Re: [soa-rm-ra] further thoughts on visibility and publish-find-bind

[Ken Laskey <klaskey@mitre.org>]

Michael,

You are correct but I was thinking more about policies and technical
assumptions.

As I said, rough capture of thoughts and that is the kind of clean-up
that would need to be incorporated if I (for project work)/we (for RA
work) take this forward.

Ken

At 08:36 AM 7/21/2006, Michael Stiefel wrote:

 "If any of these have changed with service substitution, the execution context must be reestablished."

Isn't the URI itself part of the execution context? So even if the URI is substituted for, wouldn't that be a change in execution context?

I suppose another way of asking this question is: What isn't part of the execution context?

Michael


At 12:17 AM 7/19/2006, Ken Laskey wrote:

While working on a project for my day job, I cautioned against blind 
recitation of the publish-find-bind mantra and was asked for an 
alternative.  This followed our discussion of Danny's visibility 
write-up and so I scratched down some thoughts and exchanged a few 
emails with Danny for a sanity check.  Below is a capture of our 
exchange and the attached (apologies for the length) are the ideas as 
they stand after incorporating some (but not all) of the discussion.

(Sorry for the Word attachment but (1) wasn't sure where to put it in 
the wiki and (2) it's too late to fool around with uploading and 
positioning figures.   Can remedy later.)

The email capture follows.  As typical, read from the bottom up.

Ken


From:     danny_thornton2@yahoo.com
Subject:        Re: thoughts on registries, etc.
Date:   July 13, 2006 12:19:23 PM EDT

I concur with your distinction between global and groups.  It also 
does a nice job of mentioning taxonomies.  You have captured the two 
forms of global visibility that I would expect to see the 
architecture discuss.  While an "aggregate of information expressed 
in many vocabularies and categorized using many taxonomies" is likely 
to be the more prevalent type of visibility in a global SOA, 
significant efforts are being made for the GIG architecture to 
"mediate capabilities that enable discoveries across vocabularies" as 
well as classification levels.

Danny


--- Ken Laskey <klaskey@mitre.org> wrote:

Danny,

Thanks for the quick feedback.  What I sent was somewhat of a 
ramble and a capture of thoughts without things completely worked 
out, without a clear idea of what points needed to be made, without 
a clear idea of how and/or where it eventually needs to be 
incorporated.  My following comments continue in that tradition :-)

Global could certainly be looked at as a dynamic group of groups.  
I am (and I'm sure most other people are too) uncertain to what 
extent there will be significant content generated specifically for 
global, so I'm not sure I'd characterize it exclusively as a group 
of groups.  However, that raises a possibly interesting 
distinction.  As a group of groups, global will be an aggregate of 
information expressed in many vocabularies and categorized using 
many taxonomies (or other representations).  Thus, either a global 
registry search will be akin to a search of the group(s) using that 
vocabulary and taxonomy (effectively, a search over a very small 
part of the entire registry content) or the global registry will 
have to incorporate significant mediation capabilities to enable 
discovery across vocabularies.  I think this could be a useful 
distinction between group and global and a driver for use of the 
global registry. Figuring out how it would be appropriate to 
express this will be a challenge and will require more thought.

Note, I think personal, group, and global somewhat align with 
simple, complex, and mediated, and those categories probably 
influenced my thinking but I was toying with concepts without
trying to specifically match anything else.  No matter what the 
categorization, I think the boundaries may be fuzzy so, for 
example, three people can jointly maintain a personal catalog 
without fully being a group and three groups could work together 
without getting to global.  As we proceed I expect the boundaries 
to either become more clear or get permanently fuzzy.

The initial focus on UDDI and Web services had a lot to do with the 
actual work discussion that got me to start writing this down.  I 
agree that it needs to be expanded beyond UDDI but that is the 
prevalent straw man and somehow needs direct attention because I 
think the UDDI structure exacerbates the registry problem.  Again, 
this was in no way meant to be complete and your comments are 
valid. Again more thought is needed both on scope and presentation.

Ken



On Jul 12, 2006, at 3:06 AM, Danny Thornton wrote:

Hi Ken,

The personal, group, and global categories make sense within the 
context of your document.  I can see how group can always imply 
some type of ownership domain. The question I ask myself is if 
global is another type of group.  If global is distinguished as a 
dynamic ecosystem of groups, then that would make sense to me.

 It would certainly be easy to integrate personal, group, and 
global levels into the simple, complex, and mediated sections of 
Visibility.   I currently have one and two liner examples 
equivalent to the personal, group, and global examples. Your 
examples could be substituted for those one and two liner examples.

The beginning of your document focuses on UDDI, WSDL, and Web 
Services.  I would also incorporate discussions about ebXML, the 
Process Model, and repositories.  See

http://www.globalresponsesystems.org/portal/default/forums/ ForumsPortletWindow?op=showTopic&windowstate=maximized&t=9&mask=a

for my review of the standards against the SOA RA.

You also touch on consistency between what is done now and what 
was done in the past and how that can be trusted to be true.   I 
would also mention the trust model and the establishment of roots 
of trust between groups in a SOA.

Danny


--- Ken Laskey <klaskey@mitre.org> wrote:

Danny,

You are working the visibility piece and I've been thinking about 
different levels of resource cataloguing for visibility since I 
commented on your draft.  Interestingly, it brought together some 
ideas I've been chewing on and the attached is a rambling 
capture. I'm not sure how this fits into the RA but I'd be 
interested in how this sounds to you and where it needs beefing 
up the most.

Thanks for any comments.

Ken


---
Ken Laskey
MITRE Corporation, M/S H305     phone:  703-983-7934
7515 Colshire Drive                        fax:        703-983-1379
McLean VA 22102-7508





While working on a project for my day job, I cautioned against blind recitation of the publish-find-bind mantra and was asked for an alternative.  This followed our discussion of Danny's visibility write-up and so I scratched down some thoughts and exchanged a few emails with Danny for a sanity check.  Below is a capture of our exchange and the attached (apologies for the length) are the ideas as they stand after incorporating some (but not all) of the discussion. 

(Sorry for the Word attachment but (1) wasn't sure where to put it in the wiki and (2) it's too late to fool around with uploading and positioning figures.   Can remedy later.)

The email capture follows.  As typical, read from the bottom up.

Ken


From:   danny_thornton2@yahoo.com
Subject: Re: thoughts on registries, etc.
Date: July 13, 2006 12:19:23 PM EDT

I concur with your distinction between global and groups.  It also does a nice job of mentioning taxonomies.  You have captured the two forms of global visibility that I would expect to see the architecture discuss.  While an "aggregate of information expressed in many vocabularies and categorized using many taxonomies" is likely to be the more prevalent type of visibility in a global SOA, significant efforts are being made for the GIG architecture to "mediate capabilities that enable discoveries across vocabularies" as well as classification levels.

Danny


--- Ken Laskey <klaskey@mitre.org> wrote:

Danny,

Thanks for the quick feedback.  What I sent was somewhat of a ramble and a capture of thoughts without things completely worked out, without a clear idea of what points needed to be made, without a clear idea of how and/or where it eventually needs to be incorporated.  My following comments continue in that tradition :-)

Global could certainly be looked at as a dynamic group of groups.  I am (and I'm sure most other people are too) uncertain to what extent there will be significant content generated specifically for global, so I'm not sure I'd characterize it exclusively as a group of groups.  However, that raises a possibly interesting distinction.  As a group of groups, global will be an aggregate of information expressed in many vocabularies and categorized using many taxonomies (or other representations).  Thus, either a global registry search will be akin to a search of the group(s) using that vocabulary and taxonomy (effectively, a search over a very small part of the entire registry content) or the global registry will have to incorporate significant mediation capabilities to enable discovery across vocabularies.  I think this could be a useful distinction between group and global and a driver for use of the global registry. Figuring out how it would be appropriate to express this will be a challenge and will require more thought.

Note, I think personal, group, and global somewhat align with simple, complex, and mediated, and those categories probably influenced my thinking but I was toying with concepts without
trying to specifically match anything else.  No matter what the categorization, I think the boundaries may be fuzzy so, for example, three people can jointly maintain a personal catalog without fully being a group and three groups could work together without getting to global.  As we proceed I expect the boundaries to either become more clear or get permanently fuzzy.

The initial focus on UDDI and Web services had a lot to do with the actual work discussion that got me to start writing this down.  I agree that it needs to be expanded beyond UDDI but that is the prevalent straw man and somehow needs direct attention because I think the UDDI structure exacerbates the registry problem.  Again, this was in no way meant to be complete and your comments are valid. Again more thought is needed both on scope and presentation.

Ken



On Jul 12, 2006, at 3:06 AM, Danny Thornton wrote:

Hi Ken,

The personal, group, and global categories make sense within the context of your document.  I can see how group can always imply some type of ownership domain. The question I ask myself is if global is another type of group.  If global is distinguished as a dynamic ecosystem of groups, then that would make sense to me.

 It would certainly be easy to integrate personal, group, and global levels into the simple, complex, and mediated sections of Visibility.   I currently have one and two liner examples equivalent to the personal, group, and global examples. Your examples could be substituted for those one and two liner examples.

The beginning of your document focuses on UDDI, WSDL, and Web Services.  I would also incorporate discussions about ebXML, the Process Model, and repositories.  See

http://www.globalresponsesystems.org/portal/default/forums/ ForumsPortletWindow?op=showTopic&windowstate=maximized&t=9&mask=a

for my review of the standards against the SOA RA.

You also touch on consistency between what is done now and what was done in the past and how that can be trusted to be true.   I would also mention the trust model and the establishment of roots of trust between groups in a SOA.

Danny


--- Ken Laskey <klaskey@mitre.org> wrote:

Danny,

You are working the visibility piece and I've been thinking about different levels of resource cataloguing for visibility since I commented on your draft.  Interestingly, it brought together some ideas I've been chewing on and the attached is a rambling capture. I'm not sure how this fits into the RA but I'd be interested in how this sounds to you and where it needs beefing up the most.

Thanks for any comments.

Ken

Content-Type: application/octet-stream;
         x-mac-type=5738424E;
         x-unix-mode=0644;
         x-mac-creator=4D535744;
         name=triangle thoughts.doc
Content-Disposition: attachment;
         filename="triangle thoughts.doc"


---
Ken Laskey
MITRE Corporation, M/S H305     phone:  703-983-7934
7515 Colshire Drive                        fax:        703-983-1379
McLean VA 22102-7508

--
     ---------------------------------------------------------------------------------
  /   Ken Laskey                                                                \
 |    MITRE Corporation, M/S H305    phone:  703-983-7934   |
 |    7515 Colshire Drive                    fax:      703-983-1379   |
  \   McLean VA 22102-7508                                              /
    ----------------------------------------------------------------------------------