The manager injects the instance name and manager location via environment variables now. So if you always start your models through the launcher, you don't need argc/argv. Maybe make them optional, or remove them? Do we want to continue to support launching the simulation through a launch script or docker-compose or something? And if the latter, can't we just use environment variables there too?
The manager injects the instance name and manager location via environment variables now. So if you always start your models through the launcher, you don't need argc/argv. Maybe make them optional, or remove them? Do we want to continue to support launching the simulation through a launch script or docker-compose or something? And if the latter, can't we just use environment variables there too?