The upper diagram shows the FLC software tasks and dataflows as presented at the CoDR. The lower diagram includes changes discussed at this meeting, some of which are to support our component and integrated tests. Software requirements for the tests are summarized in a separate report.

The upper diagram shows the FLC software tasks and dataflows as presented at the CoDR. The lower diagram includes changes discussed at this meeting, some of which are for our component and integrated tests. Software requirements for the tests are summarized in a separate report.

This meeting took place at 14:00 on Wednesday 27 October. Attendees: JSY, MF, EBS.

• GraphicalInterface will be implemented using C/GTK+ or C++/gtkmm

Implementation Details

• Expect to use threads for multi-tasking, with shared data (no ZeroMQ). Thread locking issues not anticipated to be excessively complex
• Main FLC process will have built-in image display capability for debugging, implemented using X11 libraries

The upper diagram shows the FLC software tasks and dataflows as presented at the CoDR. The lower diagram includes changes discussed at this meeting.

• Cambridge-supplied LocalDataCollector
• ISS Data Collector probably not suitable as LocalDataCollector (e.g. issues with retrieval of recorded data), and unlikely to be ready in time

• Will not be present in standalone role (neither will ISS Executive/Supervisor)

• TempController (EnvController?) has interface to ISS
  • Allows ISS to query state for this process
• Division of tasks between main process and TempController depends on I/O device sharing issues - to be investigated
  • TempController could control power on/off for camera ("?" on diagram)

• Main FLC process will have built-in image display capability using X11 libraries
• We will write a Data Analysis GUI in Matlab (not shown)

• GraphicalInterface will still include commonly-used diagnostic displays

• Expect to use dlmsg protocols for publishing to LocalDataCollector, with recording to DL-style FITS files
  • Facilitates collection of metrology data from DL metrology subsystem
  • Can use DL workstation software as interim LocalDataCollector

• EBS: Investigate ADC/DAC/digital I/O interface cards and device sharing issues

Changes from CoDR

• GraphicalInterface and LocalDataCollector grouped into single process
  • ISS Data Collector probably not suitable as LocalDataCollector and unlikely to be ready in time
• ISS Data Collector shown
  • Will not be present in standalone role
• TempController has interface to ISS
  • Allows ISS to query state
• TempController could control power on/off of camera ("?" on diagram)

Notes

• Matlab Data Analysis GUI not shown
  • Reads files recorded by LocalDataCollector

JSY is writing up conclusions.

• EBS: Investigate ADC/DAC/digital I/O interface cards
  • NB environmental control performed by dedicated process
• EBS: Investigate whether floating point is allowed in Xenomai real-time code
• MF: Try reading example FITS image time-series into Matlab

Discussion slides

Slides as PDF

• NB thermal control performed by dedicated process

• Software requirements for our testing
  • Image display for setup purposes
  • Temperature logging
  • Centroid logging
  • Anything else?
• Open Design Issues
  • Choose ISS interface flavour: "C" (threads/processes)/"C-no-threads" (event loop)
  • Can we use floating point in Xenomai kernel-space real-time?
  • How to transmit dark/flatfield frames to/from ISS?
  • How to receive off-axis/dispersion offsets from UTCS via ISS?
    • May need to account for future off-axis offsets when choosing initial position of subframe on FTT Mode start
  • Choose GUI toolkit
• Plan for camera loan

• Select ADC, DAC interface cards
  • NB thermal control performed by separate process

Technical Meeting #12: Software design

14:00 Wed 27 Oct, Ryle.

Agenda

Conclusions

New Tasks