I think your proposed naming change would make things more clear. To get a feel for what it would look like I have updated the OCCI Core document accordingly. Please read and let me know if there are any objections to the change. Please find the OCCI Core document attached. regards, Ralf On Thu, 04 Nov 2010 18:04:18 +0100, wrote:

Hey Ralf,

I'd like to come back to naming once more (and hopefully for the last name).

The current "Type" and "Kind" make things a bit confusing -- type is so overloaded as a term, and the thing that is currently called "Kind" does not describe a "Kind" at all.

As such, I'd propose to rename as follows:

"Kind" -> "Entity" and "Type" -> "Kind"

This removes the strange notion from the thing that really isn't a "Kind", and fix the overoveroverloading of the "Type", effectively getting us rid of sentences like "the Type type". Plus, with the new naming, the "Kind" really describes what kind a certain "Entity" is of.

Best, Alexander

Am 04.11.2010 um 13:42 schrieb Ralf Nyren:

...

Hi all,

After hard work at OGF30 sorting out the remaining open issue of the OCCI model the Core specification document has received a major update.

There is some new stuff in there but mostly the document has been restructured and rewritten to be easier to read and understand.

We appreciate your comments. If no major issues are found the document will enter the public comment phase in 7 days, that is on November 11:th.

regards, Ralf

Hi Michael, thanks for your feedback. Very good points you make. Reply below. On Sun, 07 Nov 2010 05:35:34 +0100, Michael Behrens wrote:

2-cents: Structural and Non-Structural concept might be confusing to folks reading it the first time through. Perhaps its purpose (extensibility) could be stated before their definitions in a non normative manner. Lastly, would adding two subclass of kind (structured/unstructured) help clear things a bit? (The text seems to speak as if there are two subclasses).

Perhaps an non-formal intro text in Core would solve this. We could put some more context behind the core model, why it exist etc, without going into the specifics. Any objections? Regarding structural/non-structural Kind the Kind type does not gain the structural/non-structural property until it is instantiated. The Core doc says in a few places that "a structural Kind" is an _instance_ of Kind etc. However, it can indeed still be quite confusing. Splitting Kind into two classes each inheriting Category might be a solution. That would automatically solve the issue that a non-structural Kind MUST NOT be related to a structural Kind. I'll see if I can draw up a UML example. Anyone else with an opinion on this? regards, Ralf

Michael, Please find attached a version of the core model with Kind split into two separate classes. Was it something like this you were looking for? To me it makes sense to do the split. Before we had the abstraction between Category and Kind it was tempting to stuff all functionality into the Category. I do not think it is anymore. I think this, exactly as you say Michael, definitely help clear things up a bit :) If there are any objections I need them asap, if this is going in I need to start updating the core doc tomorrow. And if anyone has a better name than "Mixin" please speak up! regards, Ralf On Sun, 07 Nov 2010 05:35:34 +0100, Michael Behrens wrote:

I see that the core UML model has been updated, interesting changes. The name changes look okay to me (Entity, Kind).

2-cents: Structural and Non-Structural concept might be confusing to folks reading it the first time through. Perhaps its purpose (extensibility) could be stated before their definitions in a non normative manner. Lastly, would adding two subclass of kind (structured/unstructured) help clear things a bit? (The text seems to speak as if there are two subclasses).

It also looks remarkably like a Transformer... On Tue, Nov 9, 2010 at 2:50 AM, Michael Behrens wrote:

The diagram looks good & reads well to me. Thanks.

Ralf Nyren wrote:

Michael,

Please find attached a version of the core model with Kind split into two separate classes. Was it something like this you were looking for?

To me it makes sense to do the split. Before we had the abstraction between Category and Kind it was tempting to stuff all functionality into the Category. I do not think it is anymore.

I think this, exactly as you say Michael, definitely help clear things up a bit :)

If there are any objections I need them asap, if this is going in I need to start updating the core doc tomorrow. And if anyone has a better name than "Mixin" please speak up!

regards, Ralf

On Sun, 07 Nov 2010 05:35:34 +0100, Michael Behrens wrote:

I see that the core UML model has been updated, interesting changes. The name changes look okay to me (Entity, Kind).

2-cents: Structural and Non-Structural concept might be confusing to folks reading it the first time through. Perhaps its purpose (extensibility) could be stated before their definitions in a non normative manner. Lastly, would adding two subclass of kind (structured/unstructured) help clear things a bit? (The text seems to speak as if there are two subclasses).

core_model.png

-- Michael Behrens R2AD, LLC (571) 594-3008 (cell) (703) 714-0442 (land)

_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

Interoperon? On Tue, Nov 9, 2010 at 8:34 AM, wrote:

Autobot or Decepticon?

I like the new structure, and also the naming, btw...

-Alexander

Am 09.11.2010 um 08:25 schrieb Alexis Richardson:

...

It also looks remarkably like a Transformer...

On Tue, Nov 9, 2010 at 2:50 AM, Michael Behrens wrote: The diagram looks good & reads well to me. Thanks.

Ralf Nyren wrote:

...

Michael,

Please find attached a version of the core model with Kind split into two separate classes. Was it something like this you were looking for?

To me it makes sense to do the split. Before we had the abstraction between Category and Kind it was tempting to stuff all functionality into the Category. I do not think it is anymore.

I think this, exactly as you say Michael, definitely help clear things up a bit :)

If there are any objections I need them asap, if this is going in I need to start updating the core doc tomorrow. And if anyone has a better name than "Mixin" please speak up!

regards, Ralf

On Sun, 07 Nov 2010 05:35:34 +0100, Michael Behrens wrote:

...

I see that the core UML model has been updated, interesting changes. The name changes look okay to me (Entity, Kind).

2-cents: Structural and Non-Structural concept might be confusing to folks reading it the first time through. Perhaps its purpose (extensibility) could be stated before their definitions in a non normative manner. Lastly, would adding two subclass of kind (structured/unstructured) help clear things a bit? (The text seems to speak as if there are two subclasses).

core_model.png

-- Michael Behrens R2AD, LLC (571) 594-3008 (cell) (703) 714-0442 (land)

_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

Generally looks good I guess - thanks for this Gary - very helpful! I'm just wondering if the Entity's mixin attribtue should be a set...That's what the core diagram says at least... Thanks, -Thijs -----Original Message----- From: occi-wg-bounces@ogf.org on behalf of Gary Mazz Sent: Wed 10/11/2010 08:34 To: occi-wg@ogf.org Subject: Re: [occi-wg] OCCI Core ready for public comment version Hi, I tool the opportunity to auto generate code from the UML. We should take a look at it and see if this is what we really mean. cheers, gary On 11/8/2010 7:50 PM, Michael Behrens wrote:

The diagram looks good & reads well to me. Thanks.

Ralf Nyren wrote:

...

Michael,

Please find attached a version of the core model with Kind split into two separate classes. Was it something like this you were looking for?

To me it makes sense to do the split. Before we had the abstraction between Category and Kind it was tempting to stuff all functionality into the Category. I do not think it is anymore.

I think this, exactly as you say Michael, definitely help clear things up a bit :)

If there are any objections I need them asap, if this is going in I need to start updating the core doc tomorrow. And if anyone has a better name than "Mixin" please speak up!

regards, Ralf

On Sun, 07 Nov 2010 05:35:34 +0100, Michael Behrens wrote:

...

I see that the core UML model has been updated, interesting changes. The name changes look okay to me (Entity, Kind).

2-cents: Structural and Non-Structural concept might be confusing to folks reading it the first time through. Perhaps its purpose (extensibility) could be stated before their definitions in a non normative manner. Lastly, would adding two subclass of kind (structured/unstructured) help clear things a bit? (The text seems to speak as if there are two subclasses).

core_model.png

-- Michael Behrens R2AD, LLC (571) 594-3008 (cell) (703) 714-0442 (land)

_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

Well, also various types are FUBAR. Anyway, the XMI model is probably rather incomplete in terms of type imports and the like, and -- in its current state -- is probably not suitable for roundtrip engineering. Would be great if someone could volunteer for polishing this towards a fully compliant XMI model, with all the gory details... I certainly won't. -Alexander Am 11.11.2010 um 10:09 schrieb Ralf Nyren:

True, overall ok but the UML-to-Java tool does not seem to have taken the association multiplicity into account. As Thijs says, there should be Set<Entity> etc.

regards, Ralf

On Thu, 11 Nov 2010 09:55:13 +0100, Thijs Metsch wrote:

...

Generally looks good I guess - thanks for this Gary - very helpful!

I'm just wondering if the Entity's mixin attribtue should be a set...That's what the core diagram says at least...

Thanks,

-Thijs

-----Original Message----- From: occi-wg-bounces@ogf.org on behalf of Gary Mazz Sent: Wed 10/11/2010 08:34 To: occi-wg@ogf.org Subject: Re: [occi-wg] OCCI Core ready for public comment version Hi,

I tool the opportunity to auto generate code from the UML. We should take a look at it and see if this is what we really mean.

cheers, gary

On 11/8/2010 7:50 PM, Michael Behrens wrote:

...

The diagram looks good & reads well to me. Thanks.

Ralf Nyren wrote:

...

Michael,

Please find attached a version of the core model with Kind split into two separate classes. Was it something like this you were looking for?

To me it makes sense to do the split. Before we had the abstraction between Category and Kind it was tempting to stuff all functionality into the Category. I do not think it is anymore.

I think this, exactly as you say Michael, definitely help clear things up a bit :)

If there are any objections I need them asap, if this is going in I need to start updating the core doc tomorrow. And if anyone has a better name than "Mixin" please speak up!

regards, Ralf

On Sun, 07 Nov 2010 05:35:34 +0100, Michael Behrens wrote:

...

I see that the core UML model has been updated, interesting changes. The name changes look okay to me (Entity, Kind).

2-cents: Structural and Non-Structural concept might be confusing to folks reading it the first time through. Perhaps its purpose (extensibility) could be stated before their definitions in a non normative manner. Lastly, would adding two subclass of kind (structured/unstructured) help clear things a bit? (The text seems to speak as if there are two subclasses).

core_model.png

-- Michael Behrens R2AD, LLC (571) 594-3008 (cell) (703) 714-0442 (land)

_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

Category attributes are there by association (a type of inheritance by associative composition). Andy andy.edmonds.be On Thu, Nov 11, 2010 at 17:33, Gary Mazz wrote:

I think the real issue is the disconnect between the Resource and the Category. Resource extends Entity and Entity is a stand alone abstract class. There is no inheritance from Category, so no Category attributes in Resource.. We need a generalization relationship between Category and Entity.

Multiplicity and Set(s) are implementation details, possibly not appropriate for this model.

cheers, gary

On 11/11/2010 2:09 AM, Ralf Nyren wrote:

...

True, overall ok but the UML-to-Java tool does not seem to have taken the association multiplicity into account. As Thijs says, there should be Set<Entity> etc.

regards, Ralf

On Thu, 11 Nov 2010 09:55:13 +0100, Thijs Metsch wrote:

...

Generally looks good I guess - thanks for this Gary - very helpful!

I'm just wondering if the Entity's mixin attribtue should be a set...That's what the core diagram says at least...

Thanks,

-Thijs

-----Original Message----- From: occi-wg-bounces@ogf.org on behalf of Gary Mazz Sent: Wed 10/11/2010 08:34 To: occi-wg@ogf.org Subject: Re: [occi-wg] OCCI Core ready for public comment version Hi,

I tool the opportunity to auto generate code from the UML. We should take a look at it and see if this is what we really mean.

cheers, gary

On 11/8/2010 7:50 PM, Michael Behrens wrote:

...

The diagram looks good & reads well to me. Thanks.

Ralf Nyren wrote:

...

Michael,

Please find attached a version of the core model with Kind split into two separate classes. Was it something like this you were looking for?

To me it makes sense to do the split. Before we had the abstraction between Category and Kind it was tempting to stuff all functionality into the Category. I do not think it is anymore.

I think this, exactly as you say Michael, definitely help clear things up a bit :)

If there are any objections I need them asap, if this is going in I need to start updating the core doc tomorrow. And if anyone has a better name than "Mixin" please speak up!

regards, Ralf

On Sun, 07 Nov 2010 05:35:34 +0100, Michael Behrens wrote:

...

I see that the core UML model has been updated, interesting changes. The name changes look okay to me (Entity, Kind).

2-cents: Structural and Non-Structural concept might be confusing to folks reading it the first time through. Perhaps its purpose (extensibility) could be stated before their definitions in a non normative manner. Lastly, would adding two subclass of kind (structured/unstructured) help clear things a bit? (The text seems to speak as if there are two subclasses).

core_model.png

-- Michael Behrens R2AD, LLC (571) 594-3008 (cell) (703) 714-0442 (land)

_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

Hi Pedro, Which document are you referring to? Could you provide a reference? I am afraid you might have found some old document which is not valid with regard to the upcoming release of the OCCI spec. regards, Ralf On Thu, 04 Nov 2010 19:58:55 +0100, Pedro Gonçalves wrote:

Hi all

in the section 1.4.2.1 (Querying) the example defines a link with relation search and then shows a response message with the mime-type application/occi

shouldn't this be instead application/opensearchdescription+xml and return a opensearch description document (according to references in table 3)

regards

Pedro

On Nov 4, 2010, at 1:42 PM, Ralf Nyren wrote:

...

Hi all,

After hard work at OGF30 sorting out the remaining open issue of the OCCI model the Core specification document has received a major update.

There is some new stuff in there but mostly the document has been restructured and rewritten to be easier to read and understand.

We appreciate your comments. If no major issues are found the document will enter the public comment phase in 7 days, that is on November 11:th.

regards, Ralf _______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

The OCCI Core is one of three documents that together form the OCCI specification. Discovery is described in Core and the specifics on finding supported resource types over HTTP is described in the HTTP rendering doc due to be made available shortly. Hope that answers your question. regards, Ralf On Thu, 04 Nov 2010 21:03:34 +0100, Pedro Gonçalves wrote:

Hi Ralf

yes, you're right :) I was checking the differences between the file you sent and the file available i had here (the one from the web site most probably) and got the docs mixed up

All the search capabilities that were in that version were removed in the version you sent, is this right ?

regards

Pedro

On Nov 4, 2010, at 8:33 PM, Ralf Nyren wrote:

...

Hi Pedro,

Which document are you referring to? Could you provide a reference?

I am afraid you might have found some old document which is not valid with regard to the upcoming release of the OCCI spec.

regards, Ralf

On Thu, 04 Nov 2010 19:58:55 +0100, Pedro Gonçalves wrote:

...

Hi all

in the section 1.4.2.1 (Querying) the example defines a link with relation search and then shows a response message with the mime-type application/occi

shouldn't this be instead application/opensearchdescription+xml and return a opensearch description document (according to references in table 3)

regards

Pedro

On Nov 4, 2010, at 1:42 PM, Ralf Nyren wrote:

...

Hi all,

After hard work at OGF30 sorting out the remaining open issue of the OCCI model the Core specification document has received a major update.

There is some new stuff in there but mostly the document has been restructured and rewritten to be easier to read and understand.

We appreciate your comments. If no major issues are found the document will enter the public comment phase in 7 days, that is on November 11:th.

regards, Ralf _______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

All docs are available from the project SVN repository. Regarding HTTP rendering there are so many pending changes from OGF30 right now that what you will find there is quite far from what it will be. We really do appreciate your help and to get started I would recommend you to read the OCCI Core doc I sent to the list. That will give you the basis for understanding the HTTP rendering. When the HTTP rendering doc has caught up with all the latest stuff we would very much appreciate your comments on that document as well. regards, Ralf On Thu, 04 Nov 2010 22:00:13 +0100, Pedro Gonçalves wrote:

Hi ralf

thanks for your quick response how can we have access to a draft version of that document, or better yet, how can we help out in the writing ? .... i know probably i should be in the next concall :)

best regards

pedro

On Nov 4, 2010, at 9:07 PM, Ralf Nyren wrote:

...

The OCCI Core is one of three documents that together form the OCCI specification.

Discovery is described in Core and the specifics on finding supported resource types over HTTP is described in the HTTP rendering doc due to be made available shortly.

Hope that answers your question.

regards, Ralf

On Thu, 04 Nov 2010 21:03:34 +0100, Pedro Gonçalves wrote:

...

Hi Ralf

yes, you're right :) I was checking the differences between the file you sent and the file available i had here (the one from the web site most probably) and got the docs mixed up

All the search capabilities that were in that version were removed in the version you sent, is this right ?

regards

Pedro

On Nov 4, 2010, at 8:33 PM, Ralf Nyren wrote:

...

Hi Pedro,

Which document are you referring to? Could you provide a reference?

I am afraid you might have found some old document which is not valid with regard to the upcoming release of the OCCI spec.

regards, Ralf

On Thu, 04 Nov 2010 19:58:55 +0100, Pedro Gonçalves wrote:

...

Hi all

in the section 1.4.2.1 (Querying) the example defines a link with relation search and then shows a response message with the mime-type application/occi

shouldn't this be instead application/opensearchdescription+xml and return a opensearch description document (according to references in table 3)

regards

Pedro

On Nov 4, 2010, at 1:42 PM, Ralf Nyren wrote:

...

Hi all,

After hard work at OGF30 sorting out the remaining open issue of the OCCI model the Core specification document has received a major update.

There is some new stuff in there but mostly the document has been restructured and rewritten to be easier to read and understand.

We appreciate your comments. If no major issues are found the document will enter the public comment phase in 7 days, that is on November 11:th.

regards, Ralf _______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg

_______________________________________________ occi-wg mailing list occi-wg@ogf.org http://www.ogf.org/mailman/listinfo/occi-wg