...
- Our code that is marked deprecated already
- Using Python's built-in deprecation warnings
- How can we focus on removing this so that we can see Django's deprecation warnings
- Jenkins can't handle that many deprecation warnings (mostly)
- Count the number of deprecation warnings and ensure that the number only ever go down.
- A ticket exists on the TE backlog (Jeremy Bowman (Deactivated) will add a link later)
- Bad imports preventing us from removing sys.path nonsense
- Advocate usage of the OEP and propose changes when the process described is not useful
- Diana Huang to add link to OEP here
- Adding dates to the DeprecationWarning to indicate when we expect things to no longer work
- Organizational focus on eliminating warnings
- Bubbling up dates into code/tickets so that product can respond
- Keep track of deprecation tickets that will impact the next release
- Could be wiki page, could be Jira filter
- Examples of situations:
- Dynamic Pacing
- We will deliver the feature
- Engineering work needs to happen as part of delivering the feature before we can move onto other projects
- Dynamic Pacing
- Percentage of engineering time dedicated to it
- Dedicate a day or two to engineering work per sprint
- Escalation teams - can they take on debt work?
Future Discussions
- How do we deprecate features?
- Getting product buy-in?
- Notifier, as an example
- Presenting to the engineering org about successes
- Figure out how much time we are spending right now on tech debt
- Increase that number?
- Are we doing enough?
- Focus on removing pylint and deprecation warnings
- IOUs
- when introducing tech debt, keep track of the new things that are added
- Track engineering pain points
- Course Modes
- access
- Questions of ownership?
- Guilds - how are they working?
Diana Huang - schedule a meeting to go through the list and update information on them