Releases: sdss/lvmgort
Releases · sdss/lvmgort
GORT 1.11.5
✨ Improved
- #55 Better messaging for alerts unavailability. A resolution alert message is now sent when the alerts become available again.
- Test Python 3.14 in CI.
🔧 Fixed
- Explicitly raise exceptions in
AGSet.reconnect().
GORT 1.11.4
✨ Improved
- Check AG cameras and power cycle them if necessary in the cleanup recipe.
- Allow to disable points upon addition to the pointing model.
🔧 Fixed
- Perform focus sweep if a periodic focus sweep is required and none has been done yet that night, even if
focus_ageis less than the periodic value.
GORT 1.11.3
✨ Improved
- Speed up AG camera power cycling.
🔧 Fixed
- Include reobserved and ancillary information in Overwatcher notifications.
GORT 1.11.2
✨ Improved
- Register the observation 5 seconds before the exposure ends to reduce overheads.
🔧 Fixed
- Fix issue with
read_yaml_fileby updating sdsstools.
GORT 1.11.1
✨ Improved
- Report Overwatcher version in the Slack message when it starts.
🔧 Fixed
- Use the guide tolerances defined in the configuration file when starting the guiders during acquisition.
- Improve handling of locked out AG cameras in the clean-up recipe.
GORT 1.11.0
✨ Improved
- #53 Add more options for focusing in the Overwatcher. The options are specified in the configuration file under
overwatcher.observer.focus. Theeverykeyword controls the frequency of focus sweeps, which can be completely disabled by setting it to zero. It's possible to require a focus sweep at least once per MJD by settingrequire_mjd_sweepto true. The optionon_erroris not currently implemented but could be used in the future to force a focus sweep if certain errors occur. - Reconnect AG cameras in the clean-up recipe.
- Force stopping the MoTAN devices when homing.
- Force telescope homing if the clean-up recipe is run more than once during troubleshooting.
- Stop observing if the clean-up recipe is run more than three times without success.
- The ping task now check if the Overwatcher main task has stalled and restarts it if necessary.
- Raise timeout errors in several places in the Overwatcher to prevent it from stalling indefinitely.
- Include reobserved and ancillary flags in the exposure headers and database.
- Say whether a tile is being reobserved or ancillary in the Slack message.
🔧 Fixed
- Fix a bug in the Overwatcher that caused the new log created when the SJD changes to be rotating at midnight UTC.
GORT 1.10.4
🔧 Fixed
- Reconnect cameras after they have been power cycled.
- Remove call to
clear_live()when clearing the progress bar.
GORT 1.10.3
✨ Improved
- Check that all cameras have been reconnected during pre-observing task.
- Implement a less verbose way to handle times when the scheduler cannot find a valid tile to observe.
- Better logging of when the dome has been actually closed during a shutdown.
GORT 1.10.2
✨ Improved
- When scheduled, run long term calibrations at least 1 hour after sunset and at lower priority then the daily calibrations.
- Always take biases in the long-term calibrations.
- Updated
sdss-cluto version 2.5.3 andlvmopstoolsto version 0.5.19.
🔧 Fixed
- Increase the timeout for the API request to email the night log.
GORT 1.10.1
✨ Improved
- Do not fail the twilight flats in the Overwatcher if the exposure time is too long for the last fibre. Instead, just log a warning and mark the calibration as done.
- Add option
clip_to_max_exposure_timeto the twilight flats recipe to clip the exposure time to the maximum defined in the configuration file. This isTrueby default.