Vladimir Ermakov

First byte 0xFD is a framing marker of MAVLink version 1, 0xFE - MAVLink version 2. Version 2 has bigger header: https://mavlink.io/en/guide/serialization.html [MAVLink 2](https://mavlink.io/en/guide/mavlink_2.html) was in all major APs for...

https://github.com/mavlink/mavros/blob/master/mavros/src/lib/mavros.cpp#L58 code still there. Could you please check again that you set `/mavros/fcu_protocol := "v1.0"` before node startup?

Some of static transforms are hardcoded into the node, some just requires config changes. Please note that tf_prefix param support dropped while ago from tf lib.

I think here need to check in packet level. My suggestion that server protocol might be changed. Can you try to test older firmware?

Ok, then next thing to analyze - diff 1.12.3 and 1.13.0 (in mavlink app).

Try to check "mavlink_ftp" module. But yes, looks like there a lot of changes...

I do not see any trouble on mavros side. Probably you need to change some firmware parameters to select mocap position source.

Sorry, but i never tried to setup it. So i mostly guessing. Perhaps you could find something useful here: https://docs.px4.io/v1.12/en/ros/external_position_estimation.html Just one common thing to check: do you update header.stamp?

When it is possible, sys_time does timesync between FCU and OBC. So mavlink messages with timestamp fields will be relatively close to FCU's time. You can also use ntpd_driver package...

I do not know why, but sys_time reports that RTT is too high to operate correctly.