Re: The Core Beliefs of Usability and Their CSS Application

On 7/5/05, David Woolley <david@djwhome.demon.co.uk> wrote:
> 
> > > - the one technology problem - most authors want to think they are
> > >   writing a document in a single language, and are prepared to confuse
> > >   HTML, CSS and EcmaScript, but not server configuration directives;
> >
> > I don't quite follow what the point of that statement is. I'm not
> > trying to be difficult, but what are you getting at?
> 
> This is the same problem that means layering won't work.  A large
> proportion of web authors want to write HTML.  By HTML they mean the file
> that causes IE6 to produce the display they want, including scripting,
> styles, etc.  Currently server configuration, and therefore proper control
> of things like caching are not done because they cannot be done in the
> "HTML" file.

How things are cached and managed is unimportant to the existance of
such a feature to the web author. They need only know that it's turned
on.

> > I don't see where there is the need for configuration at all. It
> > should be a zero-configure system. I don't have to configure my web
> > server to respond to incoming requests. It should be part of the
> > underlying algorithm.
> 
> Servers aren't HTML servers, they are web resource servers.  In some
> cases it may be important that a specific format be served, and, like
> with other negotiation types, there may be a need to quality
> rank material.

Actually most that I've seen are in fact fairly complex. They often
handle server script (PHP, ASP, CGI), they handle page instructions
(e.g. no-cache), whether we wanted them to or not. They also handle
incoming requests for resources by rerouting or denying access. They
can, and do sometimes, do this too.

> > I'm not sure we're talking about the same thing anymore. I'm talking
> > about the fallback mechanism as it relates to alternative media types.
> > I'm a firm believer in alt text.
> 
> Images are most often used as an alternative media type for HTML
> text.  XHTML 2.0 treats the provision of alternative text as a
> fallback from the replacement image to the original text, or from
> a non-directly replaceable image to a text near equivalent.

I'm sure that's exactly what the web authors had in mind with
spacer.gif . The truth I think is that visual designers think in
pictures and when asked why all there <p> elements are empty except
for a src attribute, they're going to say, "that's how you do images
now." The irony will be bittersweet. I'll be right and won't want to
be.

Orion Adrian

Received on Tuesday, 5 July 2005 13:44:26 UTC