RE: The devices-universal websites "myth" and the semantic web

> Yes, but could we have the content and navigation adjust
> automatically, by 
> combining the wishes of the content creator, needs of the user and 
> capabilities of the equipment?

> The importance of different parts of the content and the
> order in which it 
> should be displayed could be hinted at by the designer. The 
> user would 
> indicate their preferences and the UA then try to do the best 
> it could.

I guess even though this _is_ the original intention of many
device-independence approaches I'm not  sure if it can ever really work
in practice: From what we see on the Web today, most content creators
would not be satisfied by just giving "hints" on their layout and design
intentions. I do have the impression that those horrible
"best-viewed-with-internet-explorer-5.5-at-1024x768-in-full-screen"
sites are fortunately becoming less common, but web development is still
about precise visual control. It may not necessarily be about
pixel-perfect control, but at least it's about a solid predictable
result, and too much automatic adaptation might quickly make things
unpredictable.

Also, many possible patterns on how to make content scale across devices
are heavily dependent on the nature of the content: Downsizing a news
article by taking only the first paragraph and adding a "more..." link
is something that definitely works well - but only with a news article
or similar content. What about content thats more complex in structure?
How handle things like navigation sidebars, etc.? Again, I think most
designers will want to control precisely how their content should adapt.

> In practice it is the images (and sometimes Javascript) which
> take up the 
> bandwidth, not the text. So hiding some of the text wouldn't 
> mater much. A 
> clever CMS might even be able to omit the hidden connect if 
> it detects a 

I agree with that. And I definitely agree that a CMS could be just the
right spot to tackle the multi-device issue, in conjunction with a
device-profiling solution (e.g. the WURFL was mentioned in our
discussion)

Greets,
Rainer

Received on Thursday, 16 June 2005 16:04:24 UTC