Viewer Upgrade & Embed Change Order: Viewer beta release

I. Overview

Description of change

Expand the scope as originally outlined in the Project Charter to include the support of a publicly-available interim beta version of the upgraded Viewer application so that our user community can begin to leverage improvements until in-scope work on Mirador 3 is complete and all MPS manifests have been generated in production. 

Reason for change

Due to multiple dependencies on the Media Presentation Service (MPS) project, the release of an upgraded Viewer that meets in-scope requirements and that can access MPS manifests will be significantly delayed. An interim beta release would enable end users to experiment with improvements for viewing objects with eligible manifests. 

As the public-facing front end of MPS efforts, the Viewer is directly impacted by issues related to MPS manifests and manifest generation. MPS-generated manifests deliver extant content and metadata in full compliance with IIIF specifications as required by the updated Viewer, but MPS manifest generation is currently paused and there is not yet an estimated date when all manifests will become available. In addition, due to initial Viewer-driven manifest changes in May, the number of revised MPS manifests available with those changes is extremely limited, which limits both the functionality available to end users as well as development team testing. Further MPS dependencies block core functionality including the ability to display restricted images, perform text search, and craft user-friendly unique URLs for individual canvases. 

II. Requirements

• Release a publicly-available beta version of the Mirador 3 upgrade to production.
• Mitigate user expectations through educational text on the beta site.
• Create a Mirador-3-specific help resource available on https://ask.library.harvard.edu/, to be linked to from the beta application.
• Educate LTS Support so that they can field incoming issues and questions.
• Encourage users to experiment with the beta version by linking users to the beta site through succinct, Viewer-adjacent HTML links on the following Library discovery platforms: 
  
  • CURIOSity
  • HOLLIS (once restricted images can be displayed)
  • HOLLIS Images (once restricted images can be displayed)
  • (This requirement was revised September 2023. See Change order VUE-002.)
• Iteratively update the beta site as improvements are unblocked.
• Take down the beta site when Mirador 3 is deployed to production.
• Ensure the final version of Mirador 3 points to a resource on https://ask.library.harvard.edu/ that does not reference the beta release.
• Remove discovery platform links to the beta site when Mirador 3 is deployed to production. 

III. Risks & Impacts

Risks if implemented

• Delayed embed development
• Development team time and attention will be diverted from other project work, thereby prolonging project delivery and developer resource commitment. This may be mitigated by pursuing a strategy for text, design, and technical approach that requires minimal development team time. 
• Potentially frustrating user experience
• Users may be confused or discouraged by features that are either unavailable, inconsistently available, or not yet functioning as expected. This may be somewhat mitigated by supporting communication: beta site text for context, an in-application link to help on https://ask.library.harvard.edu/, and coordination with LTS Support for fielding tickets. 
• Scope creep
• It may become more challenging for LTS to enforce the project scope if stakeholders begin to submit feature requests in response to their experience with the beta release. This can be mitigated with the Executive Sponsor’s commitment to the project scope in his communication with stakeholders.

Risks if not implemented

• Extended wait for upgraded Viewer
• Stakeholders and power users invested in Mirador 3 improvements will wait an unspecified amount of time to begin to explore the updated application. 

Impact on resourcing

• Additional work requires additional time from all team members: technical product owner, project manager, development team (e.g., front-end, back-end, UX, QA), and LTS HOLLIS consultant.

Impact on budget

• There is no anticipated budget impact. 

Impact on timeline

• This change may delay delivery of the embed solution, as team resources are diverted to beta release tasks. This change is not anticipated to impact Viewer delivery, which is currently blocked by MPS dependencies. 

Impact on project goal

• There is no anticipated impact on the project goal. 

IV. Expanded scope definition

Additions in bold italics

In Scope:

• Develop an automated testing and deployment pipeline (aka continuous integration and continuous delivery aka CI/CD)
• Viewer upgrade:
  • Reduce customization by using plug-ins and out-of-the-box improvements
  • Meet Harvard University’s accessibility requirements
  • Comply with copyright/usage restriction requirements
  • Conform to end-user device expectations
  • Implement analytics to improve future value delivery
  • Upgrade Viewer application to the latest available version of Mirador
  • Incorporate the new v3 IIIF manifest while ensuring backwards compatibility with v2 IIIF manifests
  • Leverage standard Harvard authentication and authorization 
  • Achieve majority feature parity with the pre-existing Viewer application. Delivered functionality will:
• Support interim beta version of upgraded application on production until all MPS manifests have been generated in production and in-scope work on Mirador 3 is released. 
• Embed service
  • Develop an embed solution
  • Replace Viewer instances with the embed solution in the following Harvard Library discovery layers: 
    • HOLLIS
    • HOLLIS Images
    • CURIOSity
  • (This requirement was revised September 2023. See Change order VUE-002.)

Out of Scope:

• New features that aren’t included in the latest out-of-the-box version of Mirador 
• Viewer support beyond single image and page-turned objects (e.g., time-based media, geospatial data, three-dimensional objects) 
• Replacing Viewer instances with the embed solution in discovery layers other than those enumerated above
• Decommissioning/retiring supported services

V. Acceptance

Accepted by: Stu Snydman

Effective date: June 30, 2023

Prepared by: Sara Rubinow