Desktops and the Licensing of Potential Use

Why are desktops treated as single-user devices and subject to simplified user counting rules, when they're as capable of servicing multiple users as servers?

A common argument in software licensing is whether a license is given in respect of actual use or potential use i.e. whether a license is required even if the licensee makes no use of the software. This depends upon what the license agreement says but it is not unusual to find customer and publisher having different views.

This issue was relevant in a recent, now settled litigation in which one of the points of contention was the counting of user licenses for software that was installed on servers. The difference between the two sides’ counts was millions of dollars.

Setting aside the license agreement and the contractual aspects, there was an interesting inconsistency between the approaches to counting potential use for servers and desktops. Users for server-based software were counted based on the security groups to which users are assigned to determine their ability to access the servers and/or the software installed upon them. Desktops and workstations on the other hand were treated as having, and capable of having, only a single user.

This is often not the case. While it may be unlikely for desktops to be used by multiple users, it is not impossible and there are circumstances where it might be routine, such as the use of a desktop in a library or a research lab. Sharing of login credentials aside, the administration of user access for desktops and servers is more alike than might be expected.