[Moims-dai] Updated Notes on Vint Cerf discussions on Digital Preservation with CCSDS-DAI

[David Giaretta]

Notes on Vint Cerf discussions on Digital Preservation with CCSDS-DAI

5th April 2016


Vint Cerf - points made


*       He described the value of having read, so far, the first 300 pages
of David Giaretta's book, Advanced Digital Preservation (
<http://www.springer.com/us/book/9783642168086>
http://www.springer.com/us/book/9783642168086).

*       He has some knowledge of our ISO work with OAIS Reference Model,
Trustworthy Digital Repositories, Audit and Certification of TDRs, etc.

*       He acknowledges that emulation is only one tool at best.

*       "Digital Vellum" focused primarily on relatively simple electronic
documents, software, and games. 

o   Mike Kearney has requested Vint Cerf's slides

o   Digital Vellum presentation: (
<https://ecolloq.gsfc.nasa.gov/Current/announce.cerf.html>
https://ecolloq.gsfc.nasa.gov/Current/announce.cerf.html)

o   Similar presentation at:  <https://www.youtube.com/watch?v=STeLOogWqWk>
https://www.youtube.com/watch?v=STeLOogWqWk

*       Compound digital objects present special problems.

*       Interoperability is a key requirement. Mashing up emulated data with
current data is problematic. He believes we need to demonstrate multiple
trustworthy digital repositories that can interoperate to move the adoption
of the standards forward.

*       He has no suggestions for a business model for keeping the data
usable for as long as it is needed.

*       Google is working in collaboration with the Computer History Museum
collecting software much like the NIST NSRL.


Discussions arising from Vint Cerf's comments - most after he left:


*       The OAIS concept of Representation Information provides a general
approach to interoperability. However more specific standards such as EAST
or DEDSL can be used to provide some of the detailed Representation
Information required.

*       The users of two repositories with the same Designated Community
should be able to understand information they exchange, at least for members
of that Designated Community.

*       Bruce noted that Vint seemed to be presuming that all digital
repositories want to be interoperable, which they may not. DG noted that
this could also be viewed as a way to increase the value of the information
and hence help to justify preservation.

*       A related but less specific aspect of interoperability would be for
us to develop a formal API definition for the OAIS Archival Information
Package (AIP)- based on the UML diagrams in OAIS. This would at least help
to ensure that repositories can exchange AIPs and extract the data,
Representation Information etc.

*       The document DAI is working on currently will help in guiding the
capture of information needed to exploit (including interoperating with
other data) data being created.

*       We could try to map DAI's work and existing books to a more
"protocol-like" view. For example, PAIS includes the definition of an
information package which can be sent from a Producer to an archive, and the
management processes required to organize this exchange.

*       Mike Kearney asked if there are any services from DAI which SANA
could host. David suggested that it would be possible for SANA to host an
authoritative Registry of Representation Information; he has an example
which is simply a webserver with a few very small scripts.

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ccsds.org/pipermail/moims-dai/attachments/20160422/ae73edfe/attachment.html>