Revisit CONTRIBUTING.md

[hancush]

Background

Now that we've got a year+ of R&D under our belts, it would be a good exercise to revisit our documented process to determine whether it reflects our real life habits and practices.

Proposal

I propose that we have a short conversation at an R&D meeting with a designated notetaker, with a focus on discrepancies between the documented R&D process and R&D in practice. The notetaker will summarize those discrepancies in this issue. Together, leads will decide whether they merit a process change or corrective guidance to R&D participants.

Deliverables

This conversation will result in an updated R&D procedure or clarification of the existing procedures.

Timeline

I expect that we can accomplish this in a day or less.

[jeancochrane]

I will schedule a meeting for this.

[jeancochrane]

Some changes we discussed making during our meeting today:

• Clarify that the "proof of concept" is really a spike, and doesn't need to be fully-featured
  • Remove language about feature parity
• Move section 5. Pilot use of the tool on a project above section 3. Recommend adoption, further research, or abandonment
• Provide some high-level guidance during piloting on how to protect against bad choices
  • Have a contingency plan in case the tool doesn't work
  • Keep an eye out for signs that the tool is difficult to debug, e.g. differences between dev and production or hidden output
  • Keep an eye on the amount of time you spend; if you find yourself going down rabbit holes (e.g. a whole day debugging one problem) consider falling back to your contingency plan
• Clarify that we should use our discretion for step 4. Notify partners of recommendation
  • May not be necessary if:
    • A partner has been informed throughout the R&D process
    • The change to our stack is minor (e.g. recommending a package for an existing framework)
  • Probably necessary if:
    • The new tool requires a billing account
    • Staff will need to be trained on the new tool
• Update step 6. Produce adoption artifacts to mention making a plan for spreading knowledge
  • Investment time is a good vehicle for this