Recently, Kenneth Harrenstien developed a portable emulator of the PDP-10 which can run the ITS operating system. Use of this emulator would avoid the problem of maintaining the old hardware, but would still require that one know how to use the ITS operating system; we would have escaped the constraints of old hardware, only to be trapped by old software. Rather than spending the time necessary to maintain an emulator, it would be better for us to spend it translating the data to be saved into a form that can better weather the passage of time. Thus we could eliminate the need for any knowledge of the original operating system, instead of continuing to exert ourselves to port the emulator to every new ``industry standard'' platform. Also, if we used such emulators, we would have to maintain one for each operating system we decided was valuable enough to preserve. Emulators increase in complexity as the original hardware increases in complexity, so this problem only would get more difficult with time. In conclusion, there would be a serious continuing cost that would have to be addressed if this approach were used.