Re: A Rough Guide to Notation3

From: "Sean B. Palmer" <sean@mysterylights.com>
Subject: Re: A Rough Guide to Notation3
Date: Fri, 23 Aug 2002 17:45:45 +0100

> > I have a serious problem with a document on N3 that starts
> > out saying that N3 is a ``shorthand non-XML serialization of
> > RDF''. I view this statement as wrong and, worse, completely
> > misleading.
> 
> Huh? It clearly *is* a shorthand non-XML serialization of RDF, and much
> more besides. 

If you don't mention ``and much more besides'' in this context, you are
giving the false impression that RDF can capture all of N3, which is false.

> The fact that I don't *mention* the extra baggage that comes
> along in the N3 suitcase does not mean that the statement in the document
> is false. If P & Q, and I only state P and say nothing about Q, how is P
> suddenly false? If I'd have said that N3 is *only* an alternate
> serialization of XML/RDF, perhaps you'd have a case.
> 
> In fact, had you spotted it, you would have been correct to take me up on
> the grounds that N3 is not really a superset of XML/RDF; it doesn't yet
> have the xml"" literal constructs. But that is a minor point, and one that
> seems likely to change in the near future (read the section entitled "Typed
> Literals").
> 
> However, seeing as how this document is intended to be helpful to people
> wanting to learn more about N3, it may well behove me to re-write the
> introduction.
> 
> > The document goes on to state that N3 ``forms a good
> > introduction into [sic]
> 
> Me and my nonce idiomatizations :-)
> 
> > [...] if you count among the key principles of the Semantic Web
> > a lack of both syntax and semantics, then I'll go along with this
> > statement, but I would hope that the Semantic Web would not
> > continue to espouse these principles.
> 
> Format fight! Everybody start throwing BNF! :-)

I don't care about format, just about any format would do.  However, N3
doesn't have *any* definition for either its syntax or its semantics.

> Notation3's shortcomings speak for themselves. It isn't stable, there is no
> central specification for it, and for a long time there were huge i18n
> issues. However, it is possible to round trip (modulo xml"" literals) from
> XML/RDF to Notation3 back to XML/RDF again. 

This *may* provide syntax or semantics for *part* of N3, but certainly
does't provide anything that is complete.

> For all these features, the RDF
> Model Theory applies. For all the extensions, I agree that
> specification-standard verbiage from TimBL would be extremely helpful, but
> it's not going to happen. If you use Notation3, you have to put up with
> that.

Then stating that N3 exemplifies the Semantic Web is making a very strong
statement. 

> As for the syntax, there are many irregularities between the various N3
> grammars. I have done a detailed survey on this very point, and have noted
> it many times. However, the consensus is strong enough for Notation3 to
> have utility in a number of applications. No one is forcing anyone else to
> use it, and you can simply ignore it if you choose. But there are an ever
> increasing number of people that find Notation3 to be useful to them in
> some way.

Sure, you can *use* N3, and, maybe, all current N3 parsers may accept what
you write, and all current N3 reasoners treat it the way you expect.  Is it
reasonable to count on this?

> The /2002/notation3/ document was only intended to be a list of helpful
> notes to people wanting to learn Notation3; note that I called it a "rough
> guide" rather than a "fully polished guide". It was not intended to start
> yet-another-serialization-war, and I suggest that anybody who wants to
> journey into that rathole again ought to take it to www-archive (the email
> equivalent of /dev/null, except it's archived).

Well, if you want it to be a helpful set of notes, then I suggest not
putting important statements into it that can be miscontrued.

> Kindest Regards,
> Sean B. Palmer
> @prefix : <http://purl.org/net/swn#> .
> :Sean :homepage <http://purl.org/net/sbp/> .

peter

Received on Friday, 23 August 2002 13:01:20 UTC