FW: New Version Notification for draft-masinter-dated-uri-06

[Larry Masinter-3]

After a long hiatus on "tdb", I updated the document
based on some feedback.

Changed from URN scheme to URI, got rid of "duri"
and just left "tdb".



-----Original Message-----
From: IETF I-D Submission Tool [mailto:idsubmission@...]
Sent: Sunday, July 12, 2009 10:43 PM
To: lmm@...
Cc: LMM@...
Subject: New Version Notification for draft-masinter-dated-uri-06


A new version of I-D, draft-masinter-dated-uri-06.txt has been successfuly submitted by Larry Masinter and posted to the IETF repository.

Filename: draft-masinter-dated-uri
Revision: 06
Title: The "tdb" URI scheme: denoting described resources
Creation_date: 2009-07-12
WG ID: Independent Submission
Number_of_pages: 13

Abstract:
This document defines a URI scheme, "tdb" ( standing for "Thing
Described By").  It provides a semantic hook for allowing anyone at
any time to mint a URI for anything that they can describe.  Such
URIs may include a timestamp to fix the description at a given date
or time.

This URI scheme may reduce the need to define define new URN
namespaces merely for the purpose of creating stable identifiers.  In
addition, they provide a ready means for identifying "non-information
resources" by semantic indirection -- a way of creating a URI for
anything.

Note

This document is not a product of any working group.  Many of the
ideas here have been discussed since 2001.  This document has been
discussed on the mailing list <uri@...>.  Previous versions have
couched "tdb" as a URN namespace, and included a "duri" scheme for
fixing date without indirection, which seems unnecessary.  It was
originally written as a thought experiment as a way of resolving the
use/mention problem in semantic web applications, but may have other
uses.
                                                                                 


The IETF Secretariat.

[Dan Brickley-2]

On 13/7/09 17:59, Larry Masinter wrote:
> After a long hiatus on "tdb", I updated the document
> based on some feedback.
>
> Changed from URN scheme to URI, got rid of "duri"
> and just left "tdb".

Thanks for keeping this alive!

Side-question: Does it work with 404s? Oddly enough I'm looking for "the
thing described by"
http://www.ietf.org/internet-drafts/draft-masinter-dated-uri-05.txt (now
404'ing per IETF tradition), to remind myself what the "duri" piece did
exactly.


The relationship foaf:primaryTopic is, as far as I can see, an
expression of much the same notion. So a quick compare/contrast here:


http://xmlns.com/foaf/spec/#term_primaryTopic

At the moment all it says there is

"The foaf:primaryTopic property relates a document to the main thing
that the document is about.

The foaf:primaryTopic property is functional: for any document it
applies to, it can have at most one value. This is useful, as it allows
for data merging. In many cases it may be difficult for third parties to
determine the primary topic of a document, but in a useful number of
cases (eg. descriptions of movies, restaurants, politicians, ...) it
should be reasonably obvious. Documents are very often the most
authoritative source of information about their own primary topics,
although this cannot be guaranteed since documents cannot be assumed to
be accurate, honest etc. "

So wherever we have

        some URI ?x foaf:primaryTopic  some URI ?y,
(and also ?x must be a document of some kind)

then
        "tbd:"+?x is a URI naming the same thing as URI ?y ...

The tbd: spec allows mailto: URIs to be used to be tdb for people. FOAF
does this directly using foaf:mbox, or indirectly with
foaf:mbox_sha1sum, but not using the primaryTopic construct.



Temporal aspects are the other slightly awkward fit between the two
approaches. In FOAF, all that was left unconstrained. Documents can
change their primary topic over time. This feels a little too loose, but
so far nobody's pushed for a more rigid version.

In http://tools.ietf.org/search/draft-masinter-dated-uri-06

[[
    The goal, then, of the "tdb" URI scheme is to provide a mechanism
    which is, at the same time:

       permanent: The identity of the resource identified is not subject
       to reinterpretation over time.

       explicitly bound: The mechanism by which the identified resource
       can be determined is explicitly included in the URI.
]]


Hmm seems you left in some duri pieces?

"Previous versions have couched "tdb" as a URN namespace, and included a
"duri" scheme for fixing date without indirection, which seems unnecessary."

vs

"2. Syntax
    A tdb URI takes the form:
         duri:<timestamp>:<URI>"

and

"   The meaning of a duri is "the resource (or fragment) that was
    identified by the <encoded-URI> (after hex decoding) at the very last
    instant of the date(time) given"."


I'm a bit confused now about what was in tbd vs in duri. How much of the
date-stamping piece was supposed to be left in?

cheers,

Dan

[Larry Masinter-3]

Thanks! Fixed in http://larry.masinter.net/duri.txt (and .html .xml).

The intent was to leave in the timestamp.  "duri" only did
timestamping. "tdb" does semantic indirection with a timestamp.

I intended to allow the possibility of untimestamped "duri"
but discouraging it, e.g., in constructs such as tdb::data:,
but even then it's tricky.

What is the difference in meaning between
 tdb::data:application/uri,SOMETHING .

Larry
--
http://larry.masinter.net


-----Original Message-----
From: Dan Brickley [mailto:danbrickley@...] On Behalf Of Dan Brickley
Sent: Tuesday, July 14, 2009 12:53 AM
To: Larry Masinter
Cc: URI
Subject: Re: FW: New Version Notification for draft-masinter-dated-uri-06

On 13/7/09 17:59, Larry Masinter wrote:
> After a long hiatus on "tdb", I updated the document
> based on some feedback.
>
> Changed from URN scheme to URI, got rid of "duri"
> and just left "tdb".

Thanks for keeping this alive!

Side-question: Does it work with 404s? Oddly enough I'm looking for "the
thing described by"
http://www.ietf.org/internet-drafts/draft-masinter-dated-uri-05.txt (now
404'ing per IETF tradition), to remind myself what the "duri" piece did
exactly.


The relationship foaf:primaryTopic is, as far as I can see, an
expression of much the same notion. So a quick compare/contrast here:


http://xmlns.com/foaf/spec/#term_primaryTopic

At the moment all it says there is

"The foaf:primaryTopic property relates a document to the main thing
that the document is about.

The foaf:primaryTopic property is functional: for any document it
applies to, it can have at most one value. This is useful, as it allows
for data merging. In many cases it may be difficult for third parties to
determine the primary topic of a document, but in a useful number of
cases (eg. descriptions of movies, restaurants, politicians, ...) it
should be reasonably obvious. Documents are very often the most
authoritative source of information about their own primary topics,
although this cannot be guaranteed since documents cannot be assumed to
be accurate, honest etc. "

So wherever we have

        some URI ?x foaf:primaryTopic  some URI ?y,
(and also ?x must be a document of some kind)

then
        "tbd:"+?x is a URI naming the same thing as URI ?y ...

The tbd: spec allows mailto: URIs to be used to be tdb for people. FOAF
does this directly using foaf:mbox, or indirectly with
foaf:mbox_sha1sum, but not using the primaryTopic construct.



Temporal aspects are the other slightly awkward fit between the two
approaches. In FOAF, all that was left unconstrained. Documents can
change their primary topic over time. This feels a little too loose, but
so far nobody's pushed for a more rigid version.

In http://tools.ietf.org/search/draft-masinter-dated-uri-06

[[
    The goal, then, of the "tdb" URI scheme is to provide a mechanism
    which is, at the same time:

       permanent: The identity of the resource identified is not subject
       to reinterpretation over time.

       explicitly bound: The mechanism by which the identified resource
       can be determined is explicitly included in the URI.
]]


Hmm seems you left in some duri pieces?

"Previous versions have couched "tdb" as a URN namespace, and included a
"duri" scheme for fixing date without indirection, which seems unnecessary."

vs

"2. Syntax
    A tdb URI takes the form:
         duri:<timestamp>:<URI>"

and

"   The meaning of a duri is "the resource (or fragment) that was
    identified by the <encoded-URI> (after hex decoding) at the very last
    instant of the date(time) given"."


I'm a bit confused now about what was in tbd vs in duri. How much of the
date-stamping piece was supposed to be left in?

cheers,

Dan