DLF Fall Forum Notes

What is the going on in the future of DLF? (new members on p.6 of Rachel's presentation, p. 11 describes DLF's role compared to other conferences)

Continue to focus on the technological aspects of developing, sustaining, and federating digital libraries. DLF will promote standards, protocols, and best practices; it will evaluate and promulgate digital library projects and programs, and support digital architectures that most effectively allow for interoperable and extensible digital resources and tools.

Not attempt to build or develop projects in its name; rather, it will aggressively promote the digital production of its constituent institutional members, work with them to facilitate collaborations and cooperative efforts, and build a registry of projects and programs of wide interest and applicability.

Promote the concept of an international/ global digital library in which all major projects and future developments need to be understood as interrelated components. DLF will seek more partnerships with major libraries and institutions overseas.

work with the major funding agencies, private and public, to assist them in better understanding the interconnectedness of digital library development and to encourage them to fund projects in concert with one another to avoid redundancy and lost opportunities that can result from isolated, competitive funding schemes.

Build stronger ties to corporations, especially those investing heavily in digital resources and large-scale digital assets, and connecting the corporations to the communities that can best advise and complement the corporate investment.

Articulate qualities and characteristics of leadership in the coming decade, requisite for the changes and transformations ongoing.

http://www.clir.org/dlf/forums/fall2010/speakerslides.html

General Session Keynote highlighted promising projects:

Promising Projects
•WEST (distributed print journal repositories) - see attached word doc
•Digital Public Library of America (Harvard Robert Darnton)
•CRL print archives coordination
•HathiTrust
•2CUL
•OAPEN (EU academic publishing)
•CDL; Cal State

All of these projects also share an important tactical approach. They mitigate the concern of loss of individuality--loss of ‘brand’ in the common parlance--by keeping the level of tradition, history and idiosyncrasy of the institutions involved intact and building their interdependent alliances and collaborations within and among the services and programs that underpin research and teaching

Session #4 "Archiving" Digital Lives(green star)

Data Accessioner

library.duke.edu/uarchives/about/tools/data-accessioner.html

guiding principles - honor the unity of the collections, the needs of the donor for privacy, needs of the researcher, and the concern of the content with the digital context

JAVA based data recovery - GUI, mirror directory service, changes the last modified date to the original

First triage is to get the data off the hardware, forensic disk image - raw bit for bit, and make a logical copy

Change creation dates so copying is not good since it changes the context and relationships also the permissions change

Logical - capture deleted files

FRED- Forensic Recovery Embedded Device

Let the donor use the software

Hydra(green star)

p. 7/38 - Tailored applications and workflows for different content types, contexts and user interactions, A common repository infrastructure, Easily skinned UI, One body many heads

Technical Framework -Components
•Fedora provides a durable repository layer to support object management and persistence
•Solr, provides fast access to indexed information
•Blacklight, a Ruby on Rails pluginthat sits atop solrand provides faceted search & tailored views on objects

2010Technical Framework -Components
•Hydra Plugin, a Ruby on Rails library that provides create, update and delete actions against Fedora objects

•Services, providing discrete, reusable Lego bricks of functions (e.g., indexing, checksumming, MD transforms, etc.)

•Hydrangea, a web application that bundles all the RoRcomponents, user interaction and hooks to services into a single, adaptable package
•ActiveFedora, a Ruby gem for modeling, creating and managing Fedora objects
•Opinionated Metadata, tools for mapping domain specific vocabularies in application code to XML structures
•Solrizer, provides generic utilities for mapping content into a solrindex
•JavaScript Libraries, for Hydra-specific, rich user interactions with repository content

The Hydra project wiki: https://wiki.duraspace.org/display/hydra/The+Hydra+Project

VIRGO - UVA Manuscripts and images

Standford ETDs:https://lib.stanford.edu/sulair-news/electronic-theses-and-dissertations-available-searchworks-and-socrates-and-google-books

Stanford EEMS: http://lib.stanford.edu/everyday-electronic-materials-eems/about-eems

Code: http://github.com/projecthydra/hydrangea

Poster of interest -

From Trickle to Flood: An Extensible Approach to Large Scale Digitization of Manuscript Collections at UNC Chapel Hill Libraries

Reading Session #4: Digital Content and Infrastructure Needs of Research Faculty
Led by Jennifer Schaffner, Susan Kroll, Mackenzie Smith
Links: A Slice of Research Life available at http://www.oclc.org/research/publications/library/2010/2010-15.pdfhttp://www.oclc.org/research/publications/library/2010/2010-15.pdf.
Scholarly Information Practices from the Online Environment available at http://escholarship.org/uc/item/0kr8s78vhttp://escholarship.org/uc/item/0kr8s78v.

MacKenzie's objections to Kroll's Report:

  1. How valid is this research at your individual institution (each university is a snowflake...)
  2. There is tension but still trying to move ahead even though administration is telling you that you can't
  3. Senior Faculty behave differently than rank and file researchers
  4. Mission and marketing - how do we market what we are already doing

Other notes:

Coeus - MIT http://coeus.org/

Faculty want help with: structuring their notes and datasets, negotiating publication licenses, funding requirements, authority work, customizable personal webpages, and keeping up to date with their iterations.

Review Ithaka's faculty survey

www.rin.ac.uk - patterns of info use and exchange case studies

"pre-print is a more authentic representation of their work" - Kroll's study found that this is the feeling of faculty

Staffing and Workflow Solutions:

http://www.clir.org/dlf/forums/fall2010/15DLF-UIowa.pdf

Sustainability and Revenue Models
for Online Academic Resources
An Ithaka Report
May 2008
http://www.clir.org/dlf/forums/fall2010/sca_ithaka.pdf

Reimagining METS An Exploration

http://www.clir.org/dlf/forums/fall2010/METSNextGeneration.pdf

U.S. Government Printing Office’s Federal Digital System (FDsys)

http://www.clir.org/dlf/forums/fall2010/FDsysFactSheet.pdf

Closing the Digital Curation Gap

http://www.clir.org/dlf/forums/fall2010/CDCGflyer.pdf

Twitter summary:http://wiffiti.com/screens/41933summarized: http://summarizr.labs.eduserv.org.uk/?hashtag=dlf2010

Themes of DLF wrap-up:

Digital Library Federation

  • To focus on the technological aspects of developing, sustaining, and federating digital libraries:
  • Promote standards, protocols, and best practices
  • Evaluate and promulgate digital library projects and programs
  • Support digital architectures that most effectively allow for interoperable and extensible digital resources and tools
  • Facilitate relevant formal and informal education opportunities

Ideas covered -

Ideas Expressed
• Key Performance Indicators‐ develop a community of practice around assessment. #kpi
• Economic Case Studies #econcase
• Develop models for determining value, ROI, return on mission #valueassessment
• Small forum for digital archivists, working w/ electronic records in manuscript repositories, technologically neutral
solutions (esp microservices) #digarchforum
• Hot issues group, identify hot issues and share policies/plans/practices (one first one: data management plans) #dlfhot
• Microservices as way to migrate project‐>program (hackathon or similar) #mshack
• Facilitate cross domain collaboration (however) #collab
• Action Learning Groups for project managers, #dlfalg
• Talk about failures and brainstorm, #failstorm
• Developers talk and managers really listen #dlfdevtalk
• Emerging DL leaders #dlfprofdev
• Training session/workshop on software testing, good development practices, #devcraft
• Methods for counting DL objects in ARL etc environment #dlfstat
• 1/mo conf call with rotating convenor. 5 people talk for 10 mins on what they’re doing, #confcall
• Exhibit systems, how to disseminate content #accesssystems
• (make dlf.org resolve) lots better web site for DLF, #dlfweb
• tools dashboard, open source and homegrown to see what other inst’s are using #tools
• find way to allow more people to go to more sessions at the forum – don’t overlap so much. Record sessions for later
viewing. Collaborative scheduling/voting. #DLFPPC
• Supporting research methods for specific research communities (dig humanities, science data) #dlfdh, #dlfscience
• Exploring barriers to collaboration (strategic, organizational, technical) – how to convince administrators #collab
• How to deal with artificial grant cycles – allow to innovate within that structure #grantsuccess

Project Managers Meeting

Agile Project Management:

http://agilemanifesto.org/

working software, customer collaborations, responding to change, individuals and interactions

Characteristics of agile:

plan as you go, feature-breakdown structure, user stories, release plan, story boards, deliver as you go, learn every iteration, adapt every thing, manage team

"Information radiators"

empowering and trusting people, providing a constant feedback

change happens frequently

requirements are not well-defined

new technology or public domain

customer isn't sure of what is desired

Delphine uses agile/scrum at Penn State - calls some members of her team "ingesters"

NCSU - refer to powerpoint for books to read

resources:

http://agileforall.com/blog

blog.mountaingoatssoftware.com

Books: http://library.mit.edu/item/001270790

http://library.books24x7.com.libproxy.mit.edu/toc.asp?site=bbbga&bookid=8392

notes from Emily's presentation:

Product backlog is supposed to be prioritized. But library staff have a difficult time prioritizing issues against each other. They work best talking generally about what they think should be part of a given iteration. Then we negotiate with them about what we think we can accomplish. The IT product manager plays a large role in determining final priorities here.

Testing – “the weakest link”. For smaller projects w/o massive codebase, when is it faster and better to automate testing? No QA experts, so try to utilize product teams with regular updates throughout the cycle. In the end, developers and IT product managers have to do first pass at QA testing. Have not done a good job defining acceptance tests for each piece of functionality.

Our group is responsible both for real-time support and development. How to handle? Working toward documenting time spent on real-time support emergencies in JIRA and track how much effort spent on that vs. development. This feeds our estimation of how many hours are available for each person for each cycle; assume some time goes to support.

Timeboxing of efforts within a 6-week development cycle helps keep us from getting bogged down in large implementations that go on for months and months and keep team from working on other projects. Instead, focus on the outset on a simplified set of work that can be accomplished in 1 cycle and then re-evaluate institutional priorities at the end of that cycle. This reduces feeling of black box stall on any given project.

 Users hate seeing nothing. Using these methods, more people have seen some results faster. Incremental improvements that happen regularly produce much more positive feedback and energy both within the development team and out to our customers in the library.

Adapt to changing priorities. For example, move our EZProxy Administrator tool to the top of the list as needed for Shibboleth integration. We would not have been able to do this without causing problems if we had been already committed to months of development and a huge feature list for ReservesDirect, our course reserves system. Constant re-evaluation also always priorities within a project to change – since we haven’t committed beyond the current cycle this is easy to accommodate.

Laura at Emory Library:

  • Not have a set plan at the outset
  • Testing in smaller chunks
  • Not having to shield the developers from the stakeholders

p12/27 of pdf - "data dictionary" great idea - we should have that in DSpace/Dome Operations

Greenhopper & JIRA
Google docs & JIRA

Jira Studio...services in the cloud (http://www.atlassian.com/hosted/studio/). Cost is reasonable (free 30 day trial) and nobody has to be convinced to install the tools. It includes Greenhopper, Jira and Confluence...other options can be added.

My development team embraces agile, yet the management team prefers a plan-driven methodology. Resources I've found useful are _Agile Estimating and Planning_ <http://bit.ly/dyPv1S> and _The Software Project Manager's Bridge to Agility_ <http://bit.ly/b2RWGy>.

 Happy reading, Barrie Howard, Project Manager, CACI-ISS, Inc., 202-507-0387

The Art of Project Management (http://oreilly.com/catalog/9780596007867/), has been a useful resource for us as well--not in the "template" sense, but as an orientation, especially for new project managers.

ProjectsatWork.com

http://www.projectsatwork.com/content/Articles/256677.cfm

http://www.jamasoftware.com/contour/

Zepheira uses Redmine for their ticketing system for the Recollection development.  They call their system "Foundry," but it's Redmine underneath.

Starr, Joan. Basic Time Estimation (2010) http://www.slideshare.net/joanstarr/basic-time-estimation-3127687

Wrona, Vicki. "Project Risk Management: A Practical and Effective Approach." http://www.continuitycentral.com/feature0547.htm

Rick Johnson and Banu Lakshminarayanan of Notre Dame presented at Code4Lib Midwest ( http://wiki.code4lib.org/index.php/Midwest ) about their use of Taverna ( http://www.taverna.org.uk/ ) to build digitization workflows.  It reminded me of SEASR ( http://seasr.org/ ), which has been promoted for building text processing workflows.

  • No labels