Setting up new Digital Mitford workspace and image archive system

[ebeshero]

@ebeshero @lmwilson @KellieDC @RJP43 @ghbondar @djbpitt (I am "pinging" each of you from GitHub because I've been chatting with you about one aspect of this or other over the past few weeks.)
Here is a Task List for getting Digital Mitford's workspace out of the Box system. Simply getting a sense of the arc of the transition will be helpful. We won't be completing all this overnight, but most of it is already underway, starting with a major file directory backup from Box and preparation of new space on the Digital Mitford web server. My hope is that we'll have a new system and a new set of project tasks to share with the Mitford team when we next meet, which seems likely to be virtually, and perhaps in May.

Transition Tasks:

Many universities (including first Pitt and now Penn State) are terminating their arrangements with Box.com, so we are finally pressured to leave Box this spring. Here is what we are working on to create a new and optimally more robust workspace driven by our own project webspace and GitHub.

Prepare a new coding workspace on GitHub

• Our GitHub organization will host our project work with proper version control and better integration with our web hosting on Digital Ocean.

• We will need to refile our project workspace file directories on the appropriate GitHub repo (Letters, Poetry, Drama, etc). We'll keep the same file directories, but we'll be porting the image files to a special "bucket" on the https://digitalmitford.org web server. (GitHub is not optimal for image storage anyway.)
  * Our "bucket" will be serving the images by URL, which seems optimal for IIIF delivery and integration in the project web interface.
  * We will need to prepare website delivery that associates GitHub project directories with their corresponding image folders on Digital Ocean.
• Prepare to orient the Mitford active team to GitHub with help from the oXygen XML Editor's GitHub integration

Prepare a web interface for views of the images and code

On the https://digitalmitford.org server, we need to create a web interface that pulls together the following:
1) The TEI XML file
2) The associated bundle of images by their corresponding directory ("API endpoint")
3) For publishable code: the HTML published form for reading (from eXist-dB on digitalmitford.org)
4) Metadata from our box labels
5) Hides views of images from Rylands and other permission-only archives to gate them as work views for Mitford team only.

To make this work we need to do the following:

• Notify the active Digital Mitford Team that we must stop any new uploads to Box as we are migrating away!

• Retrieve the Box labels on our file directories: study up on the Box API for Developers

• Batch process the images to do the following:

  • create parallel image directories on the "bucket" in the Digital Ocean server using same directory names as workspace
  • transform them to JPEG2000 (apparently via the ImageMagick processor within Cantaloupe) Work out grid dimensions in the process.
  • Test/check URLs/APIs. Are images retrievable/viewable in web browsers?
  • Install Cantaloupe on newtfire / digitalmitford.org and see about IIIF service

• Build a website structure for workspace + publishing that creates an HTML page pulling its data from
  a) the GitHub repo and workspace directory,
  b) image URLs from corresponding directory (and gate if not Reading-CL),
  c) published view from eXist-dB if applicable.

• Activate Digital Mitford Team Hypothes.is for commenting directly on https://digitalmitford.org pages and set up system for archiving Hypothes.is comments on GitHub

• Establish a new system in TEI XML for referencing image APIs on the project. (That's a whole new set of tasks).

[ebeshero]

In progress: Digital Ocean Image API server is ready.
But blocked because directory structure isn't permitted, or I need to find out if/how to create one?
Need to make sure metadata is clear on each image file.