If you install it they will stay?
Confusing terminology: "Configuring" and "installing" don't have absolute, context-free definitions. In some circles, we install a product and then configure it. In the GNU world, we configure the package (generate the Makefile), then we make install the package. To install the package is to copy the binaries to their appropriate run-time directories, copy the documentation to the doc directory, and maybe let the package manager know what's happened. That's generally considered part of the build process, covered in the last chapter.
For lack of a better term, this chapter describes installing the product. Put more specifically, once we're done with the package manager, we still have to tell FreeTDS about your database servers, and we still have to tell your client programs about FreeTDS.
After FreeTDS has been built and installed, it still doesn't know where your servers are or what particular version of Sybase or Microsoft software each one is using. The purpose of this section is to explain how to describe your dataservers to FreeTDS. FreeTDS has not one but two ways to do this, in freetds.conf and interfaces, described below. It also can use environment variables to override settings in those configuration files. One of the more important (and arcane) settings is the TDS protocol version, described next.