D29355: Fix issue with Meridian Flip Retry

[Chris Rowland]

chrisrowland added a comment.


  The general comments  are a great help in understanding what this does but don't address what I think is needed, more description in the code of what is being done and why.  The comments here are ephemeral and in six months time when somebody else needs to work on this code they will not be available.
  
  Can some of the description of what this does be added to the source?  Thge comments hereIt will then be of use to the next person who needs to work on this.
  
  As for the overall design of this fix what seems to have happened was that in some cases the capture module would cancel a request from the mount to pause acquisition instead of leaving the request in place until the capture was in a position to pause.  The fix is to add a sort of half state in the mount where a pause request is delayed if the capture module is not in a state to accept it, then, when the capture module is in a state to accept a pause request, to send it.  This sort of merges the mount meridian flip and capture meridian flip state machines.
  
  I am uneasy about this, it seems to be adding a special case to manage another special case, a band aid on a band aid.  Is there a way that the original problem - that Capture had decided to cancel a meridian flip request - be fixed?
  
  I don't feel that I'm competent to decide that this is OK.  The capture and mount state machines are so complex that I can't tell.   I understand that it hasn't failed any testing so far and maybe that's the best that we can say.

REPOSITORY
  R321 KStars

REVISION DETAIL
  https://phabricator.kde.org/D29355

To: murveit, chrisrowland, mutlaqja
Cc: kde-edu, narvaez, apol
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kde-edu/attachments/20200504/4a4cd381/attachment.html>