What would you like
Add a test environment so that automation responses to new Node.js releases can be tested without requiring production server availability.
Add unit tests and integration tests.
Mock relevant endpoints, such as:
Determine if tests should still assume sourcing musl builds from https://unofficial-builds.nodejs.org or if this should fast-forward to a scenario where this is no longer necessary.
Why is it needed
Automation that deals with processing Node.js updates often fails, leading to manual effort to correct the situation.
Problems have been known about for some time, but efforts so far to correct the issues have had limited success. The outcome of new releases has tended to cause surprises.
Security updates, which typically cause release of updates for each supported release line on the same day, have proved particularly difficult to deal with correctly. Automation is supposed to proceed without waiting for musl builds, which can take many hours, and instead is waiting when it shouldn't.
Planned upstream changes mean that code changes will be necessary:
- promotion of Alpine to Tier 2
- Node.js introduction with alpha releases
If these can be tested before their function is needed, that would help with automation predictability.
What would you like
Add a test environment so that automation responses to new Node.js releases can be tested without requiring production server availability.
Add unit tests and integration tests.
Mock relevant endpoints, such as:
Determine if tests should still assume sourcing musl builds from https://unofficial-builds.nodejs.org or if this should fast-forward to a scenario where this is no longer necessary.
Why is it needed
Automation that deals with processing Node.js updates often fails, leading to manual effort to correct the situation.
Problems have been known about for some time, but efforts so far to correct the issues have had limited success. The outcome of new releases has tended to cause surprises.
Security updates, which typically cause release of updates for each supported release line on the same day, have proved particularly difficult to deal with correctly. Automation is supposed to proceed without waiting for musl builds, which can take many hours, and instead is waiting when it shouldn't.
Planned upstream changes mean that code changes will be necessary:
If these can be tested before their function is needed, that would help with automation predictability.