Re: Meta Tag Draft - New version.

Simon Cox ([email protected])
Fri, 22 Dec 1995 10:07:38 +0800


I'm quite concerned about this discussion.

I'm afraid that I'm only medium competent in understanding the
arcana of HTML, HTTP, URL vs URI, etc etc, (so am liable to
attract the odd "newbie" flame) but I have a current job to do
which critically depends on some sort of resolution to these discussions.

I am responsible for coordinating the posting of results from all the
projects within the "Australian Geodynamics Cooperative Research Centre"
into a sensible framework on a few servers. Most of the people I'm dealing
with are much more familiar with publishing their results in conventional
journal/conference media, plus maps and databases, so these methods require
some re-education. Furthermore, I have very little real leverage with them,
so I have to make it easy as possible and rely on carrots rather than sticks.
I also am still trying to demonstrate the worth of a WWW approach to my
bosses, so have to do a lot of this stuff more-or-less on my own pending
the supply of further resources (people!).

After successfully demonstrating both some static html and also a
simple interactive application that generates customised maps
(http://www.ned.dem.csiro.au/AGCRC/4dgm/grasslinks/)
I now have them clamouring to post stuff (well, sort-of).

One important element of our web is that material needs to indexed
& cross-referenced by theme, location, time, etc. I haven't yet
figured out hoow this is to be done (eg use an existing robot? or what?)
but since material is being prepared to be posted NOW I am trying to
encourage the use of some minimal standards, in anticipation, so's
to speak.

I had thought
that the most logical place to put such information was in <META>
fields in the header, and have been working this week on an (internal)
draft of proposed mandatory and optional fields for this. Where there
are external standards, I've tried to retain consistency with them, but
in other places I've had to be creative.

(I have in mind to write a "forms" interface to allow the less ambitous
of my "clients" to type in their stuff, check a few boxes, and
automagically create a valid .html page from it. - BTW my "wish-list"
would include a couple of additions to FORMS, posted previously - eg
nested forms, thus giving a sort of "pull-down-menus").

SO, when David Musella's doc came through yesterday, I was pleased.
I've also found that META is widely used for this purpose, eg at Sandia,
and more locally at ERIN in Australia.

Now Matijn Koster is "deprecating" the use of <META> in this way,
and says we should use LINK instead. Is he the odd-one out on this,
or does he have a real point. Someone else is sniping at the Dublin-core,
which I have also been referring to (though I'm struggling to map their
elements onto any standard http-equiv type things - is there a
decent list of "standard" http-equiv names anywhere, which the
normal robots expect to find?).

Finally, can anyone explain _with_examples_
of how LINK is to be used? And also, when to use HTTP-EQUIV vs NAME
in a META tag.

I'm inclined to just carry on along the path that I've
started down, but am nervous if we're going to left in the cold.

Typed in real quick before an appointment - please excuse the sloppiness.

Simon Cox

_________________________________________________________
Dr Simon Cox __ \
CSIRO Exploration & Mining ,~' L_|\ Australian
39 Fairway, PO Box 437, ;-' \ Geodynamics
Nedlands, WA 6009 Australia ( \ Cooperative
Phone +61 9 284 8443 + ___ / Research
Fax +61 9 389 1906 L~~' "\__/ Centre
[email protected] W
AGCRC info>> http://www.ned.dem.csiro.au/AGCRC/
_________________________________________________________