bwi_kr_execution: upward dependencies on segbot_logical_translator

[jack-oquin]

These are launch file dependencies. They would be caught if the package had a launch check unit test, but it does not. The references are in: bwi_turtlebot_launch and robot.launch.

@piyushk: the segbot_logical_translator appears to be a pure BWI component with no dependency on anything specific to the segbot platform. If so, I recommend moving it to bwi_common and renaming it bwi_logical_translator. The only references I can find are in bwi_launch and here in bwi_kr_execution.

[jack-oquin]

@piyushk: Isn't this issue fixed? If so, let's close it.

[piyushk]

@jack-oquin it's not been fixed. segbot_logical_translator still exists in the segbot repo, and needs to be moved here.

[jack-oquin]

Taking this up again, xref: #100, utexas-bwi/segbot#96

[jack-oquin]

This issue has become a problem for sharing code with the RoboCup@Home project.

[rdelfin]

@jack-oquin are you looking into this already? I look into it as well

[jack-oquin]

Yes, I assigned myself to fix it. Right now I'm planning to do that in the kinetic branch, because I expect that to receive more testing and impact fewer people in the short run. Does that seem reasonable? At some point, it may be necessary to back-port the fix to master.

I'd appreciate any testing help you can provide once I commit a fix.

[rdelfin]

Makes sense. We can try and move it in after. I'll wait for a commit and test it then

[jack-oquin]

Thanks, Ricardo!

I plan to come to the lab this afternoon for Jesse's practice talk. I'll look for you then.

[rdelfin]

Ah, I'm not in Austin at the moment, I was thinking either over email or hangouts (it would have to be in a Kinetic machine anyways, so not the main robots)

[jack-oquin]

I just committed the bwi_common part of the fix, bwi_launch still to go.

[jack-oquin]

With a quick test, simulation.launch seems to run visit_door_list.

[jack-oquin]

Roberto is currently running visit_door_list. It's been going for a while without observable problems, except that the wireless internet connection seems to have died. That does not stop the robot, but will probably cause the log files not to be saved automatically (this time).

I rebooted and saved the logs. The robot ran for a little over an hour, travelling 1.43 km.

[jack-oquin]

@nickswalker managed to get the arm robot to move with kinetic. It had xenial installed, but no display so hard to do full navigation.

I think it's time to dual boot one of the segbot_v2 robots, and try full navigation.

[jack-oquin]

I believe this is working now in the kinetic branches. Since RoboCup@Home is now migrating to Kinetic, that seems sufficient for our current needs.

I'd be happy to avoid back-porting a bunch of this stuff to Indigo, if it's not needed.