Re: Whence the HTML DTD [was] Re: More syntax details in HTML 2.0?

Albert Lunde (
Mon, 19 Jun 95 13:35:16 EDT

At 4:29 PM 6/15/95, David Morris wrote:
>The archiving and accessiblity of DTDs associated with IETF
>standards such as HTML should be as rigorous as the handling of
>the standard itself.
>Could this be as simple as getting the IESG to accept RFCxxxx.DTD as
>the DTD associated with RFCxxxx ? Another approach would be for
>the IESG/IETF to 'certify' their WWW server as providing valid
>copies of standards related materials (this may be already true)
>and define a URL like:
>where the htmldtd cgi program would expect PUBLIC or SYSTEM or
>whatever for TTTTTT and the sgml dtd id string as the value of IDID....

I'd favor getting the RFC editor to accept a file type of .dtd for DTD's
associated with RFCs and internet drafts or (more radically) to create
a new directory in the rfc/internet-draft mirror sites. Getting something
in a form that will be carried by RFC mirror sites fits the distribution
into the IETF's existing framework for publishing things and provides
acccess by many protocols.

But associating a DTD with an RFC/Internet-draftname doesn't solve
the problem of mapping public identifiers. The second idea above
would do that, but it seems like a hack.

(It would be nice to have working URNs about now ;)

Would the mapping be something that belongs in the "assigned-numbers"
RFC? (Not that this can be read automatically either.)

What does the SGML community do about issues think this?

    Albert Lunde