[Moims-dai] [EXTERNAL] Re: Architecture Description

Tue Jun 18 10:42:54 UTC 2019

Hi Steve

My view is that as long as we say this is “OAIS” then the interoperability framework should reflect the OAIS Information Model (OAIS-IM). It could do other things but I believe that limiting ourselves to that will be useful enough. 

The clearest way to do that is to make it absolutely clear that the UML in the OAIS-IM is implemented. Therefore, for example, to get the Fixity Information (FI) of a particular Information Object (IO), 

The route between the IO and FI would be

*	Identify an Information Package (IP) that contains the IO 
*	Use the IP which we have identified to identify the PDI  (not that unless the IP is an AIP then there is no guarantee that it contains PDI)
*	Use the PDI which we have identified to get to the FI (again we can only guarantee that the PDI gets us to the FI if we are dealing with an AIP).

If the IO is being preserved then we can call it Content Information (CI). We can then reasonably say that the IP we want to deal with is an AIP, and hence must have PDI and FI etc.

If the framework interfaces allowed a user to perform those steps then I think everyone would agree that it reflects the OAIS-IM.

In writing the above I realise that I am focussed on what a user can request from an archive. There is the other side of this which is “get me an IP, when I have it on my computer I can then open it up using the Packaging Information and extract the various components”. Clearly we need to be able to do that since in the end whatever the archive send us will be some kind of IP. Maybe in the book we need to make a distinction between (1) the requests we make to the archive and (2) what we can do with the IPs we get back.

Of course our interfaces may provide “convenience methods” such as – given some CI then get me the associated FI. Internally to the archive there may be a database that allows one to do that pretty directly, but that depends on the archive. If there is no such pre-prepared database then the archive would perform the steps mentioned above internally. Whichever way the archive does it is, from my point of view, irrelevant to the interfaces we are defining.

So, in summary, I say that first and foremost our interface should clearly match the OAIS-IM UML. After we have done that then maybe we can think about what convenience methods we might add - although I would argue that we should not define any such convenience methods in the normative part of the book because that would be too open-ended, with no guidance from OAIS, and I would like the book to be completed as soon as possible.

Regards

..David

From: MOIMS-DAI <moims-dai-bounces at mailman.ccsds.org> On Behalf Of Hughes, John S (398B) via MOIMS-DAI
Sent: 13 June 2019 18:10
To: MOIMS-Data Archive Interoperability <moims-dai at mailman.ccsds.org>
Cc: Hughes, John S (398B) <john.s.hughes at jpl.nasa.gov>
Subject: Re: [Moims-dai] [EXTERNAL] Re: Architecture Description

Hi all,

This is the discussion I remember.  Here is a use case.

In the PDS4 Information Model we have the following hierarchy.

1.	Bundle – A Bundle contains one or more Collections (E.g., a Bundle is typically a package containing a data product collection for a single instrument together with one or more collections of information necessary for using the data. ) The PDS4 bundle is effectively an OAIS Information_Package.
2.	Collection -> A Collection contains one or more Data Products of a single type  (e.g., all the raw images from the Viking Orbiter 1 and 2 cameras)
3.	Data Products -> A Data Product containing data from a scientific observation. (e.g., one raw image from the Viking mission)  The Raw image and its PDS4 label is effectively an OAIS Information_Object.

Viewing PDS4 as an OAIS, if a user finds and requests a Viking Image from the PDS they will typically receive an Information Object, i.e., the raw image (Data Object) and its Representation_Information. If the user then requests the Fixity Information for the raw image, under the OAIS-IF, how should system get the Data_Object’s Fixity_Information?

According to the OAIS-IF Information Model the Fixity_Information is associated with the Preservation_Description_Information and the Preservation_Description_Information is associated with an Information_Package. This suggest that given the Viking Image Information_Object, the system will navigate through the Collection to the Bundle (an Information_Package), get the Preservation_Description_Information, get the  Fixity_Information, and then locate and return the Fixity_Information specific to the Viking Image Information_Object.

Another view is that the system knows that the Viking Image Information_Object is Content_Information and directly gets the associated Fixity_Information.

Thoughts?

Thanks,

Steve

From: MOIMS-DAI <moims-dai-bounces at mailman.ccsds.org <mailto:moims-dai-bounces at mailman.ccsds.org> > On Behalf Of david
Sent: Wednesday, June 12, 2019 11:59 AM
To: MOIMS-Data Archive Interoperability <moims-dai at mailman.ccsds.org <mailto:moims-dai at mailman.ccsds.org> >
Cc: Hughes, John S (398B) <john.s.hughes at jpl.nasa.gov <mailto:john.s.hughes at jpl.nasa.gov> >
Subject: [EXTERNAL] Re: [Moims-dai] Architecture Description

Yes, necessary, but not sufficient.

David

Sent from my Samsung Galaxy smartphone.

-------- Original message --------

From: Mark Conrad <mark.conrad at nara.gov <mailto:mark.conrad at nara.gov> > 

Date: 12/06/2019 19:15 (GMT+00:00) 

To: MOIMS-Data Archive Interoperability <moims-dai at mailman.ccsds.org <mailto:moims-dai at mailman.ccsds.org> > 

Cc: "Hughes, John S (398B)" <john.s.hughes at jpl.nasa.gov <mailto:john.s.hughes at jpl.nasa.gov> > 

Subject: Re: [Moims-dai] Architecture Description 

"Preservation Description Information (PDI): The information, which along with Representation Information, is necessary for adequate preservation of the Content Data Object"

On Wed, Jun 12, 2019 at 1:38 PM David Giaretta <david at giaretta.org <mailto:david at giaretta.org> > wrote:

In OAIS terms PDI may be part of any Information Package. If the Information Package is an AIP then by implication the Information Object it will probably be Content Information.

..David

From: MOIMS-DAI <moims-dai-bounces at mailman.ccsds.org <mailto:moims-dai-bounces at mailman.ccsds.org> > On Behalf Of Mark Conrad
Sent: 12 June 2019 17:51
To: MOIMS-Data Archive Interoperability <moims-dai at mailman.ccsds.org <mailto:moims-dai at mailman.ccsds.org> >
Cc: Hughes, John S (398B) <john.s.hughes at jpl.nasa.gov <mailto:john.s.hughes at jpl.nasa.gov> >
Subject: Re: [Moims-dai] Architecture Description

" I cannot understand why Content_Information (a sub-type of Information_Object) shows methods get/putPreservationDescriptionInformation.  That would normally indicate that these are additional methods in addition to those which Information_Object has – but why? "

Because PDI does not apply to all Information_Objects - only Content_Information.

Mark Conrad
NARA Information Services

Systems Engineering Division (IT)
The National Archives and Records Administration
Erma Ora Byrd Conference and Learning Center
Building 494, Room 225
610 State Route 956
Rocket Center, WV  26726

Phone: 304-726-7820
Fax: 304-726-7802
Email: mark.conrad at nara.gov <mailto:mark.conrad at nara.gov>  

On Tue, Jun 11, 2019 at 9:17 AM David Giaretta <david at giaretta.org <mailto:david at giaretta.org> > wrote:

Hi Steve

I can understand why get/setDataObject and get/putRepresentationInformation are methods in Information_Object but I don’t understand why get/putInformationObjectMetadata is there. Also don’t understand the difference between get/setPID and get/putIdentificationInformation. Also what is the difference between put and set?

I cannot understand why Content_Information (a sub-type of Information_Object) shows methods get/putPreservationDescriptionInformation.  That would normally indicate that these are additional methods in addition to those which Information_Object has – but why?

Is there are mix-up between Information_Object and Information_Package?

I suspect that part of my confusion arises from using the ontology rather than a UML tool to create the diagrams etc. but that would not explain my questions/comments above.

Regards

..David

From: MOIMS-DAI <moims-dai-bounces at mailman.ccsds.org <mailto:moims-dai-bounces at mailman.ccsds.org> > On Behalf Of Hughes, John S (398B) via MOIMS-DAI
Sent: 11 June 2019 03:16
To: MOIMS-Data Archive Interoperability <moims-dai at mailman.ccsds.org <mailto:moims-dai at mailman.ccsds.org> >
Cc: Hughes, John S (398B) <john.s.hughes at jpl.nasa.gov <mailto:john.s.hughes at jpl.nasa.gov> >
Subject: [Moims-dai] Architecture Description

Hi all,

Attached please find the Model Specification Document (html format) and two diagrams with a few updates to some core classes. In particular the DPI get/put methods have been moved to the Content_Information class as per recent discussions.

Thanks,

Steve

_______________________________________________
MOIMS-DAI mailing list
MOIMS-DAI at mailman.ccsds.org <mailto:MOIMS-DAI at mailman.ccsds.org> 
https://mailman.ccsds.org/cgi-bin/mailman/listinfo/moims-dai

_______________________________________________
MOIMS-DAI mailing list
MOIMS-DAI at mailman.ccsds.org <mailto:MOIMS-DAI at mailman.ccsds.org> 
https://mailman.ccsds.org/cgi-bin/mailman/listinfo/moims-dai

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ccsds.org/pipermail/moims-dai/attachments/20190618/0609653e/attachment-0001.html>