XML vs. RDF vs. N-Triples vs. N3 vs. CSV saga is up again (was: Re: Toss NTriples -- RDF Reification is all we need )

Patrick Stickler wrote:

> On 2002-06-04 3:11, "ext Michael Kifer" <kifer@cs.sunysb.edu> wrote:
>
> >>>>>> "SR" == "Seth Russell" <of Mon, 03 Jun 2002 10:35:26 PDT> writes:
> >
> >   MK> NTriples can be naturally encoded in XML and exchanged.
> >
> >   SR> Is that actually true?   How?
> >
> > <triple><subject ...>subj</subject><property>...</property> <object> ...
> > </object> </triple>
>
> Why of course. Why did we not see this before?!

Hello Patrick,

I have been following this forum for at least a couple of years now and I saw
several colorful threads about the XML vs. RDF vs. N-Triples vs. N3 vs. CSV
saga that now I can not even remember how many of them. I am pretty sure that
your XML syntax for triples has been proposed on this list at least other
four times in the past [1][2][3][4] (if not, those were slightly different).
Together with them, there have been several different proposal of XML
serialisations for RDF [5][6][7] and even the other way around [8]; then RDF
as plain ASCII [9] text was proposed together with others similar syntax
trying to have CSV (Comma Separated Value-s) aka N-Triples [10] versions of
all the possible permutations of the others :) Some syntax are nicely
human-readable while others just machine-understandable; there are syntax
that compress beautifully while other less; some are UTF-8 compliant while
others not. Some are stating statements while others are quoting them. Then
we have real and dark-triples (DT) and so on and so on......and we could
probably continue for another five years trying to enumerate the pros and
cons of each approach or syntax without ending....ever! Is this process going
to end at some moment?
On the technical side, we could see instead that most of the proposed syntax
have been more or less implemented by the RDF developers, and IMO that
actually contributed to the success of the SW story today.

From a pragmatic point of view, I agree with you Patrick (and others) that we
need common concepts and paradigms to tame the RDF beast, trying to
build some infrastructure to help information interoperability (and I do like
your simplified syntax :) ; but (but!) we have to take care of not loosing
control of the thing by having "no-sense" discussions or even worse
"re-invent" our ideas over and over from time to time.

Here are my questions:

Aren't we in a dangerous loop sometimes?? :)
How can people believe us?
How could we expect developers to adopt RDF and understand what the SW is if
we keep on proposing and changing our minds every certain random number of
months?
Why XML has been a success while RDF not yet? Implementations perhaps? Or
simple because XML has an easy to use and understand spec/model?

IMHO the RDF Core WG has been set up with some of these questions in mind,
and I think they are trying to fix and address various flaws and problems in
the original XML/RDF model and syntax, providing some canonical RDF form that
everybody can understand and use. Till today the WG has been actively and
incrementally proposing concrete solutions to concrete problems i.e. they
have been enabling developers to write software :-) On our side we should
probably try to be more constructive instead of "destructive" and contribute
more to the WG work; I personally find the www-rdf-* mail archives a quite
nice knowledge-base to write software :-)

Some weeks ago a friend of mine gave a presentation about "buzzwords:
candidates" in which he was trying to explain how the "acceptance" and the
success of a spec is somehow proportional to the size of the spec self  e.g.
SGML vs. HTML vs XML, X.400 vs. SMTP, Z.39.50 vs. RDF, X.500 vs. LDAP vs.
CNRP and so on..... I think he was quite right there and that's why I like
the RDF model simplicity and I believe it could potentially be a success. But
to make it a success we should probably also answer to the following
questions as well:

What happens in the past?
Can't we learn from past experience?
How is the complexity of the standardisation process related to
simplicity/smooth of learning of a technology?
How is the duration of a standardisation process related to the market-driven
technology evolution?
Is the wait-and-see and prototype before standardising paradigm ever worked
here?

I am sorry for the large number of questions being asked in this email, but
as RDF fun and developer I would really like to see it happening at some
moment in the near future :) I do not want to worry anymore about what will
happen tomorrow, whether a new syntax will be proposed by some clever mind or
instead we will have  finally got RDF M&S 1.1 from which we can start writing
the next layer on :-)

best regards

Alberto

[1] http://lists.w3.org/Archives/Public/www-rdf-interest/2000Sep/0217.html
[2] http://lists.w3.org/Archives/Public/www-rdf-interest/2000Sep/0131.html
[3] http://www.ilrt.bris.ac.uk/discovery/rdf-dev/rudolf/js-rdf/
[4] http://lists.w3.org/Archives/Public/www-rdf-interest/2000May/0009.html
[5] http://www.w3.org/TR/1999/REC-rdf-syntax-19990222/
[6] http://www.w3.org/TR/rdf-syntax-grammar/
[7] http://www-db.stanford.edu/~melnik/rdf/syntax.html
[8]
http://lists.w3.org/Archives/Public/www-rdf-interest/2000Sep/att-0129/01-xlink2rdf3.html

[9] http://www.w3.org/DesignIssues/Notation3.html
[10] http://www.w3.org/2001/sw/RDFCore/ntriples/

Received on Tuesday, 4 June 2002 07:17:22 UTC