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

Re: Summary of the QName to URI Mapping Problem

From: Thomas B. Passin <tpassin@home.com>
Date: Wed, 22 Aug 2001 09:11:06 -0400
Message-ID: <000e01c12b0b$e7f65640$7cac1218@reston1.va.home.com>
To: <www-rdf-interest@w3.org>
<Patrick.Stickler@nokia.com>

>
> Again, I repeat, the issue is not one of certainty, correctness,
> or consistency of statements, but one of integrity and consistency
> of *data*.
>
OK, I'm going to close this out with this last remark.  You ***are*** going
to lose information putting something into rdf interchange syntax - it's the
well-known inability to round-trip.  Once you flatten a nested structure
into sets of triples, with interpolated nodes to take the place of the
nested structures, you lose the information needed to fully reconstruct the
original, right?  You may claim that nothing important is lost ***for the
purposes of RDF representation***, which may be true. Or you may add some
triples to describe the original structure, which would mean adding thngs
that weren't in the original.  In any event, something has been lost or
added.

But I really just meant to be a little provocative (and not completely
serious) by asking whether, if there were going to be some friction -
information loss - in the overall system anyway , it would matter too much
exactly where that friction occurred: "out there", or "in here" (in my own
processor). Naturally, I don't advocate that we should casually allow our
own processors to degrade data.

Cheers,

Tom P
Received on Wednesday, 22 August 2001 09:07:55 GMT

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