pocket_archive/doc/roadmap.md

Pocket Archive Road Map

The first goal is to build a working prototype, with all the basic functional components, even if not entirely developed or only usable in a specific development environment, to demonstrate the overall workflows and functionality.

The second step is to produce a basic application, which is fully functional and available for use by the intended audience.

Beyond the basic application, new features and bug fixes will be driven by usage and opportunities for expanding adoption in relevant areas.

❏ = pending; ⚒ = in progress; ⎊ = blocked; ✓ = complete; ✖︎ = not implemented.

Prototype (design -> alpha) — DONE

• Configuration + config parser
  • Application
  • Content model
  • Validation rules
• Submission module
  • SIP building
  • Metadata from LL
  • Brick structures
  • Structure inference
• HTML generator
  • Index
  • Resource
  • Static assets
• Non-HTML generators
  • RDF (turtle)
  • Transformers
  • JS search engine index
• CLI
  • Init archive
  • Deposit
  • Generate site
  • Generate LL (single resource)
  • Generate RDF (single resource)
• Front end
  • JS search engine
  • Add collections to index page
  • Basic styling
    • Default type icons
• QA
  • ~50 resource data set

Basic application (alpha -> beta -> v1.0) — IN PROGRESS

• ✖︎ Management UI & API
  • ✖︎ Deposit via single tar or zip file submission
• ✓ Deposit via local hot folder
  • ✓ Watch local folder and trigger deposit
  • ✓ Option to regenerate site after deposit
  • ✓ Option to clean up sources & LL on success
• ✓ Proper collection handling
  • ✓ Dedicated template
  • ✓ Link to markdown doc for presentation page
  • ✓ Handle artifacts as members
• ⚒ Preservation
  • ✓ Dump archive RDF
  • ❏ Backup full environment (including config)
  • ❏ Restore whole archive from RDF & data folder
• ❏ Submission
  • ❏ Deleting resources
• ❏ Content model
  • ❏ Multilingual support
  • ❏ Content model dump for users (CSV?)
  • ❏ Local overrides
  • ❏ Relatioships inference
  • ❏ Markdown support for individual fields
• ⚒ Generator
  • ⎊ Generate site for one collection only
  • ✓ Generate LL for submission
  • ❏ htmlgen option for local file or webserver URL generation
  • ✖︎ Generate RDF (multi) [addressed by dump archive RDF]
• ❏ Front end
  • ✓ Mobile-friendly layout
  • ❏ Enhanced styling and access
  • ❏ Category browsing
  • ❏ Improve search indexing
• ❏ Testing
  • ❏ Unit tests
  • ⚒ Roundtrip submission, download LL, update, resubmission
  • ⚒ >100 resource data set
• ⚒ Documentation
  • ✓ Break main sections off README
  • ✓ Watchdog process guide
  • ⚒ Submission guide
  • ⚒ Content modeling primer (archivist)
  • ⚒ Content modeling manual (sysadmin)
  • ⚒ Glossary
  • ❏ Site generation guide

Post-release

• ✖︎ Deposit via remote hot folder
  • FTP [Addressed by separate FTP server]
  • S3 [Not a good choice - implementations such as MinIO mangle all files and an additional S3 pull is required to get the deposited files. Not efficient.]
• Incremental build
• Rebuild only site assets
• Custom templating
• Deposit via S3 source pointer