Thin client computing has been around for a long time in the UNIX world. Although the implementation has evolved quite a bit, the concept is the same:

Because the clients do very little, the hardware can be small and cheap. The clients themselves are basicly maintenance-free. They last longer because they have no moving parts like hard disk drives. Furthermore, if they break no data is lost because nothing is stored on the client itself. Just swap the client with another one and go back to work. Clients get stolen or put at the trash? No data ends up on the streets.

The terminal server runs all applications and contains all the data. All the regular maintaince (software updates, administration) takes place at the server side. The server is more powerful than a normal PC but is much more efficient with resources like CPU, diskspace and memory (most capacity of normal PC’s remains unused). Overall, a thin clientserver environment is cheaper both in terms of hardware and maintainance.

A drawback of the thin client/terminal server model is that if the server fails all the clients which use it become unusable. In high-availability environments this is taken care of by fail-over mechanisms. For most situations it won’t be a big issue since Edubuntu is stable, but it is something to keep in mind. Even without fail-over mechanisms it’s possible with good practices to get the environment back online rather quickly.

see article