Research spike: Nx bonuses

[joshgamache]

A timeboxed research spike to explore the features outside of our core requirements for Nx. Will try to evaluate for any quick wins and considerations on implementation time vs value. [2-3 days]

Known features to investigate

• Self-hosted remote caching (Nx Replay)
  • Cache to GCP bucket? Artifactory?
• Self-hosted remote task execution (Nx Agents)
  • Github actions? OpenShift jobs?
• Releases (Nx Release)
  • Drop in replacement for our Release-it module?
• Github PR integrations
• Code generation
  • Automatic documentation for components, API?
• Plugins
  • Nx Playwright for E2E
  • Nx Jest for unit testing Already implemented Vitest
  • Image builds (Nx Container is already in use for Reporting, optimizations likely)
  • Linting
  • SonarQube

Questions

• Estimated effort?
• Developer adoption effort (automatic or needs training)?
• What part of the process does this target (DevOps, Dev experience, Security, etc)?
• Maintenance complexity/considerations?
• Technical limitations (esp. resource requirements)
• Reach Impact Confidence Effort score?

[patriciarussellCAS]

Should this be pulled into Sprint 11? @joshgamache

[joshgamache]

@patriciarussellCAS

Should this be pulled into Sprint 11? @joshgamache

Probably a good candidate as a stretch goal!

[pbastia]

Suggestion:

• timebox this as a short research
• write cards for each individual feature that we deem interesting for the team, that we can then estimate and schedule depending on their individual value

[joshgamache]

Research and analysis is posted here in the Sharepoint.

[pbastia]

Left to do: Write cards about the features we think we'd like to implement one day. Document the ones we won't in this card.

[joshgamache]

Tickets were created:

• bcgov/cas-reporting#232 - Nx Affected, cache
• bcgov/cas-reporting#240 - Nx Release
• bcgov/cas-reporting#239 - Nx Affected (but for Docker containers)
• bcgov/cas-reporting#241 - Nx Atomizer(E2E)
• bcgov/cas-reporting#242 - Nx Cloud, Nx Replay

Not implemented:

• Code generation: wait and see

Won't (likely) implement:

• Nx Agents