- From: Dan Connolly <connolly@w3.org>
- Date: Wed, 18 Jul 2001 13:21:43 -0500
- To: Aaron Swartz <me@aaronsw.com>
- CC: Graham Klyne <Graham.Klyne@Baltimore.com>, RDF core WG <w3c-rdfcore-wg@w3.org>
Aaron Swartz wrote: > > On Thursday, July 12, 2001, at 02:02 AM, Dan Connolly wrote: > > > from aaron: > > > > |I think all parsers should spit out > > |equivalent genids for the same document -- the spec should > > |mandate the genid to use. > > > > I'm pretty sure there is no algorithm for coming up > > with genids that works; i.e. there's no algorithm > > for sorting the triples in a graph: you can't treat a graph > > as if the edges didn't (sometimes) cross. > > This isn't necessary -- you only need to say something like > they're numbered based on the location of the XML element that > defined them, or that they random UUIDs or something. Random uuids won't meet the requirement I understand you're after, which is that different parsers would compute the same URI for anonymous nodes given the same RDF input. Random uuids are, effectively, skolem constants, which see other thread about whether reducing existentially quantified variables to skolem constants is OK or not. (my position: not). I've noodled on the approach of naming things based on their XML element position, and I've convinced myself that won't work either. I don't have a proof that there's no such algorithm, but I'm not likely to spend more time thinking about it until somebody makes a concrete suggestion (i.e. running code) of an algorithm that works. I'm pretty sure that, at a minimum, it requires the sort of tricks involved in putting a signature inside a signed document. -- Dan Connolly, W3C http://www.w3.org/People/Connolly/ office: tel:+1-913-491-0501 mobile: mailto:connolly+pager@w3.org?subject=pls%20call%20+1-...
Received on Wednesday, 18 July 2001 14:22:14 UTC