Design Session Notes

Questions from the meeting

1. Do we actually want SQL style DB

   1. Data will be very different between Midwest and north east etc.

   2. May be better to use nosql

   3. Differences

      1. Midwest → list the ways cover crop can be terminated

      2. Northeast termination has goals → ratings

      3. Stars meaning differences

      4. Precipitation gradient (needs info)

   4. How often will changes be made to data shape for the selector

      1. Once or twice a year

2. Anonymous concerns

   1. We don’t want users to be able to easily export all the ratings for the selector

   2. Maybe 4 tiers

      1. anonymous

      2. local storage

      3. DB tracking no data aggregation

      4. DB tracking with aggregation

3. 7th API for shared things

   1. Species of cover crops

   2. Location → zone

   3. Zone → region or location → region

   4. Weather (30 year averages ?)

   5. Wrap frost date into our weather API or use an existing tool

Research Action Items

1. Research region appropriate standardization.

2. Research Flexible vs single postgresql

3. Research commonalities between APIs for DSTs

4. Research a shared DB for the above

5. Research UserDB

6. Research Inputs and outputs for the DSTs

7. Create spike tickets for research above

Post Research Action Items

1. Regularize regional data sets for selector as much as possible

2. Design doc for each API

3. Build each API

4. Find commonalities for shared DB

5. Create shared DB

6. Create user auth