What version of Ardurover is currently running on the latest image from Emlid?

[esa.attia]

Hi,
What is the base of the ardurover code in the latest Raspbian image (Stretch) released by Emlid?

Looking inside ardurover code ( strings /opt/ardupilot/navio2/ardurover-3.2/bin/ardurover | grep ‘APM:Rover’ ), the git commit is 90e8ac09. However, there are no such commit in GitHub - ArduPilot/ardupilot: ArduPlane, ArduCopter, ArduRover, ArduSub source

My question is, has the Emlid team made any modifications to the Ardurover code in order for it to work with their hardware?

We are currently having issues between Mavros and the Ardurover version 3.2. The only way we could make it work the way we want (being able to send forward velocity and angular velocity simultaneously ) is if we modified the source. It would be nice to know if the version of ardurover on the current image is a different branch to that which is on github.

Cheers

[george.staroselskiy]

Hello there!

Here it is. As you can see it has only a couple of commits backported from master.

@esa.attia

We are currently having issues between Mavros and the Ardurover version 3.2. The only way we could make it work the way we want (being able to send forward velocity and angular velocity simultaneously ) is if we modified the source.

This is a very interesting problem. I hope we’ll be hearing from you when you crack it.

[esa.attia]

Hi George,

Thanks for the link!

What are these additional commits in relation to? What is the advantage of using your branch of ardurover vs the master? Is there anything specific?

Currently, we are able to make ardurover work with Mavros by hacking the latest version of ardurover3.2 (from the master). What my team and I would like to know is if it would be better if we made changes to the version you posted and work from that for now as oppose to the master if it means that the system will work better.

There is a more detailed description of the problem and how we solved the problem (for now) on the ardurover forum.

See - Mavros commands ignored by ardurover

What I am hoping is that these changes can be filtered into the next RC release of ardurover and tested by others of course. For my project, I would really like to run the ardurover as stock without making my own fork of the code.

[george.staroselskiy]

@esa.attia

What are these additional commits in relation to? What is the advantage of using your branch of ardurover vs the master? Is there anything specific?

This release was based on the latest APMRover release, so we believe this code has been tested enough to base your code on.

The backported commits’ messages say it all, actually. They fix small bugs that can be spotted only rarely like this scenario: you launch ardurover, wait for it to do its job, kill the process. You’ll see that ardurover quit erroneously with double free() and so on. These commits fix these issues.

@esa.attia

What I am hoping is that these changes can be filtered into the next RC release of ardurover and tested by others of course.

So, I hope we’ll see your PR in ArduPilot pretty soon!

[esa.attia]

Hi @george.staroselskiy,

Do you think it is worthwhile for you to submit your commits to the ardurover team to include it in the release version of ardurover 3.2?

That way we don’t have 2 different forks of the same code? Based on activity on the ardu rover forum it seems that there will be a lot of stability issues resolved in the next release.