[Moims-dai] Doc Tree
jonathan.tilbury at preservica.com
Wed Apr 27 11:19:11 UTC 2022
We have been watching these discussions as they continue. I totally agree that any standards defined should be compatible with commercial applications otherwise they won't be adopted - to ignore the suppliers would be like defining a new SQL standard without asking Oracle or Microsoft.
We would be happy to engage with the suggestions when they are closer to completion. Our APIs and interfaces are publicly available and may be a good starting point as well as our "OPEX" ingest and export formats.
Let us know how we can contribute.
From: MOIMS-DAI <moims-dai-bounces at mailman.ccsds.org> On Behalf Of David Giaretta via MOIMS-DAI
Sent: 27 April 2022 10:08
To: 'MOIMS-Data Archive Interoperability' <moims-dai at mailman.ccsds.org>
Cc: david at giaretta.org; kearneysolutions at gmail.com
Subject: Re: [Moims-dai] Doc Tree
A great improvement but just a couple of comments from the point of view of simplification and estimates of effort needed.
I think the only change to the diagram is that we would just need one Generic Adapter book. The idea was that any system could act as an information source (archive) or sink (user) at various points e.g. an Archive needs to ingest information i.e. act as a sink.
In terms of work needed - the first Specific Adapter book could be rather simple "fairly general" Specific Adapter which could be used quite widely for archives which currently just act essentially as file/web servers i.e. "File Server Specific Adapter".
My hope was that the Registry and the Switchboard could really be specialised, fairly simple archives which could be quite short, based on the "File Server Specific Adapter".
The thing we need to add would be how to deal with Representation Information in a realistic way - a "Representation Information Interfaces" book. In principle we could also create a book to cover other aspects of the Information Model in more detail than is currently in Steve's diagrams e.g. interfaces for Provenance, Context, Fixity, Reference and Access Rights. These could be very minimal with appendices pointing at various published implementations e.g. OPM for Provenance.
I've been looking at some of the commercial and free archival systems which are out there, for example Archivematica and others. We would try to ensure that out interfaces can accommodate these, as well as home grown systems (PDS? etc).
From: MOIMS-DAI <moims-dai-bounces at mailman.ccsds.org<mailto:moims-dai-bounces at mailman.ccsds.org>> On Behalf Of Mike Kearney via MOIMS-DAI
Sent: 26 April 2022 22:53
To: 'MOIMS-Data Archive Interoperability' <moims-dai at mailman.ccsds.org<mailto:moims-dai at mailman.ccsds.org>>
Cc: kearneysolutions at gmail.com<mailto:kearneysolutions at gmail.com>
Subject: [Moims-dai] Doc Tree
I had the action after the telecon today to update the Doc Tree for the Green book, factoring in the new architecture. The Doc Tree version in the DropBox<https://www.dropbox.com/s/0wz029riy0e8cfi/2021-01-18%20OAIS-IF%20Document%20Tree.pptx?dl=0> was not quite up to date to the version that was in the Green Book, but that didn't matter because we revamped all that anyway. David, you might check whether there is something on the Green Book version that you didn't want to lose. Also, David, in the drop box is a file "OAIS diagrams<https://www.dropbox.com/s/9ahnioc5duq72en/OAIS-IF-diagrams.pptx?dl=0>" which also had these drawings and several others that are out-of-date. You might want to clean that up (just delete obsolete drawings) to avoid confusion. I recommend we keep these Doc Tree pictures in this separate Doc Tree document, because they're important and get lost in the other miscellaneous drawings.
The original file in the Dropbox (link above) actually had three pages. Firstly, the "DAI WG Document Tree", secondly the "Relationship of CCSDS DAI WG standards, and thirdly, the "Specific Standards to be Developed (the blocky diagram with the complicated keyholes). I updated the first two, and just deleted that third drawing as obsolete and too difficult to update to the new architecture. I added a new third page that is from the EWA presentation, "OAIS Architecture Concept - Functions," that is the driving drawing that shows the current architecture. David, in the EWA presentation, there is also the "OAIS Architecture Concept - Resources" drawing that is the same block diagram with resource description. You might consider whether you want that in the green book, and in this PPT as well.
So, here's what changed.
* John suggested adding the code to show "published, in-work, and future" documents. I copied it from chart 2 and added it to chart 1. It does clutter it up a bit, but if we need the legend on both charts 1 and 2, this is the best I could do.
* For chart 1, all the blue books were redone, adapters added (IAW the new architecture), and the Archive Abstraction Layer (AAL) was deleted.
* I deleted PAIP and CAIP from the first two drawings, as we discussed, until we're more certain that we're going to do those. Unless they're needed in the new architecture, driven by needs of the adapters, I think we leave those out of our future plans.
* I kept CAIS as a future document in CHART 2, but combined it with PAIS in Chart 1 for legibility. Also, it occurs to me that we might put CAIS in the same book as PAIS anyway. Future question when we address CAIS on our development schedule.
* Put all the blue books directly under the ADD. With the AAL gone, it's not clear that any of these books are subordinate to anything except the ADD.
* As we discussed today, there may be only one generic adapter, but I split it into two for User and Archive just to match our architecture (Chart 3). We should discuss it and if we're really targeting one generic adapter for use by both users (producers and consumers) and archives, I should change this to reflect that. A "Generic Adapters" Blue Book.
* Steve, there are major implications with that. The ADD needs to "call into existence" all the Blue Books, including existing PAIS, EAST, DESDL, PVL and XFDU. Also the Switchboard and Registry. I imagine you'll need help from David and John to put near-term placeholders and long-term content for them in the ADD.
* Note that this is an old chart that David initiated to show the flow between Producers, Archives and Consumers. That's the purpose of the chart.
* Revamped the blue book flow to show all the adapters, deleted the AAL.
* Only showed one Archive Generic Adapter and explained (with asterisk) that there's one for each archive instantiation. Didn't address whether one Blue Book might cover all Archives, or individual archives might adapt their own instantiation of an Archive Generic Adapter. Or that the User and Archive generic adapters might be the same. (Effectively the Abstraction Layer reincarnate)
* Showed multiple examples of Archive Specific Adapters and explained them in the note.
* Left the RepInfo standards and XFDU standard "floating" without connection to this flow, per earlier note from John.
Comments, recommended changes are of course welcome. I think as it is, this is OK to plug into the Green Book and also to show to Peter for the Doc Tree discussion. Steve, note that the Blue Book you'll show him is not yet lined up with this Doc Tree. I guess the question for him is whether we should go to the trouble of revamping the content of the Blue Book per this doc tree, and if so, what parts of the Blue Book get farmed out to which adapters? Or does he suggest something else?
If updates are needed, we should update in this file on the DropBox, then copy into the Green Book or presentations or wherever needed. This file should be the master for the Doc Tree. Except maybe chart 3. It's "master" might need to be in a separate file, but not sure what.
That's it for today.
Huntsville, Alabama, USA
KearneySolutions at gmail.com<mailto:KearneySolutions at gmail.com>
MikeK at spacestandards.org<mailto:MikeK at spacestandards.org>
Member of Technical Staff
Space Infrastructure Foundation<http://spacestandards.org/>
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the MOIMS-DAI