Entry 8 added

Author: JuanFernandezEva

docs/_posts/2024-12-8-Week9.md

@@ -0,0 +1,34 @@
+---
+title: Documentation and testing
+date: 2024-12-8 19:00:00 +0100
+tags: [weekly progress]
+author: juan
+img_path: /assets/img/
+toc: true
+comments: true
+---
+
+## Documentation and testing
+
+This week I updated the documentation for the exercise Montecarlo Visual Loc and tested the double gazebo branch mentioned in the week 3 entry.
+
+## Documentation of Montecarlo Visual Loc
+
+The first thing I did was update the documentation of the Montecarlo Visual Loc exercise. Not a lot of changes needed, just clarifications in some arguments for the functions, specifying that the coordinates used in showParticles must be in the gazebo world coordinates system. I also added the explanation of the showPosition function since it was added recently.
+
+I also added a simple script in the RoboticsAcademy-solutions to test the exercise. It just tests how the functions function, not an actual solution for the exercise, that will be implemented later.
+
+## Testing double Gazebo Docker
+
+Next I tested the branch with the double gazebo working together (Gazebo 11 and Gazebo Harmonic) since it was almost done. I built the Docker image with the build.sh script, choosing the marker_visual_loc branches and tried it. It worked fine mostly, the Gazebo Harmonic exercise worked perfectly, however, there was a problem when running the Gazebo 11 exercises, the HAL functions didn't work, so the next step was looking at what is or went wrong.
+
+## Fixing the issue
+
+The problem is that when initializing a world in gazebo 11, the robot didn't create its publishers and subscribers to use the actuators and sensors. The reason for that is that gazebo couldn't find the libraries used in the sdf files of the models to create the publishers and subscribers, they weren't installed in the image. To installed them I used the next command inside the Docker image:
+
+```bash
+sudo apt-get install ros-humble-gazebo-ros-pkgs
+```
+
+The thing is that this line was also in the Dockerfile used to build the image, so at some point the dockerfile didn't work and had to rebuild the image again.
+