- From: Dan Connolly <connolly@w3.org>
- Date: Sun, 05 Aug 2001 21:29:43 -0500
- To: Brian McBride <bwm@hplb.hpl.hp.com>
- CC: rdf core <w3c-rdfcore-wg@w3.org>
Brian McBride wrote: > > I was unaware of a length restriction on URI's. Is there one? There are limitations in various implementations that deal with URIs... e.g. most HTTP servers only accept a few kilobytes in the request header line in order to prevent denial-of-service attacks. But there's no limit on URIs per se. This was recently discussed in uri@w3.org... er... I thought it was, but I can't find it out. > Brian > > Devon Smith wrote: > > > > Let me first say that conceptually, i like the idea of literals being > > resources, mostly because i think one should be allowed to make > > assertions about literals. However, the data: scheme is an unacceptable > > solution to the problem of how to assign a URI to a string of > > characters. The length limit, recognized by the RFC, What RFC? > > is a legitimate > > concern for implementors. Another concern is how strings encoded > > in UTF-8, UTF-16 and other non-ascii, non-latin encodings would be > > dealt with. > > > > I can't think of an elegant way to make Literals be part of the Resource > > set. All I can see is a way to allow Literals to be treated as Resources > > when needed. One could use anonymous resources in conjunction with > > a property like "RDF:represents" to create a resource that represents > > a Literal. > > > > <rdf:Description rdf:about="genid"> > > <rdf:represents>Some insightful quote</rdf:represents> > > <a:attributedTo>Some insightful woman</a:attributedTo> > > </rdf:Description> -- Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Sunday, 5 August 2001 22:30:47 UTC