pymmcore-plus/pymmcore-plus - CodSpeed

pymmcore-plus

Blog Docs Changelog

Performance History

Latest Results

refactor: bump ome-writers to use new global_metadata feature (#615) * refactor: Update summary metadata handling in OmeWritersSink and tests * undo example change * fix: Update references from set_summary_metadata to set_global_metadata in OmeWritersSink and related tests * fix: Update ome-writers dependency version to 0.3.2 in pyproject.toml * remove global * bump useq

main

2 days ago

fix: ensure data sink is closed on KeyboardInterrupt Move `_finish_run` into a `finally` block so the data sink is always closed, even when the acquisition is interrupted with Ctrl+C. Previously, `_finish_run` (which calls `sink.close()`) was called after the `except Exception` block but not inside a `finally`. Since `KeyboardInterrupt` inherits from `BaseException`, not `Exception`, an interrupt would skip `_finish_run` entirely, leaving the OMEStream open. This caused an "OMEStream was not closed before garbage collection" warning and risked data loss. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

ieivanov:fix/close-sink-on-keyboard-interrupt

6 days ago

refactor: bump ome-writers to use new global_metadata feature (#615) * refactor: Update summary metadata handling in OmeWritersSink and tests * undo example change * fix: Update references from set_summary_metadata to set_global_metadata in OmeWritersSink and related tests * fix: Update ome-writers dependency version to 0.3.2 in pyproject.toml * remove global * bump useq

main

8 days ago

tlambert03:clean-summary

9 days ago

tlambert03:clean-summary

9 days ago

fix: Update ome-writers dependency version to 0.3.2 in pyproject.toml

tlambert03:clean-summary

9 days ago

feat: Add optional rich logging via PYMM_LOG_RICH env var (#613) * feat: add optional rich logging via PYMM_LOG_RICH env var Re-introduce rich logging support as an opt-in feature controlled by the PYMM_LOG_RICH environment variable. When set to 1/true/yes and rich is installed, stderr logging uses RichHandler instead of the plain formatter. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * add note that rich formatting adds overhead --------- Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

main

9 days ago

feat: add per-device waitForDevice timeout and timeout registry Add a `timeout_ms` keyword argument to `CMMCorePlus.waitForDevice()` that overrides the global `getTimeoutMs()` for a single call. When set, the wait uses a Python-side polling loop (10 ms cadence, matching MMCore's `pollingIntervalMs_`) and raises `TimeoutError` on expiry. `timeout_ms=0` checks `deviceBusy()` once and fast-fails if still busy. Also adds a per-device timeout registry via `setDeviceTimeoutMs()` / `getDeviceTimeoutMs()`, and overrides `waitForSystem()` to iterate devices in Python so registry entries are honoured automatically. Timeout resolution order: per-call kwarg > registry > global C++ path. Propagates through `UniMMCore.waitForDevice` for Python devices. UniMMCore.waitForSystem waits for Python devices in parallel via ThreadPoolExecutor, matching the pre-existing behavior.

hinderling:feat/wait-for-device-timeout

12 days ago

Latest Branches

0%

fix: ensure data sink is closed on KeyboardInterrupt#616

6 days ago

6238ff0

ieivanov:fix/close-sink-on-keyboard-interrupt

0%

refactor: bump ome-writers to use new global_metadata feature#615

9 days ago

2a27f97

tlambert03:clean-summary

0%

feat: per-device waitForDevice timeout and timeout registry#614

12 days ago

3ed181d

hinderling:feat/wait-for-device-timeout

© 2026 CodSpeed Technology

Home Terms Privacy Docs