buglady@bronze.lcs.mit.edu (Aliza R. Panitz) wrote:
> Quoting from Thomas Koenig (Tue Oct 11 19:06:43 1994):
> > Henk Penning (henkp@cs.ruu.nl) wrote in list.faq-maintainers,
> ...
> > > the author puts an Url:-subheader-line in the faq pointing to
> ...
> > interested ;-). At the moment, I've got a "X-Url:" header in, but
> > I'm far from dogmatic on what the header should look like ;-)
> ...
>
> I vote yes to "X-Url:"
I second that motion.
X-Url (as opposed to "Url:") doesn't mess with the header namespace in
the son-of-RFC1036 draft. Since the [X-]Url header is mostly for the
benefit of periodic postings, we sould not impose a general header
name to the News RFC but use an X- header instead.
I would further suggest that we discuss the header format together
with the maintainer[s] of the WWW faq archive[s]. I only know of
Thomas Fine <fine@cis.ohio-state.edu>, who does a great job getting
the FAQs from news.answers (and in future maybe mirroring rtfm.mit.edu)
and converting them to HTML. (Hi, Thomas :-)
His software is great at converting digest format (RFC 1153) documents
to hypertext and in discovering URLs in any document, but only a few
FAQs are actually fully RFC 1153 conformant. My FAQ for example isn't,
because I don't want english keywords in a German language document.
So I would like to do my own HTML conversion and include it in the
X-Url header.
My idea of the X-Url header is that it is used to compile a list of
FAQs (like http://www.cis.ohio-state.edu/hypertext/faq/usenet/)
By default, the Subject: and maybe the Archive-name: headers are
evaluated, maybe including the From: line to produce the link to the
HTML version of the FAQ.
Again by default (if there is no X-Url: header), this will be the
HTML file automagically created on the server from the news.answers
posting.
If there is an X-Url: header, the server software may choose to
still use the Subject: or Archive-name: header as title for the
anchor, but substitute the _first_ URL in the X-Url: header as
link.
Alternatively, the title (maybe complete with the URL to the HTML
version on the server) may be kept, and the _complete_ X-Url header
included below (possibly between <HR>s to distinguish it from the rest
of the list).
The X-Url header format should support all these options.
Here's my suggestion for the X-Url: header format:
1. The header should be a HTML formatted text.
If the complete header content, starting just after the space in
"X-Url: " and ending before the next blank line (end of RFC1036 header)
or the next line with leading non-whitespace (next RFC1036 header line)
is included in the body of a HTML document, the result must still be
a valid HTML document.
2. The header must be in 7-bit ASCII, with special characters encoded
in HTML (not MIME) notation. This includes the characters '<', '>'
and '&'.
3. The following HTML language elements are not allowed in X-Url:
<HTML>, <HEAD>, <TITLE>, <BODY>, <Hn>, <ISINDEX>, <FORM> (and
everything related to forms). This includes also the corresponding
</HTML> etc.
Software interpreting an X-Url header that includes any of the
above may choose to discard that header altogether.
4. Use of the following HTML language elements is discouraged:
<PRE>, <BLOCKQUOTE>, <ADDRESS>, <UL>, <OL>, <MENU>, <DIR>, <LI>,
<DL>, <DT>, <DD>, <HR>, <DFN>, <CITE>, <CODE>, <B>, <I>, <TT>,
<!-- .... -->
These formatting statements may interfere with the document style
the interpreting software wants to produce. The software is free
to ignore any of these.
5. Every inlined IMG image must have an ALT= element.
The interpreting software is free to disregart the image altogether
and substitute the ALT text.
--With best regards,
Bernhard Muenzer
GSF - Forschungszentrum fuer Umwelt und Gesundheit GmbH PUC Neuherberg Phone: +49-89-3187-2425 85764 Oberschleissheim FAX: +49-89-3187-3369 Germany e-mail: mue@gsf.de
[
Usenet Hypertext FAQ Archive |
Search Mail Archive |
Authors |
Usenet
]
[
1993 |
1994 |
1995 |
1996 |
1997
]
© Copyright The Landfield Group, 1997
All rights reserved