Re: Media types and assertions

Hi,

On Thu, Mar 11, 2004 at 03:04:42PM -0600, Pat Hayes wrote:
> >If somebody on the Web can't distinguish between an RDF message which
> >says "Mark hates bananas" versus one that says "Mark hates bananas (but
> >not really)" (aka unasserted),
> 
> But that isn't the right way to express the distinction. "Mark hates 
> bananas" means what it means in either case, and its this bare 
> propositional meaning that the specs talk about and which is 
> identified by the media type (I actually think that media type isn't 
> the right tool even for this, but I know I'm in a minority on this 
> issue and so don't press the point). Not asserting something isn't 
> asserting it with a qualification: it is just, well, not asserting it.

Sorry, I should have been clearer.  What I meant by "but not really"
was "not asserted".

> >then there is a failure to communicate.
> 
> Communicate about what? The content is communicated accurately.

As it would be if it were transferred with text/plain or
application/octet-stream.

The Web is concerned with more than content, it's concerned with being
able to serialize up meaning at one end of the pipe, and have it
accurately reconstituted at the other end.  The media type tells us what
spec to use to extract meaning from the document encapsulated in a Web
message, and so is arguably the most important piece of metadata used in
this process.

> The 
> intent of the publisher is not communicated: but where is that 
> communicated anywhere on the Web?  Publishers of commercial websites,
> Blogs and webcams may all have very different intentions and goals, 
> but they all publish HTML, which is all that the web protocols 
> concern themselves with.

And they all publish their HTML using text/html.  If they published it
with text/plain, then they're publishing different information.

Mark.
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca

Received on Thursday, 11 March 2004 19:42:00 UTC