During our Melbourne meetup and fly today, in which we did several autonomous flights with several quads, we found that there is potential for unusual/unexpected/unwanted behaviour at the conclusion of a mission, if you leave the last waypoint type set to WAYPOINT. Basically the vehicle hits it's target and then doesn't know what to do, so if you're not paying attention, you can get into trouble.
Now, you may already have something built-in to deal with this operator deficiency, but a suggestion (if you don't) would be to run a quick verification on the final waypoint when you press "Write WPs". Checking that it's an unlimited loiter, or land (or some other appropriate choice... and not simply WAYPOINT) and providing a dialog box warning of a non-terminal state for the plan would be good. Here I mean a terminal state is one that is considered stable for all time (so landing, or unlimited loiter). Does this make sense?
Also, is there a simply way to make the plan continuously cyclic... e.g., WP1, WP2,...,WPn, WP1, WP2,..., etc?
it is meant to switch into loiter, if the last wp is a WAYPOINT ,depending on what rtl_alt_final is set as. if its set as 0, it will land, if above that it will loiter at that altitude. the default is 2m, so it will go into loiter and ascend/descend to 2m
To make it go around and around you need to use a DO_JUMP, and set the wp target to 1 and count to say 99, or similar.
Thanks for the quick reply Michael... I'll need to check what value rtl_alt_final is set at... but my copter definitely didn't loiter at its final point... and it had been okay at loiter within the plan (it had a 10s duration loiter within a 3m circle that it handled serveral times without problem). I'll look into this further. Thanks for the info and advice.