4 March meeting


  • Discuss @jjt’s architecture and use cases (posted here).

Notes taken in real time:

  • We need to be able to push FF toolbar buttons programmatically.
    This is in progress.
    It would also be useful to display tool bars programmatically

  • UI windows (boxes) should be draggable and resizable.
    Perhaps the default UI layout should simply be one big image display (using FF expanded view?).
    If so, how to make the command entry window appear?
    Also, should toolbars should appear automatically?

  • Priority: Expert mode should be implemented before “casual user” mode.
    E.g., command line interface is higher priority than menus and buttons.

  • A plausible use case:

  • At start there is only an image display, filling the screen.

  • When selecting a region, the command window appears ?

  • Results windows pop-up automatically ?
    Note: FF supports pop-ups, but only in the FF window.

  • We need to make a more complete list of use cases, with time ordered implementation.
    Action: @jjt: Copy the Draft Description into Google Docs for group editing. Done Document is here.

  • Everyone should send JJT their Google ID.

  • Action: @weiren2: Get the amplifier “region” calculation working (extracted from the image headers).

  • Action: @weiren2: Overlay hot pixels on the image, using a DS9 "regions file.

  • Image formats:
    Remember that FF needs single-extension FITS.
    There are two CCD versions, one for each vendor: E2V and ITL. So, cannot use hard-wired header values.

  • We need to decide (for some commands) how to specify (for example) a file by description, not by name or location:

  • Template file.

  • The latest image.

  • We also need to support:
    . Catalog browsing. (need to look for template files) works, but is not always very efficient. The server needs to tell client what is available (either via FF or separately).
    . Image URL.