Possible bug in Network Tables?

We encountered a repeatable byte-ordering bug in the following situation.

We have a custom java widget integrated into the smart dashboard which sends a double to the robot in the network tables. We can verify the number we’re sending because we also display the number.

On the robot, in C++ code, we receive the number from the network tables and then just send it back to the smart dashboard to make sure we have the data pipeline working.

When the number gets back to the smartdashboard (extremely large), it is corrupted; my guess is the byte order got swapped somewhere.

We switched to sending and recieving integers and that works great so it seems like the bug is in the code that handles doubles specifically and maybe its in the C++ recieving code on the cRio. I believe we are sending other doubles from C++ fine.