RE: [wsdm] Definition of the identification manageability capability

["Sedukhin, Igor S" <Igor.Sedukhin@ca.com>]

I still did not read why opaque identifier is REQUIRED to make the WSDM 0.5 interop scenario work. Your explanation sounds like "we need it for the principle". Could you explain why the scanario will not work if we didn't have opaque identifier in March 04?

I believe that a reference to an endpoint is sufficient for WSDM 0.5.


-- Igor Sedukhin .. (igor.sedukhin@ca.com)
-- (631) 342-4325 .. 1 CA Plaza, Islandia, NY 11788

-----Original Message-----
From: MURRAY,BRYAN (HP-Seattle,ex1) [mailto:bryan.murray@hp.com] 
Sent: Thursday, December 11, 2003 12:29 PM
To: Sedukhin, Igor S; VAMBENEPE,WILLIAM (HP-Cupertino,ex1); John DeCarlo
Cc: WSDM TC
Subject: RE: [wsdm] Definition of the identification manageability capability

The March release will include portions of MUWS as well as portions of MOWS.
MUWS needs a way to identify a resource that is globally unique and constant over time in order to determine whether 2 resources are the same. This identifier is opaque because it is meant to be used to differentiate between different resources - it is not meant to be used to locate the resource.

MOWS uses MUWS to manage web service resources. Thus, MOWS will use the MUWS mechanism to identify resources. The proposals we have seen so far do an adequate job of defining the type of this 'resource identifier'.

MOWS can add additional identity properties to help a manager understand what it is managing, but at the core, the globally unique id is sufficient to let the manager know that different resources are being managed if their ids are different.

Bryan

-----Original Message-----
From: Sedukhin, Igor S [mailto:Igor.Sedukhin@ca.com]
Sent: Thursday, December 11, 2003 9:05 AM
To: VAMBENEPE,WILLIAM (HP-Cupertino,ex1); John DeCarlo
Cc: WSDM TC
Subject: RE: [wsdm] Definition of the identification manageability capability


I don't think opaque identifier is at all needed to make the introp scenario
that we discussed at the F2F work. Could you please explain why do you think
that it will be required to make it work? I have no idea... 

It may take us a long time to agree on why exactly something like that is
needed. Design is easy - use UUIDs.

For March 04 giving a reference to the endpoint is good enough.

-- Igor Sedukhin .. (igor.sedukhin@ca.com)
-- (631) 342-4325 .. 1 CA Plaza, Islandia, NY 11788

-----Original Message-----
From: VAMBENEPE,WILLIAM (HP-Cupertino,ex1) [mailto:vbp@hp.com] 
Sent: Thursday, December 11, 2003 11:52 AM
To: Sedukhin, Igor S; John DeCarlo
Cc: WSDM TC
Subject: RE: [wsdm] Definition of the identification manageability
capability


I don't understand why the opaque identifier will take us a year to design
and I don't understand how we can release anything in March 04 if this very
basic component is not present.

All we need is something as simple as:

ResourceID[1] {ro, const} - anyURI: A unique identifier that identifies the
resource being managed by the manageability endpoint.

Making this ID unique and not changing over time is the job of the provider,
not something we need to worry about as interface designers.

There you go, I just saved the group one year of work! Where's my medal?

Regards,

William

> -----Original Message-----
> From: Sedukhin, Igor S [mailto:Igor.Sedukhin@ca.com]
> Sent: Thursday, December 11, 2003 8:45 AM
> To: John DeCarlo
> Cc: WSDM TC
> Subject: RE: [wsdm] Definition of the identification manageability
> capability
> 
> 
> John, there is an opaque identifier that lets you tell one "thing"
> from the "other", and there is identification that allows one to tell 
> what one thinks the "thing" is. I was looking for the latter. I 
> actually do no think the former is achievable in a meaningful way 
> before the Mach 04 deadline.
> I'm sure we will spend another year discussing how to make sure the 
> opaque identifier can actually be unique and not change over time.
> 
> If it becomes confusing we can rename the capability to something like
> "WhatAmIManaging" capability. Definition of such capability is what 
> I'm looking to add to the spec. It solves the problem that I know we 
> can solve now.
> 
> -- Igor Sedukhin .. (igor.sedukhin@ca.com)
> -- (631) 342-4325 .. 1 CA Plaza, Islandia, NY 11788
> 
> -----Original Message-----
> From: John DeCarlo [mailto:jdecarlo@mitre.org]
> Sent: Friday, December 12, 2003 11:35 AM
> To: Sedukhin, Igor S
> Cc: WSDM TC
> Subject: Re: [wsdm] Definition of the identification manageability
> capability
> 
> Sedukhin, Igor S wrote:
> 
> > I propose that the following text is incorporated in MUWS
> and MOWS specs.
> >
> > "The identification capability allows the provider of
> manageability to
> > tell what it thinks it is managing. In other words
> refer/address/point
> > to the resource it thinks it made manageable. The exact way of
> > "referring" conforms to the subject domain of the resource."
> >
> Igor,
> 
> This looks nice, but I have some comments on it.
> 
> 1.  I feel somewhat strongly that the "identification" be opaque and
> unique and not domain specific.
> 2.  What you describe is crucial, but it would be nice to be sure that 
> it isn't confused with the global MUWS identifier for the resource.
> 
> Thus, I propose the following:
> 
> "The identification capability provides the consumer of manageability
> with a [globally] unique identifier for each managed resource the 
> provider of manageability thinks it is managing.  This identifier is 
> opaque (not decomposable into domain-specific information about the 
> resource).  It can be used to compare the answers from different 
> providers to see if they are both managing the same resource."
> 
> "The managed resource reference capability allows the provider of
> manageability to tell the consumer information about what it thinks it 
> is managing. In other words refer/address/point to the resource it 
> thinks it made manageable. The exact way of "referring" conforms to 
> the subject domain of the resource."
> 
> What do you think?
> 
> Thanks.
> 
> John DeCarlo, The MITRE Corporation, My Views Are My Own
> 
> 
> 
> 


To unsubscribe from this mailing list (and be removed from the roster of the
OASIS TC), go to
http://www.oasis-open.org/apps/org/workgroup/wsdm/members/leave_workgroup.ph
p.