W3C home > Mailing lists > Public > www-tag@w3.org > April 2008

Re: Uniform access to descriptions

From: Xiaoshu Wang <wangxiao@musc.edu>
Date: Wed, 09 Apr 2008 14:51:21 +0100
Message-ID: <47FCC9D9.6010007@musc.edu>
To: Pat Hayes <phayes@ihmc.us>
CC: Dan Brickley <danbri@danbri.org>, "Williams, Stuart (HP Labs, Bristol)" <skw@hp.com>, Jonathan Rees <jar@creativecommons.org>, "www-tag@w3.org WG" <www-tag@w3.org>, Phil Archer <parcher@icra.org>



Pat Hayes wrote:
> At 8:58 AM +0100 4/9/08, Dan Brickley wrote:
>> Hi Pat,
>>
>> Pat Hayes wrote:
>>> At 7:52 PM +0100 4/8/08, Xiaoshu Wang wrote:
>>>> Pat Hayes wrote:
>>>>> At 5:54 PM +0100 4/8/08, Xiaoshu Wang wrote:
>>>>>> Stuart,
>>>>>>> Wrt to that resolution... a 303 response means *nothing*... if 
>>>>>>> you happen to follow the redirection and find something useful 
>>>>>>> about the thing you originally inquired of, that you trust and 
>>>>>>> are prepared to stick in your reasoning engine, then you win - 
>>>>>>> if not, of itself, the redirection has told you nothing/means 
>>>>>>> nothing.
>>>>>>>
>>>>>>> 200 tells you that the response convey as representation of the 
>>>>>>> (state of?) referenced thing.
>>>>>>>  
>>>>>> If this is what TAG accepts, i..e, 200=*representation of* as 
>>>>>> oppose to "resource of".  I have no problem and would be happy 
>>>>>> with it.  My perception is that TAG is recommending either 
>>>>>> explicitly or implicitly the latter viewpoint.
>>>>>
>>>>> Gentlemen, please both of you speak very slowly and carefully at 
>>>>> this point, as a precise understanding here is critical.
>>>>>
>>>>> Stuart, did you mean that the response conveys/ a/ representation/ 
>>>>> in the webarch sense/ of the referenced thing? It would be helpful 
>>>>> if every time the word 'represent' and its cognates are used in 
>>>>> this very special sense, such usage were explicitly flagged, as it 
>>>>> can very quickly lead to incomprehension when understood more 
>>>>> broadly (as it is almost everywhere else in the English-speaking 
>>>>> world.)
>>>>> (Xiaoshu: from which it follows that in this case, the referenced 
>>>>> thing in question must be something that/ has/ a 
>>>>> webarch-representation; so, in this case, it/ cannot/ be some 
>>>>> other kind of thing that cannot, by virtue of its very nature, 
>>>>> have such a (webarch-)representation; so, to refer to such things 
>>>>> - such, as we now might say,/ non-information resource things/ - 
>>>>> requires something other than a 200 response. Thus goes the 
>>>>> http-range-14 logic, as I understand it. Note that in order to 
>>>>> follow this, all we need to know is that there are things which 
>>>>> (a) cannot have a representation in the webarch sense but (b) that 
>>>>> we might wish to refer to with a URI.
>> (aside: perhaps 'http(s) URI' was meant here, rather than just 'URI'?)
>>>>> Their exact nature need not be specified, but I believe that the 
>>>>> language of 'information resource' boils down to  an attempt to 
>>>>> characterize this category of [/things that cannot be 
>>>>> webarch-represented by a byte stream/]. And, centrally important, 
>>>>> not having a representation in the webarch sense does/ not/ mean 
>>>>> not having any kind of representation, being unrepresentable, or 
>>>>> not being describable. The webarch sense of 'representation' is 
>>>>> very specialized and narrow.)
>>>> Pat, as I have detailed argued here 
>>>> http://dfdf.inesc-id.pt/tr/web-arch.  There can have only one 
>>>> consistent interpretation, that is: there is no so-called 
>>>> "information resource".
>>>
>>> The key issue is not what is an information resource, but what 
>>> isn't. So, in your document you ask, what makes the claim "A person 
>>> is not an information resource" true? And it seems to me that this 
>>> at least has a clear answer: because a person is/ not/ something 
>>> whose essential characteristics can be conveyed in a message.
>> I don't know what 'essential characteristics' are. Really. What are 
>> the (erm...) characteristics of the 'essential characteristics' of 
>> some [named type of] thing? Who gets to decide?
>
> I'm reading 'essential characteristics' as meaning, roughly, what in 
> OntoClean are called 'rigid properties' and what are often called 
> 'essential properties', meaning properties or aspects of a thing which 
> it has necessarily, i.e. which if it didn't have those it would cease 
> to be what it is. Among my essential characteristics, for example, is 
> my being human; or if you prefer, mammalian. And although we have the 
> word "human" in English, its impossible to convey the/ property of 
> being human/ in a message.
But, what is the rigid property of being a document? How to convey "the 
property of being document"? If you answer is "document is what is 
digitizable".  Then, you have a subclass of document - digital document 
but not the document itself.  Then, there is a subclass of digital 
Human, which is digitizable too. Take gene as another example, Is gene 
an IR? It should be not according to what you want to define.  We can 
certainly very faithfully digitize gene's sequence, yes?

The issue in the web is not about defining those properties. It is about 
communication.  That is how we communicate our viewpoint about certain 
resource from its digitized subclass (representation).  The purpose of 
the web is not about how to digitize resource but to communicate 
resource through its digitized form.

Xiaoshu
Received on Wednesday, 9 April 2008 13:52:09 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:55 GMT