As a developer I want to make sure that type inference is performant for new and chaning types. We benchmark our critical functions for that locally now. It would be better to explicitly include this in the tests. This is beneficial to ensure the performance of types and encourages developers to always test performance. The added workload is neglectible, as performance testing should already be part of the normal workflow.
As a developer I want to make sure that type inference is performant for new and chaning types. We benchmark our critical functions for that locally now. It would be better to explicitly include this in the tests. This is beneficial to ensure the performance of types and encourages developers to always test performance. The added workload is neglectible, as performance testing should already be part of the normal workflow.