No world frame in tf tree

edit

Thank you for your answer. I tried doing it like this in my launch file:

  <node pkg="tf2_ros" type="static_transform_publisher" name="world_broadcast" args="1 0 0 0 0 0 1 world tb3_1/odom" />
  <node pkg="tf2_ros" type="static_transform_publisher" name="world_broadcast2" args="1 0 0 0 0 0 1 world tb3_0/odom" />

And while it does work, I got some of these warnings:

[ WARN] [1633970165.290385117, 111.748000000]: TF_REPEATED_DATA ignoring data with redundant timestamp for frame tb3_0/base_footprint at time 111,705000 according to authority /gazebo
[ WARN] [1633970165.292532905, 111.748000000]: TF_REPEATED_DATA ignoring data with redundant timestamp for frame tb3_1/base_footprint at time 111,705000 according to authority /gazebo
[ WARN] [1633970165.293894016, 111.753000000]: TF_REPEATED_DATA ignoring data with redundant timestamp for frame tb3_0/base_footprint at time 111,705000 according to authority unknown_publisher
[ WARN] [1633970165.293965629, 111.75300000

They appear unrelated to the static_transform_publisher nodes as the warnings reference gazebo and base_footprint transforms. Here's a geometry2 issue discussing the warning. It may indicate you have multiple publishers of the same transforms, or one publisher publishing the same transform multiple times. Unless you suspect that the warnings are due to the static publishers, I would open a new question to address them.

I think there has to be something with the static publisher that's causing it, I've had issues like these before and they were due to problems with the tf prefixes. I fixed those by updating the ros_state_publisher and editing some files in the turtlebot gazebo model xacro to let it handle mutiple turtlebots, and the issue was completely fixed until I added the static publisher. But I'll open a new one because it could be related to the older issue and that it wasn't completely resolved

Edit: Saw that I had 1 0 0 0 0 0 1 instead of 0 0 0 0 0 0 0 (even though that didn't work so I used 0 0 0 0 0 0 1` ). For some reason I didn't get the warnings now.

I'm glad it's working! static_transform_publisher can take the transform as 6 numbers (x y z yaw pitch roll) or 7 numbers (x y z qx qy qz qw). 0 0 0 0 0 0 0 (7 zeros) would be invalid because it indicates a zero quaternion, but 0 0 0 0 0 0 (6 zeros) is valid because it indicates zero angles. That said, I'm not sure why 0 0 0 0 0 0 1 is fine but 1 0 0 0 0 0 1 produces warnings.

From the warning messages it does sound like gazebo and the static transform were publishing at the same time. What I noticed is that the warnings don't show up immedietly, only after a while, and it's not spamming like it usually does with the TF_REPEATED_DATA warnings, just 4 warning messages and that's it. But in Rviz all the frames seemed right so not sure what was wrong