> I suppose it could be described as "falling back" to screen rendering if 
> no projection style is available. The standard does not seem to disallow 
> this, though I can see how it could be unexpected.
> 
>> I can't know if the administrator of the site has added a @media
>> projection rule somewhere in a custom stylesheet.
> (...)
>> So the only way for MediaWiki to get correct behavior is to make sure
>> that *all* built-in stylesheets are served as screen,projection, to
>> make sure Opera uses them in full-screen mode no matter what.  Even
>> though we don't actually want them to be used for projection, and the
>> content is actually totally unsuitable for paged projection.
> 
> Well, in this case we would be in projection mode, so it wouldn't really 
> be appropriate to add styles if they are not intended to be used for 
> projection. Either that, or the projection rules that were added in the 
> first place were wrong or insufficient.
> 
> However, it might be a good idea for the browser to allow the user a 
> choice between just "regular" full-screen and projection mode. I've 
> raised this to our desktop team as a possible improvement.
> 

To my mind, this exactly the kind of area that I /want/ to see browsers 
competing in, rather than have standards dictate how a (sensible) 
browser may or may not consume the resources they are processing.  A 
browser should be completely free to choose to render a page in 
"projection" mode -- or any other mode -- if it sees fit, /provided that 
it follows the rules for that mode/.  How it decides which mode is 
appropriate is a welcome competitive point over which users may elect to 
use one browser over another.  (IMO, Opera's heuristic is sensible, 
although speaking as a bit of a control freak I would like to have the 
option to choose which mode I wanted, whatever the browser!)

This reminds me of the horrible situation we currently have with screen 
readers.  Unlike the mutant applications on the market today which cause 
authors to jump through arbitrary and nonsensical hoops, these 
applications should have either been genuine aural browsers which read 
*all* the HTML and ignore all CSS apart from aural CSS, or instead 
should have been genuine screen readers which actually know what's 
currently visible on the user's screen and ignore anything which isn't. 
  Nothing wrong with claiming to be an aural browser -- on the basis of 
whatever heuristic it likes -- provided it actually *behaves* like one!

Cheers,
Anton Prowse
http://dev.moonhenge.net
Received on Tuesday, 18 August 2009 18:05:46 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:20 GMT