Preserving the software and hardware environment that was used to access the software when it was created is probably the most intuitive approach. Especially for particularly important examples of software or hardware this appears to be appealing.
Yet, establishing computer museums of obsolete technology is not feasible in the long run. Electronic accessories are prone to a very limited physical life-time. In fact, most electronic hardware is expected to function for no more than 10 to 20 years. Moreover, it must be acknowledged that no archival organisation can hope realistically to maintain such hardware itself [Nat95]. Furthermore, spare parts of an obsolete computer system are unlikely to be supplied in the future. Consequently, any parts that have gone out of order would have to be recreated or makeshift solutions would have to be found. Even if this was technically practicable, it is a very costly process.
Experience at the Phonogrammarchiv5, the oldest sound archive in the world, underlines a negative position towards Technology preservation. Transferring audio data from the original carriers and equipment into the digital domain is considered the only viable way to preserve these recordings for posterity, if speaking in dimensions of centuries [Sch97]. At the same time, representatives from the archive leverage creating a solid structural framework first, and not to hustle for digitisation resulting in inferior quality.
All in all, maintaining computer museums cannot be considered a long-term option for digital preservation.