You can't just focus on longevity of the material medium, because then you'll be carving into clay tablets, which we have actual proof that it can last for millennia. Or maybe that is perfect if you only have a few very important kilobytes to store for millennia, don't need to retrieve data through a digital network, and can trust that there will be people who will know how to decode it.
Longevity is a function of both the material stability of the medium and usability (of storage, of retrieval, of space, of scale, of complexity, of cost). If nobody can decode those bits, it doesn't matter if they are still in the medium. The most optimal storage system is not a single technology, it is a well-funded and well-managed institution. You need people to regularly check integrity of redundant archives and replace individual media when they fail. You also need people to migrate the entire archive to new storage technology when the current generation is growing obsolete. The institution has to regularly buy and maintain drives and computers that can read it, which is often overlooked. Another major issue is longevity and documentation of the data formats and encodings. The bits can still be on the medium, and your drive can still read them, but are you're out of luck if you don't know how it was encoded.
"Tape" is not a single medium, it is a class that includes some formats that are now incredibly difficult to decode. Look at this recent thread about what is needed to read from an old IBM tape format, it's an expensive mess: https://groups.google.com/g/bit.listserv.ibm-main/c/AOlRa0qx...
Longevity is a function of both the material stability of the medium and usability (of storage, of retrieval, of space, of scale, of complexity, of cost). If nobody can decode those bits, it doesn't matter if they are still in the medium. The most optimal storage system is not a single technology, it is a well-funded and well-managed institution. You need people to regularly check integrity of redundant archives and replace individual media when they fail. You also need people to migrate the entire archive to new storage technology when the current generation is growing obsolete. The institution has to regularly buy and maintain drives and computers that can read it, which is often overlooked. Another major issue is longevity and documentation of the data formats and encodings. The bits can still be on the medium, and your drive can still read them, but are you're out of luck if you don't know how it was encoded.
"Tape" is not a single medium, it is a class that includes some formats that are now incredibly difficult to decode. Look at this recent thread about what is needed to read from an old IBM tape format, it's an expensive mess: https://groups.google.com/g/bit.listserv.ibm-main/c/AOlRa0qx...