W3C home > Mailing lists > Public > www-rdf-interest@w3.org > January 2001

Re: Dropping the redundant colon in N3

From: Sean B. Palmer <sean@mysterylights.com>
Date: Wed, 31 Jan 2001 23:54:16 -0000
Message-ID: <007701c08be1$1fb916a0$1ddc93c3@z5n9x1>
To: "Seth Russell" <seth@robustai.net>, "Aaron Swartz" <aswartz@swartzfam.com>
Cc: "Tim Berners-Lee" <timbl@w3.org>, <www-rdf-interest@w3.org>
I'll address Seth first, and then Aaron:-

From: Seth Russell <seth@robustai.net>

> Well the equivalence between a "namespace uri" and "proper namespace
> uri" could be stated in a language outside of N3.

But where would that be stated? Anyway, I think it is fully possible to
have a namespace ending with more or less anything; for example:-

     http://example.net/myns/ns1

Now if I had two properties in a piece of RDF with that as an alias (let's
say infomesh:p1, infomesh:p2), when you reconstruct them, they become:-

     http://example.net/myns/ns1p1
     http://example.net/myns/ns1p2

Now, it's fully possible that I could post something to those URI's is it
not? I don't think there are any hard and fast rules separating a namespace
from a "normal" URI, because namespaces are any URI: not a subset. Just
because it is fully impractical to use some URIs in RDF doesn't man that
that is an invalid syntax, because there are always going to be exceptions
to the rules, like I have shown above.

> I think we need to decide which is more important .. (1) that our
> different languages of the SemWeb (XML, N3, etc) be mixable
> or (2) whether we're going to be URI purists.

Well, obviously (1) is important to the SW, but we need to follow the rules
already laid down.

> If we are already in N3 we know if is a uri by default because it is
> not a punctuation or a literal.

But it could be a property, could it not? What about if I used mid:? That
would be even more confusing: is the mid:whatever a URI or a property with
the alias of "mid:"? We need some way to distinguish between the two.

From: Aaron Swartz <aswartz@swartzfam.com>

[...]
> "bob" "./foaf:wednesday" "mid:huh@w3.org" There's really no way to
> detect them with out some sort of special syntax.

Exactly.

> ban URI as a namespace prefix in N3. You could also figure out some
> other sort of delimiter for URIs that didn't conflict with XML.

I think banning namespace prefixes of any sort is a very bad idea as it
would conflict with the namespace specification. The obvious answer is to
prefix all URIs with a character that isn't allowed within URIs (or
namespace aliases...). I believe that percentage signs ("%") are only
allowed to escape characters in URIs, so why not use them?

     %http://example.net/ns/#

Note the following advice from the URI RFC:-

# E. Recommendations for Delimiting URI in Context
#
#   URI are often transmitted through formats that do not provide a clear
#   context for their interpretation.  For example, there are many
#   occasions when URI are included in plain text; examples include text
#   sent in electronic mail, USENET news messages, and, most importantly,
#   printed on paper.  In such cases, it is important to be able to
#   delimit the URI from the rest of the text, and in particular from
#   punctuation marks that might be mistaken for part of the URI.
#
#   In practice, URI are delimited in a variety of ways, but usually
#   within double-quotes "http://test.com/", angle brackets
#   <http://test.com/>, or just using whitespace
#
#                             http://test.com/
#
#   These wrappers do not form part of the URI.

- http://www.ietf.org/rfc/rfc2396.txt

--
Kindest Regards,
Sean B. Palmer
@prefix : <http://webns.net/roughterms/> .
[ :name "Sean B. Palmer" ] :hasHomepage <http://infomesh.net/sbp/> .
Received on Wednesday, 31 January 2001 18:55:30 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:47 GMT