Fix compile erros when using RoboStack (conda)

[nberliner]

The RoboStack maintainers helped to fix compile errors when building the UR drivers using a conda environment. conda is, in my humble opinion, very useful and the suggested changes might be helpful for others.

For more information, please see RoboStack/ros-noetic#155 and RoboStack/ros-noetic#157 .

[fmauch]

This seems unusual at first.

The console_bridge issues should probably be gone with UniversalRobots/Universal_Robots_Client_Library#74 merged.

As for the -D__STDC_FORMAT_MACROS definition: Why is this necessary?

[nberliner]

Hi, thanks for having a look. To be honest, I don't know what -D__STDC_FORMAT_MACROS does or why it is necessary. It was suggested to me by the maintainer of RoboStack here.

Maybe you could chime in on RoboStack/ros-noetic#157 to check where this fits best?

[gavanderhoorn]

@fmauch wrote:

As for the -D__STDC_FORMAT_MACROS definition: Why is this necessary?

From How to printf uint64_t? Fails with: “spurious trailing ‘%’ in format”:

The ISO C99 standard specifies that these macros must only be defined if explicitly requested.

#define __STDC_FORMAT_MACROS
#include <inttypes.h>

... now PRIu64 will work

[fmauch]

Thanks @gavanderhoorn

From my point of view none of the changes here should then be necessary with the latest versions.

@nberliner could you try building the driver without the changes with the latest library version (master or boost branch) and the latest master of this driver?

See the README about building the library from source, as well.

[nberliner]

I've tried to build the driver as well as ur_client_library from source. Unfortunately, building the ur_client_library within the RoboStack conda environment seems to be rather complicated (see here) and catkin_make_isolated failed. I probably need to wait until the RoboStack maintainers include the updates.

Please feel free to close this PR, it's definitely beyond my expertise.

Thanks for looking into it :)

[github-actions]

This PR hasn't made any progress for quite some time and will be closed soon. Please comment if it is still relevant.

[fmauch]

I should have a look at this again, before closing this.

[github-actions]

This PR hasn't made any progress for quite some time and will be closed soon. Please comment if it is still relevant.

[fmauch]

Sorry for not having answered for so long. I am still unsure whether I would like to add something that "only" makes it build on an unsupported build system.

The D__STDC_FORMAT_MACROS thing seems to be something that should be fixed inside the conda builds as suggested in RoboStack/ros-noetic#157.

If we could change the console_bridge dependency to a ordinary catkin dependency to rosconsole and not explicitly link against that but let ${catkin_LIBRARIES} handle this, that would be great. Unfortunately, this is out of my scope to test.

[fmauch]

It seems to make more sense to rely on rosconsole since we use ros/console.h inside the log handler.