I believe this is what is used for the Indy Autonomous Challenge?

Although their website is absolute dogsh-t, took me a fckn hour to find the goddamn download thing

Download links:

Setting up on M1 Mac

This was so painful to set up for some reason, but I finally got it:

  1. Download the x86 DMG for mac (this is like the base)
  2. Download the ARM-Clang .rtipkg thingy (this is an extension to make things work for ARM architecture)
  3. Download the license which they attached in the email (needed so you get the perms to run things)
  4. Download ConnextDDS-py so you can import rti.connextdds
  5. Set up your environment variables so it works in ROS and Python, for me I had to add: export ROS_DOMAIN_ID=0, and export RTI_LICENSE_FILE=/Applications/rti_connext_dds-6.1.1/rti_license.dat, and export RMW_IMPLEMENTATION=rmw_connext_cpp to my ~/.zshrc, figured this out from here

“RTI is literally 6 TIMES FASTER than Cyclone DDS”

So... what is the difference between ConnextDDS and CycloneDDS?

I know connextDDS is faster. But how is the underlying architecture different?


Knowledge Base


Interesting, ConnextDDS for Indy Autonomous Challenge:

Intro Video: (found from the indy page)

old version:

Okay, I can get access, that is great


Installation on M1 Mac

I run 6.1.1, and to get it working on Python on my mac, I had to checkout to v0.1.5

  • The build takes a LONG time, but it finally builttt!! Use 4 jobs to get it to build faster

This is the command that I used

python --nddshome /Applications/rti_connext_dds-6.1.1 --jobs 4 arm64Darwin20clang12.0