WINE, which stands for Wine Is Not an Emulator, is technically a software translation layer. It enables to install and run some software written for Windows® on FreeBSD (and other) systems.

It operates by intercepting system calls, or requests from the software to the operating system, and translating them from Windows® calls to calls that FreeBSD understands. It will also translate any responses as needed into what the Windows® software is expecting. So in some ways, it emulates a Windows® environment, in that it provides many of the resources Windows® applications are expecting.

However, it is not an emulator in the traditional sense. Many of these solutions operate by constructing an entire other computer using software processes in place of hardware Virtualization (such as that provided by the emulators/qemu port) operates in this way. One of the benefits of this approach is the ability to install a full version of the OS in question to the emulator. It means that the environment will not look any different to applications than a real machine, and chances are good that everything will work on it. The downside to this approach is the fact that software acting as hardware is inherently slower than actual hardware. The computer built in software (called the guest) requires resources from the real machine (the host), and holds on to those resources for as long as it is running.

The WINE Project, on the other hand, is much lighter on system's resources. It will translate system calls on the fly, so while it is difficult to be as fast as a real Windows® computer, it can come very close. On the other hand, WINE is trying to keep up with a moving target in terms of all the different system calls and other functionality it needs to support. As a result there may be applications that do not work as expected on WINE, will not work at all, or will not even install to begin with.

At the end of the day, WINE provides another option to try to get a particular Windows® software program running on FreeBSD. It can always serve as the first option which, if successful, offers a good experience without unnecessarily depleting the host FreeBSD system's resources.

This chapter will describe:

Before reading this chapter, it will be useful to: