Empfohlen
Empfohlen
Weitere ähnliche Inhalte
Andere mochten auch
Andere mochten auch (20)
Ähnlich wie Sustaining ArchivesSpace
Ähnlich wie Sustaining ArchivesSpace (20)
Mehr von DLFCLIR
Mehr von DLFCLIR (13)
Kürzlich hochgeladen
Kürzlich hochgeladen (20)
Sustaining ArchivesSpace
- 1. SUSTAINING ARCHIVESSPACE DLF Forum Denver, CO November 4, 2012 David Millman, Mark Matienzo, Katherine Kott
- 2. Topics • Background • Technology & Software Development • Technical framework • Process • Alpha preview • Interoperability • Organizational Home • Governance & Sponsorship • Structure • Rollout • Relationships with other organizations (e.g. SAA)
- 3. Review: ArchivesSpace is… • A project funded by the Andrew W Mellon Foundation for building a “next generation” archives management tool • A community effort, led by the libraries of New York University, the University of Illinois at Urbana-Champaign, and the University of California, San Diego • The development of a sustainable partnership and community to support and foster the growth of the software
- 4. ArchivesSpace Groups and Roles • Steering Group: Luc Declerck, David Millman, Beth Sandore, with Robin Dale (LYRASIS) • Stakeholder Team: Mark Matienzo (Technical Architect/Product Owner), Chris Prom, Kyle Rimkus, Scott Schwartz, Brad Westbrook (Archivists/Business Analysts), Joe Pawletko (Technical Advisor) • Hudson Molonglo Development Team: James Bullen, Mark Triggs, Payten Giles, Brian Hoffman • Testing and Release Team: Brian Tingle & Adrian Turner (CDL) with HM, Mark Matienzo, and Katherine Kott (Development Manager) • Migration Team: Joe Pawletko (NYU), Nathan Stevens (AT), Chris Prom (Archon) plus Stakeholders and UIUC programmers
- 5. Development partner • Hudson Molonglo Pty Ltd • An IT consulting firm with staff in both the US (Hudson) and Australia (Molonglo) • Deep experience with digital library and archival management systems • Selected through a formal RFP process • Team Members • James Bullen (lead) • Mark Triggs (back end) • Payten Giles (UI) • Brian Hoffman (data model, import/export)
- 6. ArchivesSpace Product Vision • A best of breed, open source archives management system, supporting core archival functions and standards … • … that can both scale up and scale down … • … that is flexible, efficient, and easy to use, maintain, and administer … • … that is extensible and can interoperate easily with other applications and systems … • … with a healthy ecosystem of users, developers, and partners dedicated to sustaining the product in the open source environment.
- 7. Technical Framework • Separate user interface & back end • REST service interface & documented API • Frameworks for rapid development • Sinatra • Ruby on Rails • Twitter Bootstrap • Emphasis on • Ease of deployment • Openness • Performance & scalability • Selected technologies • JRuby • Java virtual machine • Lucene and Solr • Relational databases • Support for multitenancy
- 8. Development Plan: July 2012-June 2013 Repository API, record All required management Full Basic front & edit, Workflow, import/export, workflow, user application back end authentication, search, EAD UI templating management testing, beta infrastructure general import/export & workflow, release (July/Aug) import/export (Nov/Dec) customization reporting (May/June) (Sept/Oct) (Jan/Feb) (Mar/Apr)
- 9. Test Driven Development • Development in HM environment • Continuous integration (Jenkins) • California Digital Library support for testing • Cloud-based deployment and testing (AWS, Travis, Saucelabs) • Automated build and deployment testing • Interactive testing by ArchivesSpace Team • Broader-based testing by community members coming soon • At launch, application will be supported and tested by LYRASIS, the organizational home (more about that later)
- 13. Alpha Releases • 0.2 (October 29, 2012) • Instance, container, and location management • Digital objects (basic implementation) • Event records • Notes (both single and multipart) • Deaccession records • Basic EAD export • 0.1.3 (October 15, 2012) • Rights management records • Basic EAD import • 0.1.2 (October 1, 2012) • Complex extent statements • External document/link associations • Local authorization • 0.1.1 (September 17, 2012) • Agents • 0.1 (September 3, 2012) • Archival description (resource and resource component records) • Accessions • Subjects/controlled vocabularies • Local authentication
- 14. Accessions
- 18. Deaccessions
- 19. Subjects
- 20. Agents
- 23. Instance and Location Management
- 24. Event Records
- 25. 25 Interoperability • User needs • Integration with repositories and digital preservation systems • Integration with discovery and patron request systems • Funders’ interest in leveraging investments
- 26. ArchivesSpace Organizational Home • LYRASIS • The largest membership organization serving libraries and information professionals in the United States • Known for its “local touch”—fostering collaboration and cooperation among members
- 27. 27 Organizational Home: LYRASIS • Sustainability through membership • Hosted solutions • Help desk services • Software upgrades • Training
- 28. 28 Governance Process • Membership Model • Sliding scale membership fees • Structured to include representation from a variety of organization types and sizes • Governance Structure • Board • User Advisory Council • Technical Advisory Council • Boards start work July 2013
- 29. Opportunities for Involvement • Charter memberships • Available through April 2013 • “Regular” membership • Available January 2014 • Sponsorship • For organizations that do not plan to use the software but would like to be involved in the community
- 30. 30 Professional Organizations • Build on SAA relationship • Training • Updates through RoundTable and conference presentations • Forge relationships with other organizations • ALA • AAM • ???
- 31. For more information and to participate • ArchivesSpace website • http://www.archivesspace.org/ • ArchivesSpace Google Group • http://groups.google.com/group/archivesspace/ • Github repositories • Development: https://github.com/hudmol/archivesspace • Releases: https://github.com/archivesspace/archivesspace • Volunteer for beta testing • Email mark.matienzo@nyu.edu
Hinweis der Redaktion
- These are the topics we’re planning to cover—with a focus on sustaining both the open source software we’re developing and the ArchivesSpace community.
- Launched by a project funded by the Mellon Foundation, ArchivesSpace has several facets, including collaboration among the three partner universities, engagement with the community, and specific plans for ongoing support for both the software and the community through a membership model.
- Even within the grant funded project, there are many moving parts. The Steering Group is heavily involved in plans for governance, membership and transition to LYRASIS, the ArchivesSpace “organizational home”. Formerly known as the technical group, the stakeholder team supports Mark in prioritizing development and testing alpha instances of the software. They are also liaisons to the broader community of archivists.The California Digital Library has set up a test environment. (More detail about that later in the presentation.) CDL will also be coordinating end user beta testing, and data migration testing to support archives that will be moving from AT or Archon to ArchivesSpace soon.
- One of the advantages of the project is that development is being done by an outside vendor. This means we are not competing with other projects for resources within the partner institutions.
- In an agile development process, it is important to start with a shared vision for the product. This vision is used to remind both stakeholders and developers of the overall purpose of the project to help set development priorities.The application must be compelling to both users of existing applications like Archon and Archivists’ Toolkit, as well as those archivists and institutions without an existing archives management or access system in place. The product must be more than just an incremental improvement over current applications and be recognized as a leader in its class.Archival institutions range in size in terms of both staff and size of collections. The application must be able to scale down to single-user deployments and scale up to large, multi-tenant and enterprise installations, and must work equally well for collections regardless of extent or level of arrangement.The application should be not only a joy to use, but also straightforward to maintain and administer regardless of institutional size or setting. The application must be configurable to allow a variety of institutions to be able to adapt it to their needs and allow for customization of the public and staff user interfaces. The application should be easily deployable on multiple platforms and operating systems.The application must be extensible using a straight forward plugin architecture and have well-defined and well-documented APIs to allow for the development of new and improved functionality. APIs and service endpoints of the application should also be designed to allow for integration with other systems and software with a minimum of programming complexity.Finally, the application must keep all of its potential communities in mind throughout its lifecycle and focus on features that provide value and meet their needs across them as much as possible. The sustainability of the application depends on a governance structure within which users, developers, and service providers can participate.
- ArchivesSpace is being built as with separated backend and frontend applications that interact over a REST-based API. Both applications are being written using Jruby, a version of Ruby that runs on the Java virtual machine. The backend, written using the Sinatra framework, interacts directly with a relational database, and we have targeted MySQL and Apache Derby as the two primary databases. The use of Derby allows us to package an embedded database for ease of testing and to simplify deployment for very small scale and single user instances of the application. The frontend is written as a Ruby on Rails application using Twitter’s Bootstrap user interface framework.
- Process using Scrum development methodology, with two week sprints and eight week cycles. Regular check ins with development team. As product owner, Mark tests and provides feedback to developers. We are currently on target with development, and this cycle will focus on user interaction and application workflow, providing search capabilities, and and import and export for EAD instances.
- We have committed to test-driven development. Testing frameworks include RSpec and Selenium. HM maintains their own instance of the Jenkins CI server, which runs all unit, integration and functional tests. In discussions with HM, we agreed to have them aim for 80% test coverage, and they’ve been able to provide 93.98% coverage in the backend application and common code shared across the backend and frontend. Since HM provides a test instance of the application to the team, we can quickly test new features as they are in progress before the end of the sprint. CDL is assisting with testing by a test instance of the application deployed on Amazon Web Services using EC2 and RDS. The stable branch of the application is tested using the Travis continuous integration service, which integrates well with Github-hosted repositories. We also use Saucelabs to run our Selenium test, and it provides us with a video walkthrough of the testing that can also easily function as a feature walkthrough. At the end of every cycle, we have also been encouraging feedback from the rest of the ArchivesSpace team and recently have rolled out a semi-formal test script for them to investigate the newly released functionality included in a given version of the application. CDL will be coordinating usertesting with archival community and we are very interested to get a broad base of interested users to assist us with that.
- Using Pivotal Tracker to manage user stories. Backlog requires continuous grooming by product owner and development team – allows us to identify redundancies and issues that need clarification. Pivotal Tracker is an excellent tool for Scrum because it automatically averages the development team’s point tallies for delivered and accepted features to accurately calculate their velocity. We’ve encouraged the stakeholders on our project to assist with user story creation, but prioritization falls solely within the hands of the product owner.
- Wireframes shared through the ArchivesSpace Google Group prompted community engagement. Payten Giles, one of the developers, jumped into the discussion to explain his thinking and gather feedback. We welcome questions and feedback. With release 0.2, we’ve seen members of the community become eager enough to kick the tires of the alpha version by downloading it and deploying it locally. This seems to be the first exposure for many in the archival community to an agile process, and we are hoping the community will continue to take an active role in trying to determine requirements and help with the process of crafting user stories.
- Development happens within branches in HMs repository, and they are responsible for managing their own merge processes. Github serves as a strong focal point for not just the code itself, but also technical and deployment-related documentation.
- High priority feature for community
- Multiple date types allowed; timestamps incorporated into data model
- Used to manage IP, license, and restriction information
- Available for download. As product owner, Mark tests and provides feedback to developers.
- Available for download. As product owner, Mark tests and provides feedback to developers.
- Available for download. As product owner, Mark tests and provides feedback to developers.
- Based on imported EAD data. Tree will need to be manipulable
- Need to provide robust collection management as provided by Archivists’ Toolkit
- Used to track administrative actions (e.g. deeds of gift) and preservation actions
- It’s early days for ArchivesSpace planning for interoperability. We are getting questions from users about integration with various products—in particular digital preservation systems and discovery systems. Our current thinking is that APIs are the way to go. An early ArchivesSpace API prototype is available on github for evaluation and will be used for data migrations from Archivists’ Toolkit and Archon to ArchivesSpace. We are less certain about how to leverage the investment in ArchivesSpace through interoperation with other project, but the user community needs to take an active stake in guiding the integration needs. We recognize that the application will need to integrate with local and shared discovery environments and digital asset management systems, and we have been looking to determine how the application could be best integrated with other emergent projects like ConservationSpace .
- LYRASIS has been selected as the organizational home to sustain ArchivesSpace once the grant funded period has completed. LYRASIS is well positioned for this role. LYRASIS will manage ArchivesSpace memberships, separate from LYRASIS memberships, offer hosted solutions and help desk services, manage software upgrades in consultation with the Technical Advisory Council, and coordinate an ArchivesSpace training program.
- The product and project will be sustained through tiered membership designed to be inclusive of all types of archives. An elected Board will provide oversight, with a User Advisory Council insuring strong connections with the community. The Technical Advisory Council will prioritize development, based on information gathered by the User Advisory Council.
- ArchivesSpace will build on the strong relationships with the Society of American Archivists that are already in place for community involvement with legacy products; Archivists’ Toolkit and Archon. SAA will continue to play a key role. In addition, LYRASIS, as the organizational home for ArchivesSpace will leverage its relationship with other professional organizations to insure there is a broad awareness of ArchivesSpace in the cultural heritage community.