Add build for rmf-dev package

[Ac31415]

[traversaro]

There are a few transitive dependency of rmf-dev, and so a few packages are built that are failing on all platforms. Let us know if you are interested in working on this and if you would like some tips on how to debug.

[Ac31415]

There are a few transitive dependency of rmf-dev, and so a few packages are built that are failing on all platforms. Let us know if you are interested in working on this and if you would like some tips on how to debug.

Hi! Yeah I'm interested in working in this. But my main focus will be on osx-arm64 for now. I've been getting errors mainly from the 'rmf_traffics' module. Copilot has been telling me it's due to scripts such as 'Database.cpp' having the 'final' tag at the end of some methods. But I'm not sure if what it's saying is exactly true. So I'd really appreciate some tips on debugging!

[traversaro]

Ok! First of all, you can always start by copying the error you want to tackle in an comment, so the error on osx-arm64 :

2025-05-21T20:05:00.2469210Z  │ │ In file included from $SRC_DIR/ros-humble-rmf-traffic/src/work/src/rmf_traffic/schedule/Database.cpp:22:
2025-05-21T20:05:00.2469920Z  │ │ In file included from $SRC_DIR/ros-humble-rmf-traffic/src/work/src/rmf_traffic/schedule/debug_Database.hpp:21:
2025-05-21T20:05:00.2470680Z  │ │ $SRC_DIR/ros-humble-rmf-traffic/src/work/include/rmf_traffic/schedule/Database.hpp:116:39: error: non-virtual member function marked 'final' hides virtual 
2025-05-21T20:05:00.2471260Z  │ │ member function
2025-05-21T20:05:00.2472890Z  │ │   116 |     std::size_t participant_id) const final;
2025-05-21T20:05:00.2475770Z  │ │       |                                       ^
2025-05-21T20:05:00.2476580Z  │ │ $SRC_DIR/ros-humble-rmf-traffic/src/work/include/rmf_traffic/schedule/Viewer.hpp:104:57: note: hidden overloaded virtual function 'rmf_traffic::schedule::V
2025-05-21T20:05:00.2477690Z  │ │ iewer::get_participant' declared here: type mismatch at 1st parameter ('ParticipantId' (aka 'unsigned long long') vs 'std::size_t' (aka 'unsigned long'))
2025-05-21T20:05:00.2597340Z  │ │   104 |   virtual std::shared_ptr<const ParticipantDescription> get_participant(
2025-05-21T20:05:00.2702790Z  │ │       |                                                         ^
2025-05-21T20:05:00.2838060Z  │ │ In file included from $SRC_DIR/ros-humble-rmf-traffic/src/work/src/rmf_traffic/schedule/Database.cpp:22:
2025-05-21T20:05:00.2954040Z  │ │ In file included from $SRC_DIR/ros-humble-rmf-traffic/src/work/src/rmf_traffic/schedule/debug_Database.hpp:21:
2025-05-21T20:05:00.3057280Z  │ │ $SRC_DIR/ros-humble-rmf-traffic/src/work/include/rmf_traffic/schedule/Database.hpp:128:39: error: non-virtual member function marked 'final' hides virtual 
2025-05-21T20:05:00.3177090Z  │ │ member function
2025-05-21T20:05:00.3279810Z  │ │   128 |     std::size_t participant_id) const final;
2025-05-21T20:05:00.3391260Z  │ │       |                                       ^
2025-05-21T20:05:00.3528720Z  │ │ $SRC_DIR/ros-humble-rmf-traffic/src/work/include/rmf_traffic/schedule/Viewer.hpp:134:40: note: hidden overloaded virtual function 'rmf_traffic::schedule::I
2025-05-21T20:05:00.3642610Z  │ │ tineraryViewer::get_itinerary' declared here: type mismatch at 1st parameter ('ParticipantId' (aka 'unsigned long long') vs 'std::size_t' (aka 'unsigned lo
2025-05-21T20:05:00.3749400Z  │ │ ng'))
2025-05-21T20:05:00.3858190Z  │ │   134 |   virtual std::optional<ItineraryView> get_itinerary(
2025-05-21T20:05:00.3961800Z  │ │       |                                        ^
2025-05-21T20:05:00.4065890Z  │ │ In file included from $SRC_DIR/ros-humble-rmf-traffic/src/work/src/rmf_traffic/schedule/Database.cpp:22:
2025-05-21T20:05:00.4170090Z  │ │ In file included from $SRC_DIR/ros-humble-rmf-traffic/src/work/src/rmf_traffic/schedule/debug_Database.hpp:21:
2025-05-21T20:05:00.4294560Z  │ │ $SRC_DIR/ros-humble-rmf-traffic/src/work/include/rmf_traffic/schedule/Database.hpp:132:39: error: non-virtual member function marked 'final' hides virtual 
2025-05-21T20:05:00.4296110Z  │ │ member function
2025-05-21T20:05:00.4419070Z  │ │   132 |     std::size_t participant_id) const final;
2025-05-21T20:05:00.4527210Z  │ │       |                                       ^
2025-05-21T20:05:00.4631550Z  │ │ $SRC_DIR/ros-humble-rmf-traffic/src/work/include/rmf_traffic/schedule/Viewer.hpp:140:33: note: hidden overloaded virtual function 'rmf_traffic::schedule::I
2025-05-21T20:05:00.4743490Z  │ │ tineraryViewer::get_current_plan_id' declared here: type mismatch at 1st parameter ('ParticipantId' (aka 'unsigned long long') vs 'std::size_t' (aka 'unsig
2025-05-21T20:05:00.4845070Z  │ │ ned long'))
2025-05-21T20:05:00.4948510Z  │ │   140 |   virtual std::optional<PlanId> get_current_plan_id(
2025-05-21T20:05:00.5058110Z  │ │       |                                 ^
2025-05-21T20:05:00.5174220Z  │ │ 1 warning and 3 errors generated.

This is helpful (as long as the test that you copy is not so long) so other users do not need to dig in the log to actually find the error, and also as log over time expire and get deleted, while comments remain and can be found by search engines and llms.

[Ac31415]

Great! What if the error log is long? Is there a file where the complete error is logged?

[traversaro]

Great! What if the error log is long? Is there a file where the complete error is logged?

Not automatically. IF the error log is long and you want to save it permanently, I would suggest to save it to a gist and link it in an issue. Going back to the actual build error, it seems that the problem is a mismatch between PartecipantId (type used in the interface) and std::size_t (type used in the implementation in child classes). It seems all these problems are fixed in the latest version of rmf-traffic, but the humble version is the old 3.0.3 one (see https://github.com/ros/rosdistro/blob/bbc6dd676fa320ac7f0eda7369ce599241997129/humble/distribution.yaml#L8138 and https://github.com/open-rmf/rmf_traffic/blob/3.0.3/rmf_traffic/include/rmf_traffic/schedule/Database.hpp#L116) . It seems that the fix was in the recent PR https://github.com/open-rmf/rmf_traffic/pull/98, probably we can try to add this patch locally and see if this fixes the PR. See https://robostack.github.io/Contributing.html for more info on how to create or add a patch, thanks!

For referece, see also this ChatGPT conversation, that seems to correctly frame the problem, even if it seems that it hallucinated the version that actually contains the fix: https://chatgpt.com/share/6831e6ca-ae34-8006-81bb-e0c40dc2f800 .

[Ac31415]

Great! I'll definitely take some time to look at these. Thanks for the resources. Please also excuse me as this is my first time contributing to a public project. And I'm still taking the time familiarizing myself with the tools haha

[Ac31415]

Yeah I see it now. I'll definitely try to patch that PR locally and see what I get. Will definitely try that after I'm done with the chores in the morning

[traversaro]

Great! I'll definitely take some time to look at these. Thanks for the resources. Please also excuse me as this is my first time contributing to a public project. And I'm still taking the time familiarizing myself with the tools haha

Sure, no problem, the learning curve for contributions can be steep, feel free to ask questions if you have doubts.

[Ac31415]

Thanks so much! Will let you know