2015 Near Term Technical Roadmap for OpenSpending

Comments:

  • Add a short summary at the top of what this spike solution does for the user e.g. “User can start from a file (CSV) and have it visualized (and browsable) in a few quick and simple steps”
  • Suggest dropping step 4. I think a way to “browse the catalog” can come later and is probably part of a dedicated flask app. In the mean time we have tools like http://data.okfn.org/tools/view as stop-gaps.
  • Step 6: index.html from step 4 is gone now (see prev point). Suggest that we probably want to have our minimal flask app (or even just html / js app) here so we can have theming and the like.
  • Overall: I think it would be worth mocking the front page for this app right off and thinking of the 1-2-3 parts of this. I think this app (whether flask or o/w) would be where we’d actually implement landing page and step 6 and possibly at least frontend of everything in between.

More generally, here are some thoughts from our discussion on possible MVPs / Spike solutions

  • Start from a file (CSV) and see it visualized (and browsable)
  • Import UI + DataStore (basic) + Aggregation (basic) + Browse and search + Visualization (e.g. treemap)
  • Not needed
    • identity (?)
    • No authorization for datastore (?)
  • MVP 2
  • Start from a file (CSV) and see it visualized
  • Import UI + DataStore (basic) + Aggregation (Basic) + Visualization
  • Not needed
    • Identity
    • Authorization
    • Any API (for viz) beyond CORS + S3
    • Search and browse
  • MVP 3
  • Take existing data from current system and see it visualized
  • MVP2 minus import UI (as we do not need to import - we script conversion)