Re: Some TAG review of "Cool URIs for the Semantic Web"

>Williams, Stuart (HP Labs, Bristol) scripsit:
>
>>  I mention this here because there seems to be a tendency in these
>>  threads to offer a definition whereby "information resources" are just
>>  those which in fact respond to an http request with a 200 response code
>
>In particular, that puts one in the silly position of saying that
>ftp://www.unicode.org/Public/README.TXT is not an information
>resource at all, whereas http://www.unicode.org/Public/README.TXT
>is an information resource.

Yes, of course this needs to be generalized over all transfer 
protocols. Ive been taking that for granted throughout.

>  > - whereas the TAG definition was intended to be broader - anything whose
>>  essential nature is information.

The trouble with philosophical-sounding definitions like this is that 
in order to use them one has to actually start doing some philosophy, 
which is usually a bad move in contexts like this. What exactly IS 
information in any case? What sense of "essential" do you have in 
mind? What exactly is the "nature" of something? These are much 
harder questions than anything the TAG is trying to answer. I really, 
honestly, have no idea what this could possibly mean. I wouldn't say 
that the *essential* nature of, say, an image was information. Surely 
its essential to the nature of an image that it be image-like.

Why do we NEED to define this concept of 'information resource'? 
These things are purely imaginary anyway. What actually happens is 
that something is sent back in response to the GET request (insert 
generalization for other xxTPs) with an indication that this is 
'correct' or 'normal'. We *call* this thing that is sent back a 
'representation' (using this word in a nonstandard way) of a 
'resource', but that tells us nothing at all about what this 
resource-thingie is, other than it is something that emits these 
'representations' of itself in response to GET requests: in fact, in 
the REST model a resource is *defined* simply as any function from 
times to representations. The one part of this story that is 
absolutely clear and can be defined quite sharply is the 
'representation': it must be something that can be transmitted over a 
network, and it must be finite. We can add, it must also be 
accompanied by a (in the http case) 200 code, i.e. (more generally) 
an accepted marker that it is a normal or correct response made 
directly to the request. Then an 'information resource' is 
*something* that emits, or which can emit, such representations when 
prodded. There doesn't seem to be any need to focus on the exact 
metaphysical nature of these things: such speculations play no role 
in the architecture or indeed the semantics of the Web, since the 
'actual' information resources are invisible. All we can detect by 
any empirical means are the representations and codes that they spit 
out at us.

Maybe "HTTP endpoint" isn't exactly correct, I don't really care. So 
replace "HTTP endpoint" by "thing which emits representations 
accompanied by codes indicating success, in response to a GET 
request": that's the definition of an "information resource".

Pat


>
>Indeed.
>
>--
>John Cowan   http://ccil.org/~cowan    cowan@ccil.org
>In might the Feanorians / that swore the unforgotten oath
>brought war into Arvernien / with burning and with broken troth.
>and Elwing from her fastness dim / then cast her in the waters wide,
>but like a mew was swiftly borne, / uplifted o'er the roaring tide.
>         --the Earendillinwe


-- 
---------------------------------------------------------------------
IHMC		(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.	(850)202 4416   office
Pensacola			(850)202 4440   fax
FL 32502			(850)291 0667    cell
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes

Received on Friday, 28 September 2007 16:25:29 UTC