TODOs from 5/12 MRO trip

[albireox]

• Evora doesn't seem to expose for more than 30 seconds. Maybe related to gunicorn workers timing out.
  • Solved but needs more testing.
• gunicorn workers time out, check if that can be configured or use development mode for now.
• Install and test new evora-wheel
• Filter wheel seems to not home in certain positions of the telescope and the filters are not well aligned sometimes. Test movement at different positions of the telescope.
  • We found that the separation between the hall sensors and the filter plate is important and if the sensors are too close, they produce false positives. Need to decide whether to keep the spacer washers we installed or reduce the size of the steel spacers.
  • Need to test if this solves the issue at different position angles of the telescope.
• Set NTP clients in Loki and Sleipner.
• Figure out how to ssh to computers from internal/external network.
• Evora: bias exposure time seems to open the shutter and exposes the entire exposure time (+0.5s)
• Fix setup.py compilation of andor_wrapper
• Write documentation about the webapp.
• UWMRO/evora-client#21 Progress bar for exposing. Requires server route to determine how long the exposure has been going.
• Add buttons to initialise and shut down the camera.
• Allow to disable cooling and make sure it remains off after sending shutdown to the camera.
• Add instructions to set up RPi with new filter wheel server.
• Move Sleipner to Loki's monitor?
• Sleipner should not go to sleep if it's going to run the Evora server.
• Server should allow to control the filter wheel even if camera is off.
• Fix filter ordering in evora-client.
  • See #15 and UWMRO/evora-client#13
• Tag all products.
• #18 evora-server reads the camera buffer as signed 32-bit integers. There's a function that uses unsigned 16-bit integers, which seems more reasonable. It's worth testing it.
• Temperature would sometimes would get lost after an exposure (?)
• Add i-band filter.
• Implement real-time exposures.