Release minification started and the build process was killed almost immediately under memory pressure. A smaller heap without a daemon was considered, but the release package was never produced. Debug compilation had already succeeded, so the kill did not point to a shrinker defect in the map changes.
- What got in the way
- On a machine with about 3.8 GB of RAM and a 2 GB build heap, the process was killed shortly after minification started. No shrink report or release package was produced, and another attempt looked likely to fail the same way.
