From: mitra@path.net (Mitra)
Subject: Re: An alternative approach to URC/URT syntax
Message-Id: <CFIwK4.159@pandora.sf.ca.us>
Date: Tue, 26 Oct 1993 21:19:15 GMT
Olle,
Thanks for this suggestion.
There seem to be two concepts in here which are usefull no matter which
format we adopt for the overall syntax. Specifically the character
issue, and fragments. I think they should be discussed seperately from the
overall proposal. I'm glad to see a concrete proposal for handling foreign
language sets coming from someone who actually has to use them all the time!
The overall syntax has a lot of similarity to my revision of Mike's
proposal. The biggest difference is that in your proposal you use ";" to
seperate elements, whereas I use CR's with leading white space for
continuation lines. I think both are totally valid. Your proposal has the
advantage that it scales nicely.
I think the issue of nesting is crucial to how this works. i.e. how to
present a URN with some meta-information, and then a bunch of URL's each
with their own meta-information. Mike and you have chosen fundamentally
different ways of doing this - you've gone for bracketing with BEGIN and
END, whereas Mike has gone for precedence. Personally I'd prefer not
to have to present the "BEGIN" and "END" string to users.
One thing that seems clear from all these proposals is a common idea of
what we are trying to achieve, i.e. a way to present in text, a collection
or URN, URL's and Meta-Information. Can I appeal that at Houston we
chose (vote if neccessary) for a single acronym for this concept, however
we end up representing it.
- Mitra