Re: #rdfms-difference-between-ID-and-about

On Monday, June 25, 2001, at 12:20  PM, pat hayes wrote:

> I wonder how you are able to know about my file name 
> conventions, but they do not conform to www.imc.org/rfc2017.

I fail to see how this is relevant. The relevant RFC seems to be:

http://www.w3.org/Addressing/rfc1738.txt

in which case it doesn't seem to matter what you name your files 
(we can always escape it).

>> You must recognize the definition of Web we are using.
>
> Maybe it would help if you could tell me what that definition 
> is. As far as I can tell (see previous message), the Web in 
> your sense is the entire universe; in fact, since (I am told by 
> our chairman) even imaginary things (unicorns) are resources 
> and so can have URI's , the Web would have to be actually 
> larger than the universe. Is that impression correct? If so, 
> Tim B-L should be given rather more credit than he is getting, 
> since He apparently invented the entire cosmos.

;-) I'm not sure if you've read TimBL's book, "Weaving the Web", 
but I recall a quote along the lines of: "At W3C meetings, we 
have a rule: if it isn't on the Web, it doesn't exist and you 
can't mention it."

I agree that we are not at all clear about what it means for 
something to be "on the Web." It seems to have several:

  - The set of all resources (the domain of discourse, includes 
things like
    unicorns)
  - someplace murky here that I think DanC was referring to with no clear
    definition
  - Everything that has a URI scheme meant to identify it 
(currently files on
    computers, books with ISBNs, etc.)
  - Everything available over the Internet.
  - Everything over HTTP

> I know there are many things that do not have names, and many 
> things (most things) do not have URLs. Maybe you could 
> enlighten me about this; how can I tell if something has a URI 
> and is therefore Webbified

I'd also quickly admit that we need better systems for giving 
things URIs and defining resources. I'd like to know if you have 
any suggestions on how to make this clearer, stemming from your 
background in semiotics, etc.

> Yes, but it isn't your URI. My point was that URI's were only 
> invented a while ago; the term is new.  Names (including file 
> names) have been around much longer. If by "URI" you just mean 
> "name", then why not just say "name" ? (Why are y'all using 
> this  pseudo-technical terminology, if you just mean a plain 
> old idea?)

My understanding is that a URI is a special type of name -- one 
that follows the rules in:

   http://www.ietf.org/rfc/rfc2396.txt

It's an attempt to create a naming system that helps prevent 
name collisions. Combined with a resolution system, it seems to 
have been very useful over the past ten years.

> But referring to something does not incorporate that thing into 
> the referring sign in any way. If a book describes Venice, it 
> would be misleading, to put it mildly, to say that Venice was 
> somehow 'in' the book, or that the book 'contained' Venice. It 
> *refers* to Venice.

Of course. There is no argument here.

> If your point is that anything, including a document, might 
> have a URI because anything might be referred to by a URI (much 
> as I might give a name to a speck of dust by simply declaring 
> that I am going to call it 'speck17'), then of course in that 
> sense anything nameable is what might be called 
> web-refer-to-able.

I continue to agree.

> But to say that anything that is web-refer-to-able *is* on the 
> Web, or *is* included in the Web, seems to me to be simply a 
> misuse of terminology.

I suppose it all depends on what the Web means, which I'm no 
longer sure of. Historically, it seems to be:

	The World Wide Web (known as "WWW', "Web" or "W3") is the universe of
	network-accessible information, the embodiment of human knowledge.
	- http://www.w3.org/WWW/ $Date: 2001/01/24 18:09:35 $

> On the face of it, it is obviously wrong, since many things are 
> not in fact referred to at all, on the web or off it, and at 
> any given time far more things are referred to off the web than 
> on it; it ignores the presence of many other sources of 
> reference (all of human spoken language, for a start); and, in 
> the case of documents, it seems to confuse reference with 
> containment, since there is an obvious sense in which an HTML 
> file with a URL *is* 'on the web' in a way that something 
> merely referred to by a document is not on the Web.

You seem to be assuming a definition of Web while providing one 
which makes this long sentence very hard for me to understand. 
However, I agree that the things are very confusing this way and 
would appreciate suggestions in fixing it.

>> Just giving it a URI
> But I thought that your position was that it already has a URI; 
> *everything* has a URI, right? (Or did you mean giving it a 
> *new* URI?)

I'm not certain that this is true. While I may have argued this 
position, I do not personally believe it strongly.

> I don't think you and I have had this conversation  before 
> (have we? When?) , though I have had similar discussions with 
> Dan Connolly.

I recall having a conversation a month or two bag which talked 
about grains of sand on the beach, although I'm not inclined to 
go dig it up.

> Indeed, I am not getting it. When people say things that seem 
> obviously silly, I tend to assume as a default that I am not 
> getting something. I find that the best way to find out what 
> they really mean is often to take what they say at face value 
> and point out how silly it seems to be. If pointing out 
> apparent sillinesses is 'stirring up trouble', then I plead 
> guilty; but in fact, I think it actually avoids more trouble 
> than it causes, in the long run.
>
> However, I do know a fair bit about names, references, meanings 
> and semiotics, so maybe a little further dialog might be 
> mutually useful, if you feel up to it.

Thank you for the offer. In this message I've tried to take a 
different approach -- instead of trying hard to argue the URI 
"dogma" (and trying to convince you of its truth) instead I hope 
that my background in the Web and your background in semiotics 
can help us find some sort of compromise to sort out the 
difficult bits of the terminology we (the RDF/Semantic Web 
Community) use.

I do ask that you bear with me if I say things that are silly, 
but please continue to point them out. I am often an optimist, 
but I do think we might actually get somewhere on this.

--
       "Aaron Swartz"      |           Blogspace
  <mailto:me@aaronsw.com>  |  <http://blogspace.com/about/>
<http://www.aaronsw.com/> |     weaving the two-way web

Received on Monday, 25 June 2001 21:20:25 UTC