Re: Issue-57

On 2011-06 -24, at 22:24, Xiaoshu Wang wrote:

> From: Tim Berners-Lee <>
> Date: Fri, 24 Jun 2011 18:42:50 -0400
> To: Xiaoshu Wang <>
> Cc: Jonathan Rees <>, Alan Ruttenberg <>, David Booth <>, Jeni Tennison <>, " List" <>
> Subject: Re: Issue-57
>> On 2011-06 -24, at 16:22, Xiaoshu Wang wrote:
>>> From: Tim Berners-Lee <>
>>> Date: Fri, 24 Jun 2011 16:14:51 -0400
>>> To: Xiaoshu Wang <>
>>> Cc: Jonathan Rees <>, Alan Ruttenberg <>, David Booth <>, Jeni Tennison <>, " List" <>
>>> Subject: Re: Issue-57
>>>> On 2011-06 -24, at 15:00, Xiaoshu Wang wrote:
>>>>> [...]
>>>>> Go try the URI and see what the author says. If he didn't, use it whatever
>>>>> way you want. If the author is not careful of making such statement, it is
>>>>> his careless. If the author says one way, and you insist the other way,
>>>>> then it is your ignorance or arrogance.
>>>> Suppose the URI is
>>>> According to your proposal, what can I use it for in RDF?
>>> First, the presumption is that we will have an assertion of the sort like the follows?
>> No, in my case there is no such presumption.
>> There is just the web.
>> That is a real URI.
>> What can I use it for in RDDF, according to your proposal?
>> You say "Go to the URI and see what the author says".
>> What does that author say?
> I think you can use the URI in anyway you want because from a logic point of view, it contradict to the URI's owner's point of view. 
> I hold the same view as Larry's. That is: all semantic problem is a communication problem. When I make a statement saying:
> a ex:Person.
> I am merely expressing my opinion. Upon receiving the document, it is up to you to decide if it is consistent with your world so to accept or reject it.  In fact, even if the author has asserted that,
> a ex:Image. 
> I can still be able to make the earlier statement. It does not imply that I am *wrong* in an absolute sense, it simply suggests that I am offering a model that is inconsistent with the author's (assuming there is an accompanying assertion that Image and Person are disjoint).  But, that is how our knowledge evolves over time. I don't see why TAG should intervene.
>> Can I use that URI for statements about a picture?
>> Can I use that URI for statements about Barak Obama?
>> for example.
> Yes, either way is fine.  

In that case, your proposed system is not the WWW, and not the Semantic Web.
It is a different imaginary system, which you are free to develop, but you should not use the term "HTTP".

In the Semantic Web,  I can say  
	<ex:i>   <fb:like>    <>.

and it unambiguously means that I like the image, not the person.
This is very valuable.

Jonathan can say
	<> cc:licence ex:lic1.

and then those who know that can copy and paste the image with impunity,
and their automated editors can help them do it.  That is very valuable.

We can define a "parallel property" 
	<ex:i>   <ogp:like>    <>.

which implies statements about the subject of the picture,but that is done in the definition of ogp:like, not
in the way the URI's referent is defined.

The semantic web is a very useful system in which people can independently 
publish information about different things.  That information can be gathered and aggregated, and useful
deductions can be made from it. 

The TAG made that decision because that its its job: to define how the parts of
the spec fit together so that the whole system works.  It was asked to resolve that issue.
Interoperability for the web and for the semantic web is key.
Everyone has to agree on the basic way it works, so they can build interoperable programs.
Sometimes, almost always, the designs we agree on are not the ones we would have designed
ourselves.  But we gracefully accept that  we will work wit the standards because we value
the incredible power of interoperability.

My software would assume that <>
was a document before doing any HTTP request, just as it assumes that <> is a mailbox
before sending any mail.  However, others wanted to be able to use http: URIs like that for arbitrary things.
The TAG defined a compromise, so that you can only assume that it is a document if you get back 200.
I changed my code.  I would have preferred otherwise, but I went along with the compromise 
in order to get consensus and get interoperability.
Then, the community invented the 303 response, and I think soon we should invent a 209 response
as 303 is in general  too inefficient in some sorts of system.

You clearly have that urge to be creative about this area.  As we work in it,
we base our work on the existing designs, possibly help extend them in places, and
do not overturn the interoperability of a lot of existing data and software.


> Xiaoshu

Received on Saturday, 25 June 2011 07:18:41 UTC