W3C home > Mailing lists > Public > www-archive@w3.org > December 2002

Re: "Universal Document Identifier"

From: Tim Berners-Lee <timbl@w3.org>
Date: Thu, 26 Dec 2002 08:26:37 -0500
Cc: www-archive@w3.org
To: Sandro Hawke <sandro@w3.org>
Message-Id: <A8DA349E-18D5-11D7-A957-000393914268@w3.org>


On Wednesday, Dec 25, 2002, at 15:02 US/Eastern, Sandro Hawke wrote:

>
> I just came accros your mention of "Universal Document Identifier" in
> http://www.w3.org/DesignIssues/Architecture.html ...
>
> It seems to me that what is identified is not a document, but a shared
> memory region, a virtual location where a document can conceptually be
> stored.   Something very much like a file, actually.
>

Until you can distinguish by a test between levels like the name and 
person
or location and contents one can argue for hour to little use about 
philosophically
which is identified.

The fundamental difference between a document and a file is
that a file has only content bits, while a document has pairs
of  (content bits, metadata which explains how to interpret them).


> So why not "Universal Filename" ?
>
> I have a lot of trouble with this talk of URI's naming documents,
> because a documents seem like static things.

Well,  I'm sorry -- there never will be a perfect match between the 
english
words we have used before and the architectural concepts of a new 
system.
It is worth going to a certain amount of trouble to use words in a way
which matches the english as well as possible. But a certain amount 
only.

> Of course we get this
> whole weird notion of "living documents" from this.

Well, we had forms of living documents before - ask anyone in a stds 
cmttee.
But we will have new things which do not have a good analogy in the 
paper
world - but thats OK.

> The popular web lingo seems to have gravitated toward the location
> terminology: it's all about "visiting" "sites" with "addresses".
> People have no problem understanding why they should keep their
> advertised address constant; it's the idea of deep-linking (that their
> site might be considered to be made up smaller, also-addressible
> locations, even down to fragment ids) that hasn't, I think, really
> sunk into the public consciousness.  (after all, a few clicks of
> navigation isn't so bad for humans.)
>

Yes, it is true that the "deep linking" seems to be less understood in
some way.  It could be that people like to have in any space
an orienting structure.  They feel uncomfortable without it.
They pick up an idea of the structure of where they are as they
go into a web site.  If they are catapulted in they feel uneasy.
But nowadays a link to a book on Amazon is more acceptable because
so many people understand where they are -- and know the surrounding
structure -- when they arrive.

In general, we must get the architecture right more important that
finding that it matches existing terms.  That is the thing
about  a paradigm shift. The world after can't be explained in the
old terms.

>   ruminatingly,
>     -- sandro
>
>
Merry Christmas!
Tim
Received on Thursday, 26 December 2002 08:26:07 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 7 November 2012 14:17:24 GMT