Re: Tests that depend on node being inserted for default attributes

Mary Brady wrote:
>Did anyone try the suggestion of creating an internal dtd instead of
external?
>Will this fix the problem without requiring validation?

I would consider those as a distinct new tests.  We do not have clear enough
guidance from the WG to know if default attributes from an internal subset
should be considered "known" and therefore should be provided.

We do know from the WG that they believed the existing tests over-reached
and tested behavior that was not always required.  However, they didn't
state the conditions under which the behavior was required.  If the behavior
was never required, then the tests should be deprecated.  By my changes, I'm
asserting my interpretation that the behavior is required when the parser is
validating since validation presupposed knowing the content of the DTD.

p.s. Marking the tests as requiring validation will only cause them to be
reported as Not Applicable to non-validating implementations.

Received on Friday, 8 March 2002 10:44:21 UTC