D21531: Bug fix in Scheduler for pausing after capturing sequence
noreply at phabricator.kde.org
Mon Jun 3 07:27:43 BST 2019
TallFurryMan added a comment.
First I need to say I haven't built your differential, so my comment is from code and explanations only. So I'm not sure I understand the diff properly.
Originally, the pause button was aborting the capture, but kept guiding running. I don't know precisely what the historical implementation wanted, but I assumed the idea was to stop the scheduler, fix things in the observatory manually, then continue. Examples would be adjusting focus, restarting a capture that would be thrown away or simply fixing cable mess before it would interfere. Your differential now has the scheduler finish the capture before pausing. To me this is a drastic change in behavior which requires clarification. I can't confirm the examples I gave are still manageable with the change you provide.
Second concern is about Scheduler being paused or aborted by the Observatory module: this is not what we agreed on. We wanted Scheduler to remain the orchestrator of the observation. As such, Observatory had to notify states, and Scheduler had to manage those states. Unless there was a safety issue, in which case Observatory still notified the emergency but took action directly in its own domain.
More on this later on when I have time :(
To: wreissenberger, mutlaqja, TallFurryMan
Cc: kde-edu, narvaez, apol
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the kde-edu