Re: REL and REV attributes (Was: More comments on HTML 3.0)

Michael J Hannah (mjhanna@sandia.gov)
Tue, 9 May 95 19:21:18 EDT

I have recently subscribed to this list, and have caught up on the
thread mentioned in the subject, thanks to the archive at my sister
Dept of Energy Lab a few miles up the road. I thought I would try to
add a fresh, outsider's, view of some of the issues in this thread. :-)

1) Need for both REV and REL.
Part of the thread discusses the need for *both* REL and REV.
I seem to have observed that there *is* a semantic difference.
REL seems to be the attribute to make a *statement* about a relationship
that this document is willing to accept/declare/etc. concerning the
referenced document/object.
REV, on the other hand, seems to be the attribute to make a *claim*
about a relationship to the reference document/object that can only
be verified by checking (somehow) with that document/object.
These are very different issues, and don't seem to have anything to
do with the argument about what list of relationships exist, or
whether that list implies some tree/hierarchy

For example, B can *claim* to have a "gworp" relationship to A by use
of the REV in B referencing A. If A is a HTML document, then A could
accept/verify that B has a "gworp" relationship to A by use of the
REL in A referencing B.

This has always been my understanding of the semantics of these
two attributes. Is this wrong? inappropriate? unneccessary?

It would seem that the purposes of REV and REL need to be clear first,
regardless of any effort concerning "Link types in HTML", or the
issue of registered relations lists. Perhaps this is the core of the
need for a clear link model as described by the messages in this
thread from Paul Burchard (burchard@horizon.math.utah.edu)?

2) Use of LINK by browsers.
As Dan Connolly (connolly@w3.org) said in this thread:

> There are pretty clear applications for this mechanism,
> but nobody has needed it badly enough to code it up. Or ...

I have never understood why browsers did not come with an "about"
feature concerning the currently displayed document. Excercising the
feature (clicking the "About" button?) would display the HEAD
information, including all the LINK information, listing the relation
with its reference as an armed hyperlink. Such a simple mechanism
would instantly make the HEAD information useful/visible and would
make the LINK references useful regardless of the discussion of REL/REV
and the entire link model. I have always thought that the purpose of
the HEAD element was to make this information optionally visable to the
viewer of the document. Isn't this the purpose both in Version 2.0 and
Version 3.0? If this group can make it clear that this is the purpose
of HEAD elements, maybe the browser vendors will at least provide this
simplistic presentation of this information, and then users will use it.

3) Registered list of REV/REL relations
I prefer the model of Headers in e-mail. A few are registered/fixed
with defined meanings and use. But others can be added at will using
the "X-gworp" syntax. Once an "X-" becomes clearly wide-spread and
useful, it can be standardized to remove the leading "X-"
That allows a group to propose an RFC for registered/fixed names, but
also allows expansion/flexibility.
__________________________________________________________________________

Michael J. Hannah "Opinions are mine only,
Scientific Computing Systems and will change without notice
Department 13918, Mail Stop 0806 whenever appropriate."

_/_/_/ _/ _/ _/ _/_/_/
_/ _/_/ _/ _/ SANDIA NATIONAL LABORATORIES _/_/
_/_/_/ _/ _/ _/ _/ Scientific Computing Systems _/_/_/_/_/_/
/ _/ _/_/ _/ Albuquerque, NM 87185-0806 _/ _/_/ _/
_/_/_/ _/ _/ _/_/_/_/ _/ _/_/ _/
_/_/_/
"Exceptional Service in the National Interest"

__________________________________________________________________________