From: weibel@oclc.org
Date: Fri, 30 Sep 1994 12:54:35 -0400
Message-Id: <199409301654.MAA02246@ws02-00.rsch.oclc.org>
To: uri@bunyip.com
Subject: Syntax for external representation
Dan's right... lets not formulate a new syntax when there are good
candidates available.
I endorse Dirk Hoymand's note about using SGML for the formal
expression of URC syntax. Yes, it's ugly, but the formal parsing tools
are available, its growing in popularity, it's already ingrained as the
Lingua Franca of electronic publishing.
Dick's reasons:
> 1) SGML is becoming increasingly popular is an electronic document format.
> It makes sense to create meta-data in this same format, thus a URC might be
> contained within a document. We just saw a proposal for one on this list
> today.
> I know that TEI has tags for meta-data (quite a few of them actually). So,
> this is coming, if not already here.
>
> 2) SGML authoring tools are hitting the market. This means we should see
> an easy means to do data entry, and have it be syntaxtically correct.
In fact, the SGML TEI tagging standard is done and ready for use, so there is
an enormous investment of intellectual effort that can be leveraged for
URC use right now.
Will TEI headers solve the entire problem? Certainly not, but it is a
great start. The fact that it is SGML encoded makes it easier to
exchange, validate, and process using the same tools that are used for
electronic documents.
My expectation is that URCs will fragment quite naturally into a common
tag subset with additional domain- or medium-specific tags. The formal
expression of a given URC type as a DTD will be very helpful for
exchanging such specifications, for building them into tools that we
will need to promote the development of URCs.
stu
Stuart Weibel
Senior Research Scientist
OCLC Office of Research
weibel@oclc.org
(614) 764-6081