But in 6 months from now it will be a matter of starting the client and connecting to one of the servers showing in your roster, a step forward compared to having to mind the protocol to use, the IP address, etc. These functions are not available yet (mDNS discovery is in testing and will be part of the 4.1, planned for the end of the year, while the directory service and NAT traversal will take a bit more), so I understand that, compared to the simple combo box of the 3 it shows more information and thus can appear more complicated. New computers are automatically showing in the roster based on mDNS discovery or a global directory and NAT traversal service we currently have under development. All the new GUI is designed to be 'dynamic'. Then there is the initial GUI, by which you must choose the computer to connect to. It is more configurable and so more 'complex', but this sense of complexity should disappear as soon as you start to use it.

The biggest change in version 4 is that you can do all the configuration you did in the old client at run-time, while in version 3 you had to take a number of decisions before connecting to the remote server and, once connected, you could do very little. I'd just like to understand how the new version is more complicated compared to the old version, so that we can improve it. I'm not going to praise NoMachine if you are not comfortable with closed source software.

Is nomachine free