headers
Timur Shemsedinov | 2 Apr 14:33
Picon

URN for person and organization

Hello

Does anybody know URN schema for person and organization except OID (http://www.oid-info.com/) and PIN URN Schema (Network Solutions) ?
Mentioned two schemas are not suitable for my purposes because they supposed to have one central authority for registration.
It is likewise if there was only one organization in the world for domain names registration, which would not be able to delegate sub-ranges to someone else by region or activity field... So, I would prefer something like person://user <at> authorityname
I am in search of such solution...

Timur 

_______________________________________________
Apps-Discuss mailing list
Apps-Discuss <at> ietf.org
https://www.ietf.org/mailman/listinfo/apps-discuss
Permalink | Reply | 22 comments |
headers
Stephane Bortzmeyer | 2 Apr 14:50
Picon

Re: URN for person and organization

On Wed, Apr 02, 2008 at 03:33:04PM +0300,
 Timur Shemsedinov <timur.shemsedinov <at> gmail.com> wrote 
 a message of 46 lines which said:

> It is likewise if there was only one organization in the world for
> domain names registration, which would not be able to delegate
> sub-ranges to someone else by region or activity field... So, I
> would prefer something like person://user <at> authorityname

Why not simply http://timur.shemsedinov.example/ ? No, the use of the
"http" scheme does not imply that there is a HTTP server.

See <http://norman.walsh.name/2006/07/25/namesAndAddresses> and
<http://www.w3.org/TR/2008/WD-cooluris-20080321/> (for solutions that
are a little different)

[There are many other solutions but let's see them one at a time.]
Permalink | Reply | 1 comment |
headers
Keith Moore | 2 Apr 18:10

Re: URN for person and organization

> Why not simply http://timur.shemsedinov.example/ ? No, the use of the
> "http" scheme does not imply that there is a HTTP server.

perhaps not, but it's pretty confusing to use an HTTP URL in this way 
when a huge body of existing practice uses HTTP URLs to point to 
resources that can be accessed via HTTP.
Permalink | Reply | 0 comments |
headers
Randy Presuhn | 2 Apr 20:27
Picon

Re: URN for person and organization

Hi -

> From: "Timur Shemsedinov" <timur.shemsedinov <at> gmail.com>
> To: <discuss <at> apps.ietf.org>
> Sent: Wednesday, April 02, 2008 4:33 AM
> Subject: URN for person and organization
...
> Does anybody know URN schema for person and organization except OID (
> http://www.oid-info.com/) and PIN URN Schema (Network Solutions) ?
> Mentioned two schemas are not suitable for my purposes because they supposed
> to have one central authority for registration.
....

Uhhh, I'm missing something.  How does the use of OID require a central
authority for registration?  A key feature of the way OIDs work is that
they do not require administrative coordinate, much less centralization.

Randy
Permalink | Reply | 18 comments |
headers
Timur Shemsedinov | 2 Apr 22:10
Picon

Re: URN for person and organization

On Wed, Apr 2, 2008 at 9:27 PM, Randy Presuhn <randy_presuhn <at> mindspring.com> wrote:

Uhhh, I'm missing something.  How does the use of OID require a central
authority for registration?  A key feature of the way OIDs work is that
they do not require administrative coordinate, much less centralization.

But they need central register (database) to guarantee uniqueness or central one and sub-registers (hierarchical structure of child authorities).
For example urn:oid:1.2.250.1 - {iso(1) member-body(2) f(250) type-org(1)} so it is a authority tree like domain names, but I prefers domain names to identify authority because it is mode human readable and good for understanding and remembering

Timur
_______________________________________________
Apps-Discuss mailing list
Apps-Discuss <at> ietf.org
https://www.ietf.org/mailman/listinfo/apps-discuss
Permalink | Reply | 17 comments |
headers
Frank Ellermann | 2 Apr 22:46
Picon
Picon

Re: URN for person and organization

Timur Shemsedinov wrote:

> I prefers domain names to identify authority because it is
> mode human readable and good for understanding and remembering

Some of your goals match what RFC 4151 (tag: URI scheme) offers.

 Frank
Permalink | Reply | 0 comments |
headers
Timur Shemsedinov | 2 Apr 22:51
Picon

Re: URN for person and organization

Oh, I forgot following alternatives:
3) DOI - http://doi.org/ - The Digital Object Identifier System
4) OpenID - http://openid.net/ - OpenID Foundation (OIDF)
5) GUID (Globally Unique Identifier), UUID (Universally Unique Identifier) and other integer-numeric based random or calculated methods

OpenID offers us to use HTTP schema - I do not like it - it brings misunderstanding in URL/URN usage. I think URN can be resolved / reflected into one or many of URLs
GIUD - is not human-friendly
DOI is much better approach but who have any other suggestions on person and organization identification?

Timur

_______________________________________________
Apps-Discuss mailing list
Apps-Discuss <at> ietf.org
https://www.ietf.org/mailman/listinfo/apps-discuss
Permalink | Reply | 7 comments |
headers
Nicolas Williams | 2 Apr 22:55
Picon

Re: URN for person and organization

On Wed, Apr 02, 2008 at 11:10:37PM +0300, Timur Shemsedinov wrote:
> On Wed, Apr 2, 2008 at 9:27 PM, Randy Presuhn <randy_presuhn <at> mindspring.com>
> wrote:
> 
> > Uhhh, I'm missing something.  How does the use of OID require a central
> > authority for registration?  A key feature of the way OIDs work is that
> > they do not require administrative coordinate, much less centralization.
> >
> 
> But they need central register (database) to guarantee uniqueness or central
> one and sub-registers (hierarchical structure of child authorities).
> For example urn:oid:1.2.250.1 - {iso(1) member-body(2) f(250) type-org(1)}
> so it is a authority tree like domain names, but I prefers domain names to
> identify authority because it is mode human readable and good for
> understanding and remembering

The OID namespace is very similar to the DNS namespace in that it allows
for delegation.

The main issue with OIDs is the lack of standard protocols for: a)
finding authorities for any one arc, b) name resolution (which means
that users are routinely faced with meaningless (to them) numeric OIDs.

Frankly, I'd like to see the ITU-T put forward such protocols, perhaps
based on the DNS itself (hey, why not?  It almost fits the bill as
it is -- I think all that's missing are appropriate TLDs).

Also, everyone knows how to get a domainname.  Few know how to get an
OID.

So, I agree that the DNS is a better option.

Nico
--

--
Permalink | Reply | 7 comments |
headers
Nicolas Williams | 2 Apr 23:02
Picon

Re: URN for person and organization

On Wed, Apr 02, 2008 at 11:51:01PM +0300, Timur Shemsedinov wrote:
> 5) GUID (Globally Unique Identifier), UUID (Universally Unique Identifier)
> and other integer-numeric based random or calculated methods
> 
> GIUD - is not human-friendly

But GUIDs may have useful privacy protection semantics, namely,
pseudonymity.  So I think GUIDs should not be rejected outright.

Of course, GUIDs are not at all user-friendly, certainly not without a
name resolution protocol (which ruins the pseudonymity aspect).  But I
think the way they'd be used is as follows: one peer hands you a GUID-
based person URN, and you then ask the same peer for whatever personal
information associated with the URN that you're allowed to have, and
this is what you display to users -- you'd then store {URN, authority}
in a database or {URN, authority, data} in a database and query it as
necessary.  That's not at all unfriendly, and public references to such
URNs retain the pseudonymity benefit.

Nico
--

--
Permalink | Reply | 4 comments |
headers
Timur Shemsedinov | 2 Apr 23:14
Picon

Re: URN for person and organization

On Wed, Apr 2, 2008 at 11:55 PM, Nicolas Williams <Nicolas.Williams <at> sun.com> wrote:

...
The main issue with OIDs is the lack of standard protocols for: a)
finding authorities for any one arc, b) name resolution (which means
that users are routinely faced with meaningless (to them) numeric OIDs.
...
Also, everyone knows how to get a domainname.  Few know how to get an
OID. So, I agree that the DNS is a better option.
 
So, we have OID, DOI, OpenID, GUID, UUID, PIN, IUID, OUI (from IETF), etc...
Some of them uses integer-numbers, alpha-numeric random-generated or algorithmic-generated, URL/URN/URI syntax or other syntactic constructions, using HTTP or other inappropriate schemas (why not FTP: or TELNET: - it is comically situation at all)
So is it a task for IETF or ISO or IEEE or someone else to organize this field?
We need something simple and practical... I see that all approaches have its benefits, even similar constructions
Problem is that URL/URN/URI gives unnecessary flexibility in objects identification (not fixed by other documents), so as a result we have multiple approaches.

Timur
_______________________________________________
Apps-Discuss mailing list
Apps-Discuss <at> ietf.org
https://www.ietf.org/mailman/listinfo/apps-discuss
Permalink | Reply | 3 comments |

Gmane