Revision history [back]

The main function of rospack is to crawl through the packages in ROS_ROOT and ROS_PACKAGE_PATH, read and parse the manifest.xml for each package, and assemble a complete dependency tree for all packages.

And more specifically for rospack profile:

rospack’s performance can be adversely affected by the presence of very broad and/or deep directory structures that don’t contain manifest files. If such directories are in rospack’s search path, it can spend a lot of time crawling them only to discover that there are no packages to be found. You can prevent this latency by creating a rospack_nosubdirs file in such directories. If rospack seems to be running annoyingly slowly, you can use the profile command, which will print out the 20 slowest trees to crawl (or use profile --length=N to print the slowest N trees).

So the directories without *'s are the one that contain packages.

This can all easily be found at the rospack documentation

The main function of rospack is to crawl through the packages in ROS_ROOT and ROS_PACKAGE_PATH, read and parse the manifest.xml for each package, and assemble a complete dependency tree for all packages.

And more specifically for rospack profile:

rospack’s performance can be adversely affected by the presence of very broad and/or deep directory structures that don’t contain manifest files. If such directories are in rospack’s search path, it can spend a lot of time crawling them only to discover that there are no packages to be found. You can prevent this latency by creating a rospack_nosubdirs file in such directories. If rospack seems to be running annoyingly slowly, you can use the profile command, which will print out the 20 slowest trees to crawl (or use profile --length=N to print the slowest N trees).

So the directories without *'s are the one ones that contain packages.

This can all easily be found at the rospack documentation