Re: [xml-dev] Is the set of languages expressible using XML asuperset of

[Hans-Juergen Rennau <hrennau@yahoo.de>]

I would have thought that triples is about resources and properties, and that these are more fundamental things than a binary relational data model. (Thought experiment: could you explain to a person living 1000 years ago what you mean by resources and properties? I think - yes.)

Am Freitag, 28. Januar 2022, 10:05:42 MEZ hat Michael Kay <mike@saxonica.com> Folgendes geschrieben:

Indeed, triples are no more semantic than anything else. Triples are essentially an expression of a binary relational data model, and the advantages of that model are about normalization and universality, they have nothing to do with semantics. As with any other notation, the semantics are conveyed by attaching out-of-band meaning to strings, in this case the URIs used to represent relationship names.

Michael Kay

Saxonica

On 28 Jan 2022, at 07:49, Alexander Johannesen <alexander.johannesen@gmail.com> wrote:

Hans-Juergen Rennau wrote:

Goal: express document semantics

Proposal: map document to triples

Alternatives: [please list]

Plenty of "document semantic" formats out there, why do you need triples for that? What problem is triples solving that others can't?

I know what you're after, btw, and those pesky triples (or are they?! What exactly are triples over anonymous nodes and massive repetitive tree structures, really? Is every SI truly singular in meaning? Where does the triple, tuple, n-ary expression start and stop?) provides a way to express complex tree structures to some satisfying definition of ontologically magnificence, but ... even Topic Maps can do that (while also being RDF compatible, both ways). Semantic expression and interpretation is quite different from serialisation, be it in triples and RDF, or Topic Maps and XTM, or leaves in the bottom of a tea cup.

All I'm saying is that your mode of transport isn't so much the problem, but more of a distraction from epistemological problems that lies underneath even the most brave triples parsers out there ...

Cheers,

Alex