Re: XHTML bugs

In practice, I find it easier to do without the XML declaration, which will show in older
browsers, anyway. I don't miss out on anything, validationwise, do I?

Einar

"Shane P. McCarron" wrote:
> 
> JakieChan wrote:
> >
> > Under 3.1.1 of the XHTML recomendation, it says you
> > are allowed (encouraged even) to include an XML
> > declaration. The validator doesn't agree.
> 
> I have successfully validated documents that use the xml declaration.
> 
> > The same example shows the !DOCUMENT entity with a
> > relative URI:
> >
> > <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0
> > Strict//EN" "DTD/xhtml1-strict.dtd">
> >
> > The validator doesn't like that either.
> >
> > It validates http://www.w3.org/TR/xhtml1/ so I guess
> > this must be a bug when using the upload file service?
> 
> the xhtml1 draft does not use an xml declaration because of a horrible
> bug in the Macintosh version 4.5 of Internet Explorer.  Newer documents
> the HTML working group has produced do use xml declarations, and they
> work fine with the validator.
> 
> As to the relative path problem, this is an issue that the working group
> has struggled with.  The SYSTEM identifier for an DOCTYPE should point
> to a DTD. However, if the PUBLIC portion is a well known FPI, the
> validator will use its local copy of the DTD rather than trying to
> follow the SYSTEM identifier.  So an uploaded or local document that
> uses a SYSTEM identifier of "foo" but uses a W3C-defined FPI should
> validate without error.  At least, that is how it works for me.

Received on Wednesday, 12 July 2000 17:16:32 UTC