This presentation aims to explore the possible integration of Samvera digital object repositories with additional web services using message brokers. There have been cases in which it is necessary to synchronize content updates between repositories and additional library systems such as library catalogs or digital exhibit publishing software. Within this context, developers may benefit by exploring architectural pattern in which a dedicated message broker receives asynchronous notifications of repository content updates, new ingestions, and deletions. In response to having received these messages, the broker may then broadcast these events to other listening library systems. The library systems then may reindex or update their own content accordingly. A conceptual overview of this architectural pattern shall be provided, followed by an overview of an implementation local to the systems within the Princeton University Library (synchronizing content between implementations of Valkyrie and Spotlight using RabbitMQ). The outcome of this presentation would be to identify other Samvera adopters who may also be utilizing message brokers, with the ultimate aim of determining whether or not this approach would be beneficial to a larger number of community members. A video recording of this session is available at the 'Related URL' below. and A presentation given at Samvera Connect 2018 described thus
I will use Natural Language Processing and Machine Learning techniques to analyze issue backlogs in applications from institutions throughout the community. I will apply a variety of techniques in an attempt to answer questions like, What kinds of open issues do we have in general as a community? Can I extract an interesting set of widely-desired features or widely-held use cases? Can I identify connections that might lead to collaboration across institutions? What solutions already exist that might advance open issues? Can I link open issues in one backlog to merged PRs in another repository? What have people been working on recently? Can we characterize the full set of issues that have been closed over the past year? What patterns of development do repositories follow? Can we describe the life cycle of repository development by aligning issues based on their creation / completion dates relative to the initial commit? These may or may not be the exact questions my talk will address, depending on the direction the project naturally takes. I will focus on applications in use or under development at institutions, as opposed to community-maintained engines and core gems. This talk will describe my process, results, and evaluate the success of the endeavor., and A presentation at Samvera Connect 2019 described thus
A presentation given at Samvera Connect 2018 described thus and A project description of adding our first A/V materials to Princeton's repository management software, Figgy. I'll briefly describe the project history, the collections in question, and project management strategies. Will demo the resulting ingest workflow. A video recording of this session is available at the 'Related URL' below.
We will present our use case for and development of a GraphQL API in Figgy, our Valkyrie-based digital collections management application. We'll give a brief summary of GraphQL itself, demo an in-broswer query tool called graphiql, and show how we used the graphql gem to quickly develop and deploy a GraphQL API endpoint. A video recording of this session is available at the 'Related URL' below. and A presentation given at Samvera Connect 2018 described thus
A presentation given at Samvera Connect 2020 On-line described thus and After trying to navigate deployment, configuration, performance, and scaling issues of several different image servers and support infrastructure (Cantaloupe, Aware, Riiif, nginx, and SquidCache, to name a few), we decided to see if we could build something less general/configurable but far more suited to our use case and runtime environment. serverless-iiif started out as a bare bones, proof-of-concept demonstration of how a scalable, high-performance IIIF image server could be implemented in a small, inexpensive AWS Lambda function. Just over a year later, the project serves as the basis for high-volume IIIF services running in production at Northwestern University, Princeton University, the University of Notre Dame, and the Royal Pavilion & Museums, Brighton & Hove. This presentation will cover the project from its beginnings (as a small demo repository carved out of Northwestern's cloud repository infrastructure), through a number of forks, merges, performance enhancements, deployment improvements, and into production. We will also include performance benchmarks, current production stats, and some thoughts on future work. The 'Related URL' below links to a video recording of the session. The video has closed captioning.
A presentation given at Samvera Connect 2020 On-line described thus and Figgy is Princeton University Library’s staff-facing repository management application. This presentation will share screenshots, user stories, and technical overviews of all the forms, magic buttons, storage integrations, drag-and-drop targets, rake tasks, and directory watchers that Figgy provides to support the different workflows our users have for ingesting content. The 'Related URL' below links to beginning of this presentation in the day's YouTube recording.
A presentation given at Samvera Connect 2020 On-line described thus and Princeton University Library’s digital projects and initiatives were seriously disrupted by COVID-19. Digitization of materials for projects and forthcoming exhibitions came to an unexpected halt at the same time as patrons and staff were separated from physical objects and library spaces. Necessity, however, provided an opportunity to reassess digital projects and how staff members interact with and contribute to our repository (Figgy). We focused on the creation of workflows and documentation for new contributors who would be working in the repository, helping them enhance existing digital objects with OCR, item level organization, structural metadata, page labeling, and IIIF display attributes. We describe how we were able to use Figgy and unexpectedly-available staff time to make more effective research tools and provide a better user experience for patrons and staff working with our digital collections. Such enhancements add immense value to our collections as well as to our applications, and the work can be done effectively by a wide range of staff from different departments with variable skill sets. The 'Related URL' below links to beginning of this presentation in the day's YouTube recording.
A presentation given at Samvera Connect 2020 On-line (originally titled 'Cloud storage service uploads for a Valkyrie repository') and described thus and This presentation aims to outline and discuss attempts throughout 2019 and 2020 to integrate an early pre-release of the BrowseEverything 2.0 component for supporting Google Drive file uploads into a Samvera repository. While this shall be restricted in scope for cases which were specific to the Princeton University Library and a Valkyrie-based repository Figgy, the hope is to encourage discussions regarding obstacles which were encountered and to aim to generalize the solutions which were discovered in this integration. The 'Related URL' below links to beginning of this presentation in the day's YouTube recording.
A presentation at Samvera Connect 2020 described thus and Update on recent and coming work for the Valkyrie gem. The 'Related URL' below links to the start of this presentation in the YouTube recording of the day's events.
Princeton and Northwestern recently underwent a two-week spike to explore a set of new technologies we might use in our respective teams. We looked at ElasticSearch, Elixir, and Phoenix. This presentation will go through our expected outcomes, strategies for a successful collaboration, our eventual output, and a retrospective on how the process went with advice for any others looking to do this kind of exploratory work. The video recording of this segment is available at the 'Related URL' below.
A lightning talk at Samvera Virtual Connect 2019 described thus and This will be a shorter presentation to give an update on changes to Valkyrie since Samvera Connect, as well as provide some guidance on when 2.0 will be released and what to expect.
A lightning talk at Samvera Virtual Connect 2019 described thus and I'll present the many contexts in which pair programming can be beneficial in different ways, reasons to use pairing as part of the regular practice of your team, the basic mechanics of how pairing works, prompts for staying mindful of power dynamics while pairing, and ideas for introducing the practice to a team that has never really done it before.
Valkyrie is a new persistence layer for Samvera, designed to address performance and sustainability problems. It was developed through the existing Samvera Working Group process, showing that the current Samvera community governance structure can be used to tackle big problems. Valkyrie features pluggable persistence options, allowing Samvera applications to use not just the historical combination of Fedora and Solr, but also other options like Fedora or Solr by themselves, PostgreSQL, and local disk. Allowing Samvera applications to use different persistence options refocuses the Samvera community, shifting away from persistence in Fedora as the defining aspect of the community. Instead, the focus shifts to the shared tools built by the community. and A presentation at Samvera Virtual Connect 2018 described thus
A presentation given at Connect 2017 described thus and Princeton University Library has spent the last two years developing Plum as a staff-only ingest platform for books, manuscripts, geo-spatial data, and archived ephemera. This talk would go through a history of the design goals, the successes we've had, the problems we've run into, and a look into the future.