Evaluate django-simple-history for history tables

[fyliu]

Overview

We would like to know if django-simple-history can fulfill our needs of keeping history of certain tables. We can potentially save time and developer efforts.
The package automatically keeps track of change history metadata like who made the change.

Details/Requirements

• should be useful for Create Table: event_history #49
  • An event record has a long life with cancellations and overrides, and metadata about who made those changes should be useful
• not useful for Create Table: permission_history #37
  • needs to hold the old permission records that are no longer in effect
  • new permission records are created for each update, so there's no updating each record to keep track of
• should be useful for Create Table: project_program_area_status_history #67
  • needs the history of the status field for each project

Action Items

• install plugin into django and test the functionality
• evaluate whether it can fulfil our needs
• document usage if it's useful
• release the dependency on the following issues and if there are no other dependencies, move them from the icebox to the new issue approval column and add the ready for dev lead label.
  • Create Table: event_history #49
  • Create Table: permission_history #37 no longer applicable
  • Create Table: project_program_area_status_history #67

Resources/Instructions

• This issue was a part of Create Table: event_history #49
• django-simple-history

[fyliu]

We need to come up with requirements for what we want to be able to do with a history model. Maybe document it in this issue or create a separate issue for it.

[ExperimentsInHonesty]

We will definitely create a separate issue for requirements. That issue will be this issue's dependency

[fyliu]

I'm trying to figure out why I crossed out the project_status_history table from requirements. Looks like that table is still part of the schema and needs a history. It could be that I thought it's not something that djang-simple-history can solve. I think it's still worth keeping it in the list since this is an evaluation issue, to see if it can be satisfied as well.

I'm going to un-cross it.