Another problem which hasn’t been mentioned here is whether data can be understood later. Some data has already been lost becuase it was in a proprietary undocumented format and the only program that can read it is unobtainable or can’t run on any current systems. MS office documents are probably the main concern of that nature looking forward, and using any type of DRM format is just asking to lose it. Once again, the answer is pretty much just being careful and doing a bit of research. Keep on converting data to later formats, or just use simpler, non-binary ones in the first place if the data may be needed past the near future.