Page 1
Open @ FAO
Online Information Conference
November 21, 2012
[email protected] Knowledge & Information Officer
Management Officer, David Lubin Memorial Library
Steve Katz. (Twitter: @SteveK1958)
Chief, Knowledge Management and Library Services
Food and Agriculture Organization of the United Nations
Page 2
• FAO is a specialized agency of
the United Nations
• Millennium Goals. Goal 1
Eradicate extreme poverty and
hunger
• ≥190 Member Countries
• HQs in Rome, plus ≥5000 staff
in:
• 5 Regional Offices
• 11 Sub-regional Offices
• 82 FAO Representations
• 36 country or liaison offices
• Increasing decentralization
Food and Agriculture Organization of the United Nations
(FAO)
Page 3
1996 – First appearance of FAO website www.fao.org
and, SGML Repository Proposal; FAOSTAT on-line
1997 – Corporate Document Repository (XML
Compatible)
2003 – Document Repository (PDF)
2007 – Open Archive Proposal (Fedora Commons)
2013(?) – openarchive.fao.org
Open @ FAO : A Bit of History
Page 4
General Advantages of OS Free/low initial
investment
Low administration
costs, No negotiation
with vendors, no legal
clearances
Customizable
Standards
Enhanced
interoperability
Philosophy of OS-
champions at FAO
Stability against
vendor withdrawing a
product or support
Vulnerability to
security is diminished
by quick and large
community response
Page 5
Potential Risks with OS Time and money wasted
from lack of direction
Funding insufficient for
unforeseen costs
Staff capacity insufficient
Staff time monopolized with
OS, and other work suffers
Trained staff are lost
Lack of collaboration leads
to narrow system produced
--proprietary systems can
be generic but also well-
rounded
Threshold of patience for
community process expires
Weak governance
Documentation poor—
internal & community
Potential for forking
Potential for vulnerabilities
being targeted & attacked
Low
credibility/trust, reputational
risk, legal exposure?
Development of Koha
serials module
Page 7
Consequence of the risks
Page 8
Selecting a System:
Goals, Objectives & Requirements
Technical
Open Source
Scalable
Extensible
Modular
Interoperable with other FAO
systems, e.g. Data.fao.org &
departmental websites
Administrative
De-duplication of work (e-pubs &
Library catalogue)
Workflow & Content control
Preservation
OAIS model
Some native basic preservation
actions
Discovery & Access:
Multi-lingual interface
Compliant with OAI-PMH & other
standards (external and internal
interoperability)
mobile access & delivery:
developing countries
Accessible across old
platforms, operating systems and
software versions: developing
countries
Full-text and advanced search
with Agrovoc subject headings
and metadata field searching
Advanced & full-text searching
Page 9
Fedora’s Features: Advantages
Open source: Fedora Commons & Creative Commons & open standards
Modular: flexible & extensible Ingests, stores, and manages digital content of any type
Metadata in any format can be managed and maintained
Uses a variety of front-and-back ends for user ease
Access: full-text search (Gsearch); multilingual interfaces; mobile delivery
Disseminator allows specific types of content to express itself as
needed, egg zoom in on photos; can create specific semantics for
books, images, maps, texts, etc.
Can repurpose content for specific context
Interoperable: data accessed by Web APIs
OAIS & OAI-PMH
Scaleable—accomodates millions of objects
Some Preservation actions included—Rebuilder Utility, checksums; virus
checks; format verification & validation
Page 10
Disadvantages of Fedora
Challenging, difficult & complex to implement.
Requires significantly more staff resources to
customize than other OS repository software
(e.g.DSpace)
Necessary staff development process is slow
Requires significant financial investment to
deploy:
Requires significantly more finances to customize
than other OS repository software (e.g. DSpace)
Contracting out expensive
Limited pool of contractors with Fedora expertise
Page 11
Fedora Commons Community
Over 300 registered Fedora Repositories
Active in development and archiving and information sustainability.
Reputable and prestigious partners, allies & sponsors
Andrew W. Mellon Foundation, Arrow, Cornell University
Information Science, DSpace, eSciDoc, FIZ Karlsruhe, Johns
Hopkins University, MediaShelf, Gordon and Betty Moore
Foundation, Mulgara, NSDL, OhioLink, PLOS One, Rutgers, Sun
Microsystems,DTU Technical University of Denmark, Topaz and
VNS
Community provides good documentation: DuraSpace provides
documentation, wiki, tutorials, brochures, newsletter, repository of
users, logos and more (www.fedora-commons.org)
Page 12
FAO Open Archive Working
System Interface
EPM for departmental submissions
Customized to reflect FAO departmental-
specific workflows—e.g. FI has 3 levels of
control; AG 1 level of approval
Data producers take & retain responsibility
for content
Basic metadata at working system level
Page 14
DLML Cataloguing Module
CM: Enriched metadata added by
cataloguing module
OA will merge the cataloguing functions
now in an CDS/isis system (late 1960s)—
also OS—but no longer supported and the
current CDR
Page 16
FAO Open Archive User
Interface
Page 20
FAO’s experience
Benefits
Repository completely tailored
to FAO’s particular needs
Inspiration to begin Digital
Preservation
Developing in-house expertise
Contributions to Fedora
community
Low administrative overhead
No negotiations with vendors
Standards for exchange
Challenges
Staff has moved to another
department but project
remained—expertise walked
away, tacit knowledge lost
Documentation of procedures
poor—high-level
documentation
Strong capacity required
More expensive than
anticipated.
Working silos counter
productive to process
Lengthy development time— 5
years + ?
Page 21
FAO’s OS Lessons Learned. OS projects require: Adequate staff capacity
Slow & expensive process to develop staff capacity
Document every decision , not only general descriptions of what was done
Adequate funding—can be expensive
Hidden costs: staff training, time spent in research, hardware, costs of external
maintenance and hosting, and if so consider if there are any costs for repatriating
your own data.
Clarity of concrete goals and objects
Know the difference between what you need and what you want
Know your limits and tolerances—what can you give up?
Consider what the software can do and not do
Be realistic about the minimum required to launch repository
Articulate goals before you start and stick to them, if possible, to avoid wasting
effort and resource
Page 22
Still More OS Lessons Learned.
The Organizational experience OS projects require:
Inter-departmental collaboration to avoid work
silos. One skill set focuses only one aspect. If
only IT is involved, the access end, user
interface, and policies will be neglected
Solution: Collaborate. Cross-fertilize across work units and staff skill sets for
development, implementation and administration→
develop whole system uniformly.
Page 23
A mixed bag: Advantages &
Disadvantages & vice-versa Open source free/ hidden costs can be large
Independence from vendors/community support may be
limited and/or slow
Commercial vendors may drop a software / community
may be more stable & invested
Security—open code means hacking/more solutions to
hacking
No real marketing by the community; no vested interest /
no one really accountable
FAO remains a proponent of OS
Page 24
Thank You!
Questions? Comments? Please get in
touch with us:
[email protected]
and/or
[email protected]