Re: MIME types

Wendy A Chisholm wrote:

> One technique is, "Verify that the page is usable when programmatic objects
> are disabled."  This means that it must first identify programmatic
> objects.  We had assumed that we could identify which OBJECT elements
> included programmatic objects by looking at the "type" or "codetype"
> attribute.  However, there is not a type defined for Java or Flash in the
> current list of MIME types [2].

The current list or registrered media types covers a small fraction of what
is actually used.

The "type" is advisory and some people within IETF want to have it removed
altogether.

> I had expected to find a subcategory of
> application for Java and flash.

That is a reasonable expectation, and totally unlikely to ever be realised.

Many modern media types are sent out as text/plain or as
application/octet-stream. This reflects the Herculean efforts that were
required to register a new type in the past, plus the fact that most people
don't control the servers they use anymore and can't get any new types
added.

> Should this be taken to the HTML working group?  Have we overlooked something?
> 
> Perhaps we should say that OBJECTs with type="application/*" are potential
> programmatic objects?  We don't want to limit it only to Java and flash,
> but I expect a lot of what we'll be seeing is Java or flash and if we know
> that is what we have we could suggest specific techniques for those
> technologies.

Surely javascript should be added tothe list of programatic stuff?

Also, is this going to be a problem? *Is* it possible to use things with
programatic content disabled, in the general case? Or are people thinking
in narrow terms of current HTML pages in current browsers?

I don't want to see another well-meaning but impractical and detrimental
techniqe joining the infamous "usable without stylesheets"

--
Chris

Received on Thursday, 24 February 2000 06:14:43 UTC