@arthurcolle@teknium They haven't been putting up any of the new Mistral or Llama models on launch day at all, no community fine tunes, but instead testing countless prerelease ones from OAI with obfuscated names. They're basically their PR subsidiary now and the pretense is slowly being dropped.
@abemurray@suzannegildert You could actually prove it if you had concrete insight into the internal state and observed the inner thoughts and emotions. Impossible for people, but possible for software. It's actually funny that machines might be one day proven sentient before any living organism.
@jwt0625 Do yourself a favor and get the M9N clone instead, it costs a few bucks more but can see more constellations and has a much higher accuracy. All of my outdoor robots run them, they're fantastic.
@roboticseabass The real question is why aren't those arguments the default for just "rosdep install", since that full line is the only rosdep command most people will ever run. Really antipragmatic design and bad UX imo.
@lekkerkerkerder Ah no it's mainly for the TF links, I don't currently really have a gazebo setup for it since well... I can just test on the real thing. That's where the bumpers come in ;)
@lekkerkerkerder Yep, the lidar's the ToF LD-19 which works in full sunlight ;) I bought just the bare chassis since I wanted to develop my own motor controller for it with a Pico and DRV8871s, I think I only really used the original urdf, the rest of the ROS 2 part is custom as well.
@A_Reichenbach_@audrow I really don't get why they didn't reuse the parts of ROS 1 that were solid, rclpy runs 20x slower than rospy and ROS 2 based the entire cli on it๐คช But at least we can now pick which half broken DDS we want to use, yay
@OpenRoboticsOrg @CubedUnit @audrow Well given that most of the complexity increase in ROS 2 is to cater specifically to industrial clients, perhaps it's on them to help with documenting it in return.