W3C home > Mailing lists > Public > public-sweo-ig@w3.org > April 2007

RE: Comments to Semantic Web FAQ [1/2]

From: Uldis Bojars <uldis.bojars@deri.org>
Date: Tue, 17 Apr 2007 01:28:59 +0100
To: "'Sandro Hawke'" <sandro@w3.org>
Cc: <public-sweo-ig@w3.org>
Message-ID: <E1HdbZB-0001f7-9m@maggie.w3.org>

Thanks to Sandro for comments.

>> > 4.1. What is RDF?

>> -> "for describing
>> data on the Web".
>> "data interchange" 

> I disagree.   RDF is for exchanging data, not just metadata.

> That should probably be a FAQ.   (I thought there was at least something
> about it on the mythbusting page, but I can't find it now.)   

Absolutely - RDF is about data. 100% agree.
Note: it does not say "for describing data about data", no suggestion about
metadata at all.

What I want to do with this change is to differentiate RDF from XML. If we
use the same explanation as XML has been using for a long time now - "data
interchange on the Web" - then it makes one think: "wait, isn't this what
XML is for already?". While "data interchange" is technically correct it may
help if we find a different wording, different from XML.

That's why I suggest "for describing data" [or "describing knowledge /
information on the Web"]. 
It means more than just data interchange to me as RDF is used to describe
the world, things in it (resources) and relations between them.
You may find a better wording for what I wanted to express here.

Note: there is another problem with the formulation "for data interchange on
the Web". RDF can be used to describe and interchange data in corporate
environment which technically is not "the Web". 

>> > 4.3. Where is the "Web" in the Semantic Web?

>> "URI-s are used to to name resources in RDF triples" -> "URI-s are 
>> used to name and uniquely identify resources in RDF triples"

> While I think the term "uniquely identify" is technically correct, it's a
bit of a problem because it suggests to some people that each resource has
only one name (a unique identifier), which is not the case.

"Uniquely identified" helps to better understand how statements merge
together and form graphs. 
We will need to introduce it somewhere and I tried to do that carefully.

Note that the saying "named" can lead to the very same understanding in some
people - that there is only one name.

In fact, technically they will be correct. Every resource in RDF has only
one URI. If two resources have the same URI they are the same, and if they
have different URIs they are different resources. (Unless I am mistaken).
This concept is important enough to use "uniquely identified" in the FAQ
(without going into details described here). 

It is only when we get to ontology languages like OWL that we can say that
two different URIs are actually the same.

>   - in 4.3 Where is the "Web" in Semantic Web?

>    Here's a rewrite in that direction.  Maybe it goes too far, but
>    hopefully it makes my point clear:

It's a nice answer which explains URIs in enough detail. 
No need for "uniquely identified" here any more but it may make sense to
mention it at least once elsewhere in this FAQ (when talking about naming

The sentence talking about globally-shared meaning for URIs may be too much
detail for this question though. 


[ http://captsolo.net/info/ ]
Received on Tuesday, 17 April 2007 00:29:06 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:28:53 UTC