[Moims-dai] FW: Comments on your blog post.

Wed Feb 24 15:42:26 UTC 2016

This is the email I sent to David Rosenthal at the same time as publishing
the post.

..David

From: David Giaretta [mailto:david at giaretta.org] 
Sent: 24 February 2016 15:41
To: David S. H. Rosenthal (dshr at stanford.edu) <dshr at stanford.edu>
Subject: Comments on your blog post.

Dear David

I've put the following text on the DPC forum about OAIS -
http://wiki.dpconline.org/index.php?title=Comments_on_David_Rosenthal%27s_%E
2%80%9CThe_case_for_a_revision_of_OAIS%E2%80%9D .  I hope it is a helpful
contribution to the discussion.

Regards

..David

Comments on "The case for a revision of OAIS"

From
<http://wiki.dpconline.org/index.php?title=The_case_for_a_revision_of_OAIS>
http://wiki.dpconline.org/index.php?title=The_case_for_a_revision_of_OAIS by
<http://wiki.dpconline.org/index.php?title=User:DRosenthal> David Rosenthal 

COMMENTS by David Giaretta on behalf of the working group responsible for
OAISrevision

The following contains comments to David Rosenthal's posting "The case for a
revision of OAIS" at
<http://wiki.dpconline.org/index.php?title=The_case_for_a_revision_of_OAIS>
http://wiki.dpconline.org/index.php?title=The_case_for_a_revision_of_OAIS.

The normal process for ISO standards involves a review after 5 years, which
means that OAIS is due for revision in 2017.   However, it is important to
understand OAIS before proposing revisions. As indicated in the comments
below, the case laid out is built on some fundamental misunderstandings of
the standard, in particular not realising that OAIS provides a reference
model as it very clearly states in the following way (see page 1-2): "This
reference model does not specify a design or an implementation.  Actual
implementations may group or break out functionality differently".  

The comments below (indented and in bold) seek to correct the statements in
the original post.

The official title of ISO 14721 is
<http://www.iso.org/iso/home/store/catalogue_ics/catalogue_detail_ics.htm?cs
number=57284> Reference Model for an Open Archival Information System
(OAIS). The role of a reference model is to provide abstract concepts and
terminology by means of which concrete systems can described and analysed. A
reference model is not of itself a standard against which concrete systems
can be assessed for conformance, that is the role of criteria based on these
concepts and terminology. In the case of ISO 14721 this role is performed by
<http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnu
mber=56510> ISO 16363 and its predecessor TRAC. The effectiveness of ISO
14721 must be judged by the effectiveness of its concepts and terminology in
describing concrete archival systems, and audits under TRAC and ISO 16363
provide a valuable opportunity to do so. 

COMMENT: The effectiveness of ISO 14721 is not best judged by how precisely
it is able to describe any particular archival implementation, but much more
on how widely it has been adopted to facilitate comparisons of archival
implementations and issues.  A reference model able to describe all
implementations in detail would be huge, extremely complex, and effectively
useless.

In July 2014 the  <http://www.clockss.org> CLOCKSS Archive was
<http://www.crl.edu/archiving-preservation/digital-archives/certification-an
d-assessment-digital-repositories/clockss-report> certified by CRL after a
rigorous audit against the TRAC criteria, the process for certification
under ISO 16363 not then being available. CLOCKSS gained an overall score
that equalled the previous best, and the first ever perfect score in the
"Technologies, Technical Infrastructure, Security" category. All
non-confidential documents submitted to the auditors are available
<http://documents.clockss.org> here. Four blog posts describe
<http://blog.dshr.org/2014/07/trac-certification-of-clockss-archive.html>
the certification,  <http://blog.dshr.org/2014/08/trac-audit-process.html>
the audit process,  <http://blog.dshr.org/2014/08/trac-audit-lessons.html>
the lessons learned, and
<http://blog.dshr.org/2014/08/trac-audit-do-it-yourself-demos.html> how to
run the demonstrations we showed the auditors. 

In general, basing the description of the CLOCKSS Archive on the ISO 16363
criteria, and thus on the concepts and terminology of ISO 14721 worked well.
Documents describing in detail the way significant OAIS concepts apply to
the CLOCKSS Archive are available
<http://documents.clockss.org/index.php/Main_Page#OAIS_Conformance_Documents
> here. But the  <http://blog.dshr.org/2014/08/trac-audit-lessons.html>
"lessons learned" blog post includes a section OAIS vs. CLOCKSS, reproduced
here: 

Writing the OAIS Conformance Documents made the mis-match between the theory
of the OAIS reference model and the practice of digital preservation in the
Web era, and in particular that of the CLOCKSS Archive, evident. The
conceptual mis-matches between the OAIS Reference Architecture, upon which
ISO 16363 is firmly based, and the CLOCKSS Archive's architecture fall into
four broad areas: 

*       CLOCKSS is a dark archive. Eventual readers of the archive's content
are unknown, and have no influence over when, whether and how content is
released from the archive. The OAIS concept of Designated Community is thus
difficult to apply.

*	COMMENT: This is a misunderstanding of the definition of Designated
Community. The Designated Community is defined (see page 1-11) by the
archive. The archive does not have to see into the future - they just have
to make it clear what they are doing. For example, are the CLOCKSS holdings
to be directly understandable to those who only understand Japanese? There
must be some criteria being employed, if only implicitly, and this should be
documented as the Designated Community - however narrow or broad it may be. 

The "eventual users" may or may not be part of that Designated Community,
and are not required to have any influence on when, whether and how content
is released. The archive will have some process for making these decisions
but OAIS does not cover those.

*       CLOCKSS ingests streams of content. Content ingested by crawling the
Web, as much of the CLOCKSS Archive's content is, is not pushed from the
content submitter to the archive but pulled by the archive from the
publisher. The publishers of academic journals emit a continual stream of
content; any division into units is imposed by the archive, not by the
publisher. The OAIS concept of Submission Information Package, (SIP) and the
relationship it envisages between the submitter and the archive, is
difficult to apply. The concept of Archival Information Package (AIP) also
has some detailed mis-matches, since to collect a stream an AIP must be
created before it contains any content, and subsequently accumulate content
over time instead of, as OAIS envisages, being wrapped around a pre-existing
collection of content at creation time.

*	COMMENT: The AIP is certainly defined by the archive. The SIP is a
general concept and the Producer is a role rather than a specific person or
organisation (see page 1-14). Someone or something is collecting the content
and submitting it to the archive. That person or system is playing the role
of the Producer. An individual actor can play multiple roles.

The AIP is not assumed to be created before there is any content. One could
talk about an AIP container or structure that is prepared before any
streaming is started.  Until it has all the required components it is not a
valid AIP.  The archive decides how to create the AIP. OAIS specifies the
kinds of information which must be logically contained in it.

*       CLOCKSS has a centralized organization but a distributed
implementation. Efforts are under way to reconcile the completely
centralized OAIS model with the
<http://purl.pt/24107/1/iPres2013_PDF/Creating%20a%20Framework%20for%20Apply
ing%20OAIS%20to%20Distributed%20Digital%20Preservation.pdf> reality of
distributed digital preservation, as for example in collaborations such as
the  <http://www.metaarchive.org/> MetaArchive and between the
<http://www.kb.dk/en/> Royal and University Library in Copenhagen and the
<http://library.au.dk/en/> library of the University of Aarhus. Although the
organization of the CLOCKSS Archive is centralized, serious digital archives
like CLOCKSS require a distributed implementation, if only to achieve
geographic redundancy. The OAIS model fails to deal with distribution even
at the implementation level, let alone at the organizational level.

*	COMMENT: OAIS is a Reference model - not an implementation model
(see page 1-2). There is nothing in the OAIS Reference model that would
preclude a distributed implementation of an OAIS (see pages 2-2, 4-3, 6-1
and 6-3). 

The Functional Model is a logical representation, not a design for a
centralised archive. OAIS does not specify how the various Functional
Entities are implemented or distributed. Standards for various aspects of
implementations would be better placed in a separate standard which follows
the OAIS Reference Model concepts and terminology.

Note for example that NASA's Planetary Data System (PDS) has been in
existence for many years and is a large distributed archive.   PDS staff had
no difficulty applying OAIS to the PDS. 

*       The CLOCKSS Archive contracts-out its operations. The CLOCKSS
Archive not-for-profit achieves its low cost of operations by contracting
them all out under two contracts with Stanford University. This enables many
costs to be shared with the other users of the LOCKSS technology, to the
benefit of both. The OAIS model fails to deal with organizational divisions
such as this.

*	COMMENT: Again the Functional Model does not specify how the
Functional Entities are implemented (see page 4-3).

Another mis-match between OAIS and web archiving would have been a problem
had CLOCKSS not been a dark archive. Access to archived Web content, via
<http://www.mementoweb.org/> Memento (RFC7089), direct link or text search,
occurs at the level of an individual URL. The OAIS concept of Dissemination
Information Package is difficult to apply to access of this kind; it says: 

In response to a request, the OAIS provides all or a part of an AIP to a
Consumer in the form of a Dissemination Information Package (DIP). The DIP
may also include collections of AIPs, and it may or may not have complete
PDI. The Packaging Information will necessarily be present in some form so
that the Consumer can clearly distinguish the information that was
requested. Depending on the dissemination media and Consumer requirements,
the Packaging Information may take various forms. 

Although there is obviously a lot of room for interpretation here, it does
not appear to cover the case where the Consumer requests, and the archive
delivers, a digital object (the headers and body of a URL) in exactly the
form it was ingested with no Packaging Information. This is what Consumers
of archived Web content want. It is true that, for example, Memento adds
header information to its response, but that information serves to point to
other archived digital objects, potentially in other archives, so it can't
be considered Packaging Information for the requested DIP. Fortunately for
us, the trigger process of the CLOCKSS Archive does deliver a package
containing many URLs, so it more closely matches the OAIS DIP concept. 

COMMENT: The DIP is a general concept and OAIS does not say how any
particular DIP is constructed or what it will contain. If/when required, an
archive must be able to provide the details of how the information in the
DIP links back to the original information which the archive ingested. Not
all DIPs need to contain that provenance. Packaging Information is defined
as: The information that is used to bind and identify the components of an
Information Package. If the response (the DIP) is sent using HTTP then the
fact that it is HTTP is part of the Packaging Information - normally taken
care of by the browser without the knowledge or intervention of the human
user.

Our experience in the TRAC audit of the CLOCKSS Archive reveals a number of
areas in which the concepts and terminology of ISO 14721 are inadequate to
describe a real, functioning system. There are two ways to react to this. If
you believe that ISO 14721 is not a reference model, but a definition of an
archival system, your response is to say the CLOCKSS and any other system
that cannot be described using only ISO 14721 concepts and terminology is
not an archival system. Whatever it is doing is not archiving. Over time, as
technology and the requirements of the marketplace evolve, the terminology
of ISO 14721 will describe fewer and fewer systems, so the field of
archiving will shrink to encompass only legacy systems. 

If, on the other hand, you believe that ISO 14721 is a reference model, your
response is to say that it needs updating with additional concepts and
terminology adequate to describe the systems that are doing archiving is the
sense in which that word is generally used. Our experience has identified a
number of areas in which updating is needed, and I hope to adress them in
detail in subsequent posts. I'm sure others have found other such areas, and
I hope they will address them in posts to this Wiki. Lets get to work to
ensure that a revised ISO 14721 matches the reality of current archival
systems. Once that is done, we will need to revise the standards based upon
it,
<http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnu
mber=56510> ISO 16363 and
<http://www.iso.org/iso/catalogue_detail.htm?csnumber=57950> ISO 16919. 

COMMENT: OAIS does not claim to be a reference manual to design archives. It
claims to:

-               provides a framework for the understanding and increased
awareness of archival concepts needed for Long Term digital information
preservation and access;

-               provides the concepts needed by non-archival organizations
to be effective participants in the preservation process;

-               provides a framework, including terminology and concepts,
for describing and comparing architectures and operations of existing and
future Archives;

-               provides a framework for describing and comparing different
Long Term Preservation strategies and techniques;

-               provides a basis for comparing the data models of digital
information preserved by Archives and for discussing how data models and the
underlying information may change over time;

-               provides a framework that may be expanded by other efforts
to cover Long Term Preservation of information that is NOT in digital form
(e.g., physical media and physical samples);

-               expands consensus on the elements and processes for Long
Term digital information preservation and access, and promotes a larger
market which vendors can support;

-               guides the identification and production of OAIS-related
standards.

The last point is particularly relevant here. No one standard can cover
everything. If it attempted to do so, then it would be too large to read and
would be out of date very quickly.

OAIS is an abstract standard which identified additional standards which
need to be developed. ISO16363 is an example of such an additional standard
and there are others which have been created or which are under development.
Other examples include the XFDU (ISO 13527:2010) standard which describes
one specific implementation of OAIS packages while the PAIS (ISO 20104:2015)
describes one possible implementation of the Producer-Archive Interface.

Surely the fundamental question when proposing revisions to OAIS is whether
the core, abstract, concepts need to be updated/corrected, or whether
additional standards are needed - or perhaps both. The OAIS terminology and
core, abstract, concepts are logically consistent and widely applicable.  

Taking distributed archives as an example, which are mentioned in the
original post as being beyond OAIS. We noted above that mapping PDS to OAIS
indicates that this is not true and the core concepts of OAIS do apply. It
may be sensible to create new standards for the implementation of
distributed archives, for example to define new ways to implement
federations or special storage systems. This would not in itself imply
changes to OAIS, ISO 16363, or ISO 16919.

As noted at the start, OAIS is scheduled for review/revision in 2017. It
will be important to collect ideas/comments/corrections but it is essential
to distinguish between changes in OAIS itself versus suggestions for new,
separate, standards.  Our comments indicate that the points made in the
original post fall in the latter category. However, if there are other new
considerations, or if you feel we didn't understand your post, we would be
happy to discuss this.  

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ccsds.org/pipermail/moims-dai/attachments/20160224/350ae536/attachment.html>