TCR-39-ui-serials-management evaluation

[maccabeelevine]

TC provisionally approved this module on 2024-03-18. TC and the dev team unanimously agreed that the dev team will fix the major (and critical) code smells and get the test coverage to the 80% threshold, and that it should be doable in around 2 weeks. The dev team suggested and TC agreed that substantive test cases should be prioritized rather than hitting 80% as quickly as possible.

[maccabeelevine]

When the Technical council's decision is contrary to the evaluation, or after any provisional acceptance or deferral, the Technical Council is required to provide written justification for their decision.

TC approved the module despite the four failing criteria because:

1. The third-party dependency licensing issues were issues with upstream Stripes modules, and not with ui-serials-management itself. TC earlier decided to raise that issue with #stripes and then to consider changes to the process next time we do a TCR Process Improvements cycle.
2. The WCAG compliance issues are also apparently upstream issues, not inherent in this module. The dev team has agreed to address them with the stripes team. From the self-eval:

• We have addressed the majority of issues reported by the axe DevTools Chrome Extension. However there are still two outstanding issues, both of the same type (MCL rows generating the issue: "Certain ARIA roles must contain particular children"):
  * UISER-71
  * UISER-75
• We see this issue in other MCLs in Folio, for example: Calendar settings -> Current calendar assignments and we believe there is the need for changes to Stripes to support MCLs that do not provide linkable row/row elements. If this is something we can fix in the application we will do so. If a Stripes level fix is required we will work with the Stripes team to work out the fix

3. The dev team has been making active progress on the other two failing criteria (code smells and test coverage) and there is unanimous agreement on the solution and timeline. So they are pending as part of the provisional approval, see above.

[maccabeelevine]

TC provisionally approved this module on 2024-03-18. TC and the dev team unanimously agreed that the dev team will fix the major (and critical) code smells and get the test coverage to the 80% threshold, and that it should be doable in around 2 weeks. The dev team suggested and TC agreed that substantive test cases should be prioritized rather than hitting 80% as quickly as possible.

This TCR was brought up in TC today as not yet merged. It was provisionally accepted with work to be done on code smells and test coverage. Looking today:

• The test coverage is at 69%, so not yet at 80%. We did agree "that substantive test cases should be prioritized rather than hitting 80% as quickly as possible", but I think the assumption was that we would still get to 80% eventually.
• There are 19 code smells but I don't see where they would be marked "major" if applicable. Discussing in Slack.

[julianladisch]

The test coverage should go up, not down: https://sonarcloud.io/project/activity?custom_metrics=coverage&graph=custom&id=org.folio%3Aui-serials-management

One of the nineteen issues is high severity: https://sonarcloud.io/project/issues?issueStatuses=OPEN%2CCONFIRMED&id=org.folio%3Aui-serials-management

The high severity issue is nesting of 4 functions: https://sonarcloud.io/project/issues?issueStatuses=OPEN%2CCONFIRMED&id=org.folio%3Aui-serials-management&open=AZQndphgDaPCYKNj2Fzt
This can easily been fixed by moving the calculation of starting values into a method of its own.