... His total lack of knowledge of the subject was obvious by what he thought the 10 primary issues were. He seems ignorant of the serious problem of verifying that a program was copied in its original form (as opposed to have been hacked, or with viruses) or the need for EVERY archival container to have something like an MD5 sum to detect corruption in the future. One of the problems I have as a CHM curator is trying to decide how much of the on line stuff that is out there from the pre-PC time period to snapshot (there's a LOT!). As someone else mentioned, the monoculture and proprietary systems are the real preservation problems, mid 80's or so and beyond. The best I've been able to come up with is to try to save as many SDKs as I can find, which at least gives some level of detail on how they worked.