Samvera Newspapers Interest Group Call: 2018-11-01

Owned by Eben English
Last updated: Nov 01, 2018 by Brian McBride
2 min read

Time: 1 PM EST/ 12 PM CST / 11 AM MST / 10 AM PST

Call-In Info: 712-775-7035 (Access Code: 960009)

Moderator: Eben English (Boston Public Library)

Notetaker: TBD (Etherpad: https://etherpad.wikimedia.org/p/Samvera_Newspapers_Interest_Group_Call__2018-11-01)

Attendees

Agenda

  1. IMLS Grant Project Update

    1. General development update (https://github.com/marriott-library/newspaper_works)
    2. NewspaperWorks fixtures (https://github.com/marriott-library/newspaper_works_fixtures)
    3. Newspapers vagrant box for community testing (https://github.com/marriott-library/samvera-vagrant)
    4. Public Newspapers app for UI testing (similar to nurax) is still in progress

  2. Batch ingest workflows
    1. UI vs. command-line

  3. Page interaction UX
    1. Single page vs. page-in-issue
      1. Single page:
        1. https://digital.bentley.umich.edu/midaily/mdp.39015071755511/133
        2. https://chroniclingamerica.loc.gov/lccn/sn86063774/1918-10-31/ed-1/seq-2/

    2. Page-in-issue:
      1. https://www.lib.umd.edu/univarchives/student-newspapers/id/71ac23ab-95e9-434a-a473-ad6bef3ef247?query=flooding

  4. Content Examples
    1. https://drive.google.com/drive/folders/0BwKKtxaBVqjEbE5zMFdWUEU4WGM?usp=sharing
      1. Still need: CONTENTdm, TEI, Olive, Veridian

  5. Intel sharing from other groups/projects

  6. Next meeting: 
    1. Change schedule to every other month? - Yes
    2. Thursday January 3rd 
    3. Meetings will be held using zoom.us https://zoom.us/j/993200218 


Notes:

Project Update:

Source code for latest work is available at github repo. The past month has been spent on working on developing NDNP ingestion functionality. Implementing JSON word coordinate structure will persist which will be used for IIIF search highlighting. Additional work has been applied to implementing and storing newspaper specific derivatives. Also, Hyrax does not currently have mechanism to differentiate between derivatives and multi-part works. 

Hyrax does not currently store derivatives in Fedora Commons, only the file system. We're working on implementing a mechanism to create pointers for objects on the file system. This is a challenge due to complexities in the hyrax actor stack and the a-sync processing. 

Newspaper_Works is currently working with Hyrax 2.3 The Newspaper_works gem will be updated shortly to be compatible with the latest version of Hyrax.

Create a separate repository for newspaper_works_fixtures contains only selected newspaper data. Newspaper_works does not require this gem by default.

Batch Ingest workflows

How do users currently work with batch ingest workflows?

  • University of Michigan - Images are stored in DLXS, reference to images are updated in blacklight-based application via a rake task. 
  • Indiana University - (page turning application) currently use CLI for batch ingests. Use a YAML file that container metadata and structure. Automated workflow, a script will scape a specific folder for YAML files.

Page Interactions UX

User experience, providing a single page view vs providing a single page with as part of the entire newspaper issue. What are the pros/cons of either approaches? 

  • A multiple display of pages, can be a bit confusing if there is not article segmentation. If search only returns pages, the user just assumes that a result will be a page, not an issue. 
  • If searching for a term, subject, their might be multiple results for an issue. Assumption being that users would return to search results rather than going through the issue viewer for other highlighted search terms.

Content Examples

IUPI added CDM newspaper data to google drive