Closed Issues
- Proposal: Reduce Pipeline.map's IPC overhead with chunking ([484](https://github.com/pipefunc/pipefunc/issues/484))
๐งช Testing
- Explicitly set reason in `skipif` ([499](https://github.com/pipefunc/pipefunc/pull/499))
- Skip shared memory test in CI on nogil Python (3.13t) ([498](https://github.com/pipefunc/pipefunc/pull/498))
โจ Enhancements
- Allow providing an int to `chunksizes` ([497](https://github.com/pipefunc/pipefunc/pull/497))
- Add `chunksizes` argument to `Pipeline.map` and `Pipeline.map_async` ([493](https://github.com/pipefunc/pipefunc/pull/493))
๐ค CI
- Revert `pull_request_target:` for CodSpeed ([495](https://github.com/pipefunc/pipefunc/pull/495))
- Use `pull_request_target:` to trigger CI on fork ([494](https://github.com/pipefunc/pipefunc/pull/494))
๐ Documentation
- Mention HPC vs cloud based running ([492](https://github.com/pipefunc/pipefunc/pull/492))
- How is this different from Dask, AiiDA, Luigi, Prefect, Kedro, Apache Airflow, etc.? ([491](https://github.com/pipefunc/pipefunc/pull/491))
- Add Discord shield ([490](https://github.com/pipefunc/pipefunc/pull/490))
๐ Stats
- `.yml`: +2 lines, -2 lines
- `.md`: +62 lines, -1 lines
- `.py`: +172 lines, -8 lines