[Moims-dai] Desirable preservation ecosystem characteristics
david at giaretta.org
Wed Jul 11 12:48:57 UTC 2018
You have a good list there so I'd like to come at it from another direction.
The OED has two definitions for ecosystem (1) "A biological community of
interacting organisms and their physical environment" and (2) "(in general
use) a complex network or interconnected system."
While the second definition is general and can be aligned with technology
rather more easily, I like to bring in the first definition because
1. It brings in people
2. It leads one to think of things changing and evolving
3. It leads on to think about specialisations, death, food chains and
That leads one to say that in 50-100 years we should have:
* people aware in the requirements of preservation
* producers of information creating/recording
* data management plans linked to preservation i.e. capturing all the
types of "metadata" required along the way. This must include the semantic
information about the data/documents
* support systems should help as much as possible
* preservers of information being aware of:
* who they are preserving for i.e. the Designated Community and the
"metadata" likely to be required (OAIS Representation Information including
formats, semantics and software etc)
* potential uses of data including numerical data
* the costs of preservation
* funders (public and commercial) who are aware of:
* the potential value of information
* systems/processes which can cope with change and evolution:
* implementation independent designs
* standard, long lived / slowly evolving / backward compatible
interfaces to data and systems e.g. programming interfaces for
accessing/searching archives and for extracting components from Archival
* appraisal processes which help to decide if it is worth continuing
to preserve specific information
* ways to monitor the evolving needs of Designated Communities so that
repositories can ensure that the information they are preserving can be
* specialisations, death, food chains and dependencies
* each repository is probably good at certain things and not at
others, so it should be clear which is which. As things such as people,
technology and funding change, a repository which is good at one time may
not be good later on.
* effort can be shared e.g. one repository may look after a variety of
Representation Information to be used by other repositories
* any particular repository may cease to operate so it will need to be
ready to hand over its holdings to another repository
* information is linked so that as new information is created it may
affect the value of other previously existing information
From: MOIMS-DAI <moims-dai-bounces at mailman.ccsds.org> On Behalf Of Mike
Sent: 10 July 2018 21:36
To: 'MOIMS-Data Archive Interoperability' <moims-dai at mailman.ccsds.org>
Subject: [Moims-dai] Desirable preservation ecosystem characteristics
Vint Cerf is a preservation evangelist, and in discussions and presentations
he makes, he's considering adding this topic: Desirable preservation
ecosystem characteristics. The idea is that if we know what we want the
preservation environment to be 50 or 100 years from now, we will have be
better equipped to chart a course (at least the first steps) to get there.
Vint asked me my opinion on that, and I'm asking the DAI WG community.
I'm doing this off the top of my head right now, so here goes my first cut
Trustworthiness of archives:
* A majority of archives that are certified trustworthy.
* Broad community agreement on what constitutes trustworthiness and
the process for certifying it.
* Global prioritization of the need for certification of key archives.
Interoperability of archives:
* Well understood and widely available user interface systems that
function broadly for multi-archive access.
* Archive-to-Archive communications that are easily and quickly
* Distributed archives that allow transparent access to all preserved
data within their distributed "realm."
Access to archives (besides above):
* (From the RDA) Data sharing without boundaries. Social and
technical infrastructure that enables open sharing of data.
* Well-understood and easily-waiverable system of Intellectual
Property management - for example, a system to quickly secure rights to
download YouTube videos by national archivists.
* Effective archive security that does not interfere with authorized
* Storage capacity: A better, faster, cheaper More's Law.
* AI Access to preservation archives; AI helpers that can assist
archive users whether they are part of the Designated Community or not.
* Auto-preservation. Systems that automatically compile the *correct*
metadata (Representation Info) as object data is generated.
* AI systems that can generate warnings like "Uh oh. you're not
compiling the correct metadata!"
What's on your list?
Huntsville, Alabama, USA
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the MOIMS-DAI