At 10:05 AM 2/5/96 -0600, Scott E. Preece wrote:
>All of the things recommended here would be useful, to *some* users.
>Most end users, however, have no likely interest in whether the document
>they are looking at is valid HTML.
I think it is reasonable to require HTML Users Agents to tell their users if
they have "inferred" the structure of a document by guessing at the meaning
of incorrect markup. This is different, of course, from inferring the
structure of a document that was created in a different version of HTML.
You can tell the difference through the DOCTYPE. Documents without a
DOCTYPE are considered HTML 2.1.
A browser with "HTML-error standard" compliant error checking would have to
be able to display "Bad HTML", "Newer HTML", "Valid HTML" or "Unknown HTML
I have a couple of questions, though:
* Should the error checking be required every time a document is displayed,
or is it required as a feature that could be buried in menus or on a toolbar.
* What is "correct HTML?" Do we just check conformance against the DTD, or
also conformance with specifications within the text of the standards? What
about related standards (i.e. URIs, mime types, etc.)?
I think that the idea is good, but it sounds like it needs a lot more