- From: Sandro Hawke <sandro@w3.org>
- Date: Sun, 13 Jun 2010 21:31:30 -0400
- To: Axel Polleres <axel.polleres@deri.org>
- Cc: public-rif-wg <public-rif-wg@w3.org>
[replying to myself replying to Axel]
> > * note: round tripping including RDF(S) inference is a noble goal but in general simply
> > not feasible, particularly in RIF-RDF(S)-combinations.
> >
> > Imagine a RIF ruleset R and a graph G consisting of the single triple:
> >
> > rif:allTrue rdfs:subProperty rif:anyTrue
> >
> > This almost certainly causes troubles, and I see no way to avoid these troubles.
> > So, I would weaken the claim/requirement for round-tripping even with inference.
>
> Hmmm. I wonder how to spell this out properly. Maybe I just need to
> write the demo code to clarify what I'm talking about, ... but I don't
> have time to do that on our current schedule. And this is a working
> draft. Could there just be an editor's note just talking about the
> issue?
Intuitively, what's wrong with the example you provide is that triple is
wrong. There's no way to make a ruleset survive malicious or incorrect
processing; I just want it to survive legitimate, sensible, sound (but
perhaps incomplete) inference, with the result either being clearly
incomplete or being actually intact. As a programmer, I'm confident
this mapping does that, but obviously I still haven't figured out how to
state that properly.
-- Sandro
Received on Monday, 14 June 2010 01:31:39 UTC