Closed GoogleCodeExporter closed 9 years ago
There doesn't seem to be any particular sequence of events that cause this.
The source of the problem is likely the motoman controller or some way in which
we are loading/unloading the inform job from motoplus.
A possible work around is check for the existence of the job when a trajectory
is received and then recreate it from a hard-coded string if necessary.
Original comment by shaun.ed...@gmail.com
on 4 Mar 2012 at 1:56
Looks like this problem may be of my own doing or left over from Somudro's
work. The mpDeleteJob function is called from within joint_motion_handler.cpp
(http://code.google.com/p/swri-ros-pkg/source/browse/trunk/industrial/motoman/mo
toplus/motoros_lib/joint_motion_handler.cpp#230 ) The desired behavior is to
stop the job and reset the program/motion pointer. Ted Miller (motoman)
suggests the following:
First call mpHold will stop your job. You should then wait about 4-8 msec and
turn mpHold off (this will just disable the HOLD signal, it will not
automatically restart your job). Then you can reset the current job (and line
number) with mpSetCurJob. Then call mpStartJob to begin motion.
FYI...The same code is found in controller.cpp (
http://code.google.com/p/swri-ros-pkg/source/browse/trunk/industrial/motoman/mot
oplus/motoros_lib/controller.cpp#153 )
Original comment by shaun.ed...@gmail.com
on 20 Mar 2012 at 9:06
Fixed problem in joint_motion_handler. Still needs to be fixed in
controller.cpp.
Original comment by shaun.ed...@gmail.com
on 22 Mar 2012 at 6:54
Closing this issue because the existing dx100 c++ code is being retired.
Original comment by shaun.ed...@gmail.com
on 29 Jul 2013 at 2:19
Original issue reported on code.google.com by
ger...@willowgarage.com
on 27 Feb 2012 at 8:31